SB Nation User Blog: Florida Danny

Optimism and the 2009 San Francisco 49ers: Will They Make the Quantum Leap?

Florida Danny — Tue, 02 Jun 2009 16:12:00 -0000

AUTHOR'S NOTE: I haven't posted in over a month, and we have a fair amount of newcomers to Niners Nation, so I figure a quick (re)introduction is in order. As you can see, my name's Danny. I'm currently getting my PhD in sport psychology at the University of Florida (but actually reside in Fort Lauderdale, FL...don't ask). I've been a 49er fan since the team's Super Bowl XIX victory over the hometown Dolphins on my 7^th birthday in 1985. My job title on Niners Nation is "Resident Stats Nerd" or something to that effect. During the season, I do a statistical preview and review of each 49ers' game. In addition to these game-specific posts, I also write up general stat-based articles on 49er-related news of the moment. You can find my previous posts here. Hopefully, you enjoy reading my articles, or, at the very least, learn something 49er-related from them. Two warnings, though: (1) My articles tend to be on the longish side; and (2) I fully realize that stats can only go so far in explaining and predicting football performance. If you don't like what the stats seem to suggest, feel free to ignore them. Oh, and one last thing...I mostly rely on defense-adjusted value over average (DVOA) and other alternative statistics developed by Football Outsiders (FO) because they're the best measures of NFL performance currently available to the public. For an explanation of FO's stats, see here.

For the second time in 3 years, there's a good bit of optimism surrounding our beloved 49ers (not here, of course). Many fans (including this one) feel a real sense that 2009 will be the season that ends SF's 6-year playoff drought. The reasons for such optimism are numerous. First, there's the whole, "they finally have a competent head coach" thing. Second, there's the fact that they finally have a real WR corps (thank you Al Davis). Third, they actually seem to be - gasp - tailoring their offensive and defensive schemes around the players' strengths and weaknesses. Fourth, rather than just putting up a words-not-deeds-style banner, they've actually enumerated a "Formula for Success" that tells players and fans exactly how the team plans to "Win the West."

So the 49ers' goal this upcoming season - in the minds of players, coaches, and fans alike - is to make the playoffs. As ARI showed last season - and as PIT showed in 2005 - anything can happen once a team gets in the postseason tourney. In this article, I'm going to address the general question, "Is our optimism justified from a statistical perspective?" More specifically, I'm going to look at the major statistical characteristics of teams that make the quantum leap into playoff participation from one season to the next, and evaluate whether or not the 2009 49ers appear to fit the profile.

After the jump, I'll try to stay optimistic...

IDENTIFYING GROUPS

In the current NFL, there are two groups of teams each season: A group of 12 who make the playoffs and a group of 20 who don't. But what happens to these teams the following season? Do the teams who make the playoffs one season, follow it up with another playoff appearance the next? Conversely, do the teams who miss the playoffs one season, follow it up with another January chock-full of tee times? If we take into account each team's playoff status in Season A, as well as its playoff status the next season, Season B, then we get 4 groups of teams:

Repeaters - teams that made the playoffs in Seasons A and B.
Fallers - teams that missed the playoffs in Season B after making them in Season A.
Leapers - teams that made the playoffs in Season B after missing them in Season A.
No-Shows - teams that missed the playoffs in Seasons A and B.

Using this categorization, there have been 39 Repeaters, 45 Fallers, 45 Leapers, and 94 No-Shows since 2002 (The 2002 Texans aren't included because they didn't play in 2001). As Don Banks of SI noted, these totals indicate that about 6 teams per season make the quantum leap into playoff participation. Based on the fact that the 2008 49ers didn't make the playoffs, this season's incarnation has the potential to be one of these Leapers. Of course, they also have the potential to be a No-Show. Therefore, it's useful to identify what differentiates these two groups. In other words, what changes occur from one season to the next that distinguish teams who make the playoff leap from those who have consecutive seasons ending in Week 17?

Obviously, there are a whole host of usual suspects here. Perhaps the Leapers change coaches, perhaps the Leapers change QBs, perhaps the Leapers get substantially better on offense and/or defense, or perhaps the Leapers simply end up with an easier schedule than the one they played the previous season. Not coincidentally, these are exactly the factors I'll be looking at for the rest of this artricle.

HEAD COACHING CHANGES

It's not news to anyone that the Niners' head coach in Week 1 of this upcoming season will be different from the dapper fellow who manned the sidelines in Week 1 of 2008. In my statistical world, that qualifies as a coaching change. We can quibble about how Mike Singletary was a mid-season replacement last season, but the value of a full offseason and maximum flexibility over assistant coach hires can't be overstated.

Since 2002, 36 of the 139 non-playoff teams had a new head coach in Week 1 of the following season. Not surprisingly, these 36 teams were generally the worst of the worst before the coaching change, especially when it came to passing offense. Specifically, among these 139 non-playoff teams, the 36 teams that changed head coaches averaged about 2 fewer wins (4.8 vs. 6.6), and had an average pass offense DVOA* ranking 4 spots lower, than the 103 non-playoff teams that didn't change coaches. So lesson #1 is that, for the most part, non-playoff teams change head coaches when they (a) suck reeeeeeeeally bad, and (b) can't pass the ball to save their lives. Looks like the Niners were a year late on firing Mike Nolan after all.

So how have these coach-changers fared in the season after making the switch? Well, 10 of the 36, or 27.8%, made the playoffs the next season. Three examples that fit the bill in 2008 were the Dolphins, Falcons, and Ravens. Considering that these new head coaches were taking the helm of sinking ships, these seem like Jack Sparrow-esque performances (without the grog-induced slurred speech). However, to get a valid read on this "leap rate," we need to compare it to the performance of the 103 non-playoff teams who didn't change coaches. In other words, how much did a coaching change matter among non-playoff teams?

It turns out not all that much. The leap rate for coach-keepers was 34.0%, or 35 out of 103. The 6.2% leap rate difference between the coach-changers and the coach-keepers is not even close to being statistically significant, which means that the coaching change had no meaningful impact on making the playoff leap above and beyond what would have been expected from dumb luck. Basically, about 30% of non-playoff teams have made the playoff leap the following season regardless of whether they changed coaches or not, a percentage that corresponds perfectly with the average playoff turnover of about 6 teams per season (i.e., 6 out of 20 is 30%).

Obviously, each specific coaching change occurs in its own context, and the move from Nolan to Singletary is no exception. All the stats tell us here is that, based on the last 7 years of the NFL, our optimism about a 49er playoff appearance in 2009 should not rest on the shoulders of Samurai Mike.

QUARTERBACK CHANGES

Another major change (for the better) this offseason has been the jettisoning of J.T. O'Mulligan as the 49ers' Week 1 starting QB. As was the case with the coaching change, a full offseason of first team reps in OTAs and training camp trumps "he actually became the starter last season" when it comes to classifying "Shaun Hill, 2009 Week 1 starter" as a QB change. Also, in figuring out the "starter" for a given team in a given season, the preferred measure is whether or not he started a majority of the team's 16 games. When there's no majority, the tiebreaker goes to the QB who was the originally designated starter that season. As this relates to the 49ers, neither Shaun Hill (8 starts) nor O'Mulligan (8 starts) reached the 9-start threshold. Therefore, I gave O'Mulligan the "2008 SF starter" classification because he was the originally designated starter last year, and classified "Shaun Hill, 2009 Week 1 starter" as a QB change because, well, he's not O'Mulligan (thankfully).

Since 2002, a whopping 75 of the 139 non-playoff teams had a different starting QB the next season. Not surprisingly, the non-playoff teams that made a QB change were coming off a particularly offensive season on offense (pun intended) when compared with non-playoff teams that didn't end up changing QBs. Specifically, the QB-changers ranked, on average, 5 spots lower than QB-keepers in offense DVOA. What might come as a surprise to some is that these QB-changers were statistically significantly worse in both pass and rush offense. What this says to me is that non-playoff teams change QBs when they experience systematic offensive ineptitude, not necessarily when the passing game is all that's rotten.

So has a QB change made any difference when it comes to non-playoff teams making the quantum leap? Surprisingly, no. Although Chad Pennington, Matt Ryan, and Joe Flacco brought sexy back for their teams in 2008, the last 7 years of NFL football seem to suggest that these were exceptions rather than the rule. That's because, as was the case with coaching changes, about 30% of non-playoff teams made the quantum leap a year later regardless of whether they changed QBs (33.3%) or not (31.2%).

Now, you might be saying, "Danny, not all QB changes are created equally!" You'd be right, so I classified QB changes into 4 groups based on the type of change. Based on this classification, 50 of the 75 non-playoff QB-changers simply got rid of the previous year's starter in favor of a new one (See "2008 New York Jets"), 11 had to replace their previous year's starter due to a major injury (See "2008 Detroit Lions"), 9 benched their previous year's starter in favor of his backup (See "2008 Minnesota Vikings"), and 5 had an established starter return from major injury to reclaim his earlier role (See "2008 Carolina Panthers"). For our purposes, the move from O'Mulligan to Shaun Hill qualifies as your run-of-the-mill "kick ‘em to the curb" type of QB change.

So is there an impact if we look at type of QB change? Once again, the answer is no, at least in terms of statistical significance. However, it's of practical importance to note that the "kick ‘em to the curb" club, which the 2009 49ers will become a member of in September, actually had the worst leap rate of the 4 types (28.0%). Therefore, taken together, the stats I've presented in this section seem to suggest that our optimism about a 49er playoff appearance in 2009 should not necessarily rest on the shoulders of Shaun Hill. You'll see in a moment why I've thrown that "necessarily" in there.

The next two potential differences between Leapers and No-Shows have to do with changes on the field; namely, offensive and defensive performance. I'll tackle offense in the next section (pun intended), and then attack the defense in the section after that (pun also intended).

OFFENSIVE PERFORMANCE CHANGES

So far, I'm probably raining on your playoff parade. Don't fret. Stick with me here. I promise the clouds are going to part very shortly...OK, how about right now?

It shouldn't come as a shock that, since 2002, the 140 non-playoff teams have been worse on offense than their 84 playoff-bound counterparts. Here's a table showing statistically significant differences with respect to offense DVOA:

Category	Playoff Teams	Non-Playoff Teams	Difference
Avg. Pass Offense DVOA	17.1%	-6.0%	23.1%
Avg. Rush Offense DVOA	3.0%	-5.5%	8.5%
Avg. Offense DVOA	10.4%	-5.5%	15.9%

Obviously, playoff teams have been better on offense than non-playoff teams. However, breaking down offensive performance into passing and rushing, the stats suggest that the vast majority of the difference between these two groups is in the passing game. Given that playoff teams are better in the passing game, it stands to reason that the main offensive avenue through which a non-playoff team makes the quantum leap is by vastly improving its aerial attack the next season. To investigate this, let's only focus on non-playoff teams, and examine differences between Leapers and No-Shows with respect to how their respective offensive stats change from Season A to Season B. Here are the relevant stats:

Category	Leapers	No-Shows	Difference
Avg. Pass Offense DVOA Improvement in Season B	+15.8%	+0.5%	15.3%
Avg. Rush Offense DVOA Improvement in Season B	+5.4%	+1.7%	3.7%
Avg. Offense DVOA Improvement in Season B	+10.7%	+1.0%	9.7%

Although the average non-playoff team has improved on offense the following season, those that go from playoff spectator one season to playoff participant the next season (i.e., Leapers) improve almost 11 times more. Furthermore, the vast majority of offensive improvement occurs - as expected - in the passing game. If we compare the differences in pass offense and rush offense improvement, we can conclude the following: If a non-playoff team hopes to make a quantum leap the following season, it's somewhere around 4 times more important for that team to improve its pass offense than its rush offense.

Relating these stats back to the 49ers, their pass offense DVOA last season was a pathetic -18.5%. The average improvement for Leapers, as shown in the table, has been 15.8%. So in order for us to get optimistic about a Niner playoff berth in 2009, we have to ask ourselves, "Do I think the 49ers' passing offense is going to be about as good as the league average (i.e., approximately 0.0%) next season?"

At the moment, I tend to believe that the answer to this question is yes. First, there's the "look what I found" addition of Michael Crabtree. Second, there's the maturation of Jason Hill and Josh Morgan. Third, there's the apparent (and much-needed) change in VD's role. Fourth, there's the return to a run-based offense, which is likely to help the passing game by proxy. Finally, and most importantly, there's the simple fact that the Niners' pass offense DVOA was about 30% better with Shaun Hill starting the final 8 games of 2008.

In total, then, based on the last 7 years of the NFL, our optimism about a 49er playoff appearance in 2009 might indeed rest on the shoulders of Shaun Hill after all. At the very least, it rests on the shoulders of Hill and his WRs.

DEFENSIVE PERFORMANCE CHANGES

As was the case on offense, playoff teams since 2002 have been better on defense than non-playoff teams (Remember, negative DVOA numbers on defense indicate above average performance):

Category	Playoff Teams	Non-Playoff Teams	Difference
Avg. Pass Defense DVOA	-4.8%	7.9%	12.7%
Avg. Rush Defense DVOA	-5.9%	-0.1%	5.8%
Avg. Defense DVOA	-5.3%	4.1%	9.4%

Again, we see here that the passing game - this time on defense - is primarily what separates playoff and non-playoff teams. It stands to reason, then, that, if we focus on the non-playoff teams, Leapers tend to improve their pass defense considerably in Season B, right? Here are the relevant stats (for the sake of clarity, I've made it so that plus signs indicate improvement):

Category	Leapers	No-Shows	Difference
Avg. Pass Defense DVOA Improvement in Season B	+8.2%	-1.5%	9.7%
Avg. Rush Defense DVOA Improvement in Season B	+3.8%	-1.1%	4.9%
Avg. Defense DVOA Improvement in Season B	+6.3%	-1.4%	7.7%

As expected, Leapers improve all aspects of their defense, while No-Shows don't improve any aspects of theirs. Also as expected, the majority of the difference in season-to-season defense DVOA change is in the passing game. The conclusion here is that it's about twice as important for a non-playoff team to improve its pass defense as it is for them to improve their rush defense.

Going back to the 49ers, their pass defense DVOA in 2008 was ranked 20^th at 15.7%. The average improvement for Leapers, as shown in the table, has been 8.2%. So, if we want to determine an appropriate level of optimism, the question becomes, "Do I think the 2009 49ers' pass defense will be half as bad as it was in 2008?"

This one's a little tougher to answer. On the positive side, there's the "it's about freaking time" demotion of Mark Roman, the apparent (and much-needed) change in Manny Lawson's pass rush role, the return of Shawntae Spencer, and the overall establishment of an attacking defensive identity. However, on the negative side, Walt Harris is done for the year, and Singletary's pass defense in the second half of 2008 actually was 23.2% worse than Nolan's during the first half of last season. So I don't know. Based on the stats, it's a tough call. If you want to err on the side of optimism here, be my guest.

STRENGTH OF SCHEDULE CHANGES

The final factor I'll look at that might distinguish Leapers from No-Shows has to do with changes in strength of schedule (SOS) from one season to the next. Specifically, do non-playoff teams make the quantum leap because their schedules get considerably easier?

The answer to this one is a resounding, "Yes!" Of all the comparisons I did in preparation for this article, SOS change had the strongest impact on playoff turnover from season to season.

First, though, it's useful to point out that playoff teams have had statistically significantly easier schedules than non-playoff teams over the past 7 seasons: opponents of playoff teams have had an average winning percentage of 48.0%, whereas opponents of non-playoff teams have had an average winning percentage of 51.2%. You might think that a 3.2% difference isn't that big. Remember, though, that the entire range of SOSs for the 224 teams from 2002-2008 is 41.4% to 59.0%, with 95% of these teams having an SOS between 43.2% and 56.8%. So that 3.2% difference is actually huge, representing a 9-spot difference in SOS ranking (13^th-toughest vs. 22^nd-toughest).

In terms of season-to-season change, the SOS difference between Leapers and No-Shows is almost perfectly symmetrical with that of the playoff vs. non-playoff difference. Specifically, whereas No-Shows' SOSs do not get any easier in Season B, Leapers' schedules get 3.4% easier. Again, that's not a big difference until you look at rankings: a 3.4% easier schedule translates to about a 10-spot difference in the SOS rankings. So the moral of the story here is that, if you want to be optimistic about the 49ers in 2009, hope that their schedule gets a lot easier. But how likely is that?

Well, the Niners' SOS in 2008 was an unbelievably easy 44.7%, meaning that an average Leaper improvement in SOS would take their 2009 SOS into uncharted territory. As I said a minute ago, the easiest schedule over the past 7 years was 41.4%, which the 2007 Seattle Seahawks happily converted into a playoff berth. In this context, a 3.4% easier schedule for the Niners in 2009 means that their SOS for next season would be 41.3%, or, in other words, the easiest schedule of any team since division realignment. So basically, we're banking on one hell of a statistical anomaly here.

Now, you might say, "Well, NFL.com tells me that the 49ers have the 7^th-easiest schedule next year, so maybe they'll pull off that miracle." My response to this is that, as I said in a previous post, a team's SOS based on opponents' records from last season bears no resemblance to their actual SOS based on opponents' records from this season. OK, let me rephrase that. Around 95% of a team's actual SOS has absolutely nothing to do with its before-the-season projected SOS, and projected SOS doesn't exhibit any of the huge statistical relationships that are consistently demonstrated by actual SOS with performance measures like, you know, wins, DVOA, and - particularly relevant to this discussion - playoff turnover. Essentially, you should treat projected SOS like Paris Hilton. It's fun to talk about, but totally meaningless.

BOTTOM LINE

The goal here was to determine whether our optimism about the 49ers in 2009 is justified based on the statistical characteristics of previous teams who made the quantum leap into playoff participation. Here's what the stats suggest:

Despite the MIA-ATL-BAL trifecta in 2008, changing head coaches doesn't seem to forebode an immediate quantum leap into NFL playoff participation. Therefore, don't rest your hopes on the 49ers' change from Nolan to Singletary.
Despite the Pennington-Ryan-Flacco trifecta in 2008, changing QBs also doesn't seem to forebode an immediate quantum leap into NFL playoff participation. Therefore, don't rest your hopes on the 49ers' change from O'Mulligan to Shaun Hill.
On offense, the majority of performance improvement for Leapers is in the passing game. The addition of Crabtree, the maturation of Jason Hill and Josh Morgan, the proper use of VD, the return to offensive balance, and the demonstrated passing game improvement with Shaun Hill at QB over the last half of 2008 all bode well for the Niners. Therefore, optimism seems justified when it comes to pass offense improvement.
On defense, the majority of performance improvement for Leapers is also in the passing game. For the 2009 49ers, some offseason changes bode well (e.g., Dashon Goldson replacing Roman, and Lawson rushing the passer more), while other changes don't bode so well (e.g., Harris blowing out his knee, and the pass defense sucking the big one during the last half of 2008). Therefore, optimism remains in the eye of the beholder when it comes to pass defense improvement.
The primary identifying characteristic of Leapers is that their SOS gets much easier the next season. The problem is that we have no way of knowing just how much easier (or more difficult) the 49ers' schedule is going to be in 2009. Therefore, break out your voodoo dolls and get to work on Kurt Warner et al. Interestingly enough, optimism is directly proportional to how talented you are in Black Magic.

**DVOA statistics used to produce this article were obtained from Football Outsiders.

Poll

How do you think the 49ers' pass defense in 2009 will compare to its pass defense in 2008?

Much better in 2009

444 votes

Slightly better in 2009

538 votes

Just as good/bad in 2009

81 votes

Slightly worse in 2009

20 votes

Much worse in 2009

8 votes

1091 votes | Poll has closed

Movin’ on up in the 2010 Draft: How Low can the 49ers Go?

Florida Danny — Tue, 28 Apr 2009 02:53:37 -0000

AUTHOR' S NOTE: When I discuss QBs in this article, it's just for the sake of argument. What I'm mainly trying to show here is what picks the 49ers will be most able to acquire by packaging their two #1's in 2010 if they choose to do so. Don't get hung up on the QB thing. This stuff can be applied to trading up for any player, not just a QB.

As we all know by now, the 49ers traded their 2^nd- and 4^th-round picks this past weekend for the Panthers'1^st-round pick in next year's draft. The reaction on Niners Nation has definitely been mixed. However, I don't want to get into that argument here. Rather, I'm going to suspend disbelief, and accept Scot McCloughan's reasons for making the deal. One of these reasons is of particular interest to me, and can be found in McCloughan's own words:

As everybody is well aware, ones are huge, especially, if we want to do anything with that pick anytime here out to next year, which of course, going into next draft with two #1 picks.

Grammatical errors aside, one "anything" that McCloughan is likely to want to do with those two #1 picks next year is to package them for a higher (lower?) #1. For the sake of argument, let's assume that the reason to make such a deal is because the 2010 QB draft class looks to be a good one. Below are the top 5 (or 6) QBs in next year's draft according to various "draft gurus" (they're in alphabetical, not rank, order; a * indicates a 2009 junior):

Scout	Bleacher Report	NFL Draft Scout	Kiper
Sam Bradford*	Sam Bradford*	Sam Bradford*	Sam Bradford*
Jimmy Clausen*	Dan LeFevour	Dan LeFevour	Dan LeFevour
Dan LeFevour	Colt McCoy	Colt McCoy	Colt McCoy
Colt McCoy	Zac Robinson	Zac Robinson	Tony Pike
Jevan Snead*	Jevan Snead*	Jevan Snead*	Tim Tebow
Tim Tebow		Tim Tebow

As of right now, it would appear that Bradford and McCoy are the most likely QBs to be selected in the early part of Round 1. This is obviously debatable, so let me repeat that I'm just in "for the sake of argument" mode here. Therefore, if we assume that a good part of the 49ers' increased flexibility is to move up in 2010 for either Bradford or McCoy, the question becomes, "How low (high?) can they go?" Furthermore, "What are the odds that the 49ers can go as low (high?) as possible?"

On Saturday, I briefly addressed these questions in the comments section of Fooch's post detailing the trade. Today, I'm going to go into it much more thoroughly from a statistical probability perspective. Basically, I'm going to provide and discuss the following information:

The pick that an NFL team is most likely to have given its record
For each combination of SF and CAR team records in 2009, the lowest (highest?) possible 2010 pick the 49ers can acquire if they choose to package their two #1s
The odds against SF being able to acquire any single 2010 pick by packaging their two #1s
The odds against SF being able to obtain a top 2 or top 5 pick by packaging their two #1s
The odds against SF being able to obtain any single, top 2, or top 5 pick given reasonable assumptions about SF's and CAR's 2009 records

After the jump, I'll drop some probability knowledge...

TEAM RECORDS AND DRAFT POSITIONS

The goal here is to attach a draft pick probability to each combination of SF and CAR win-loss records. To do that, we first need to figure out the draft position that any NFL team can expect to have given their record during the previous season. For the sake of your brain, I'm not going to go into any detail here. Just believe me when I tell you that, based on the 2004-2008 seasons and 2005-2009 drafts, here are the expected draft positions associated with every possible NFL win-loss record (excluding ties):

Record	Expected Pick
0-16	1
1-15	1
2-14	1
3-13	2
4-12	4
5-11	7
6-10	10
7-9	12
8-8	16
9-7	21
10-6	24
11-5	26
12-4	29
13-3	30
14-2	30
15-1	30
16-0	31

Based on this table, if the 49ers go 10-6 this upcoming season, for example, they can expect to have the 24^th pick in the 2010 draft. Obviously, because not all picks are represented in the table (e.g., there's no #13 in the picks column), there's some variability around the expected pick. However, #24 is the pick the Niners are most likely to have given the NFL's team records and associated draft positions over the past 5 seasons.

DRAFT POSITIONS AND TRADE VALUE

In a must-read post a couple of weeks ago, briandean talked about the draft pick value chart that NFL GMs adhere to - with varying strictness - when talking turkey about trading picks. This chart enables us to attach a value to each expected pick in the above table. For example, the 24^th pick in the draft is worth 740 points according to the chart. Therefore, if the 49ers go 10-6 in 2009, then they'll have 740 points to work with should they choose to deal the pick.

Of course, the assumption here is that the 49ers won't just deal their pick to move up in 2010. Rather, they're likely to trade their pick and the Panthers' pick in exchange for a lower (higher?) pick. So, for example, if SF goes 10-6 and CAR goes 9-7, then the Niners would likely use their 740 points plus the Panthers' 800 points to move up in the draft. And what pick would SF be able to trade for using those 1540 draft pick points? Well, according to the chart, the #7 pick is worth 1500 points, so, if we assume a 40-point surcharge for moving up, then packaging their 21^st and 24^th picks in the 2010 draft would get the 49ers the 7^th pick overall.

This kind of "what if" exercise can be done for any combination of SF and CAR records in 2009. In fact, it can be done for any combination of records for any two teams in any season. Luckily for you, I've already done all the heavy lifting. Below is a chart I created that shows the lowest (highest?) possible pick SF can acquire given (a) the draft picks associated with SF's and CAR's 2009 records, (b) the draft pick values associated with those draft picks, and (c) the assumption that SF will package both picks in a deal to move up in the 2010 draft:

To read this chart, just pick out win totals for SF and CAR, and then look at the number in the box where the two win totals intersect. For example, 2 wins for SF and 14 wins for CAR corresponds to the #1 overall pick being the lowest (highest?) possible pick that the 49ers can acquire were they to do a package deal.

Before going into the odds of any specific pick, let me just point out a couple of interesting things about the information displayed in this chart. First, you'll (hopefully) recall that SF went 7-9 last year and CAR went 12-4. If we look at the chart entry associated with 7 SF wins and 12 CAR wins, we find that a mere repeat of 2008 would allow the 49ers to trade up to the #4 pick. The second interesting thing about this chart is that the worst pick the 49ers can possibly get in a package deal is #12. In other words, not even simultaneous perfect seasons by SF and CAR can prevent the Niners from trading into the top 12 if they choose to do a package deal. OK, let me make the point even clearer. Simply by having two #1s, the 49ers are guaranteed the possibility of a top-12 pick regardless of their 2009 record. Now that's flexibility.

SPECIFIC ODDS COURTESY OF DUMB LUCK

Based on the above chart, you can go through any combination of SF and CAR wins, and see what pick the 49ers can acquire in a package deal. Feel free to do so in your spare time (or on bathroom breaks). For those of us with a life - I'm certainly not included in that - our resident mock draft guru, Dumb Luck, has summarized below the likelihood and odds against the Niners being able to acquire a specific 1^st-round pick in 2010 via a package deal:

Dumb Luck's Crib Sheet
Pick	Likelihood	Odds Against
1	44.29%	1.26 to 1
4	15.57%	5.42 to 1
3	10.03%	8.97 to 1
12	6.23%	15.06 to 1
8	4.84%	19.64 to 1
10	4.15%	23.08 to 1
6	3.81%	25.27 to 1
2	3.11%	31.11 to 1
9	2.77%	35.13 to 1
11	2.42%	40.29 to 1
5	1.38%	71.25 to 1
7	1.38%	71.25 to 1
Top 2	47.40%	1.11 to 1
Top 5	74.39%	0.34 to 1
Picks 6-12	25.61%	2.91 to 1

Before I discuss this table, I should note that its contents aren't magical at all. It's actually pretty simple to figure this stuff out for yourself. For instance, 128 of the 289 boxes in the chart have a 1 in it. Therefore, given all possible combinations of SF and CAR records, the likelihood that the 49ers can get the #1 pick in a package deal is 128/289 or .4429, which is the same thing as 44.29% To translate likelihood into odds against, you just subtract the likelihood from 1, and then divide by the likelihood. In this example, that would be 1 minus .4429, which equals .5571, divided by .4429. That .5571/.4429 is what gets you the 1.26 to 1 that's shown in the table for the #1 pick. Now, onto the discussion...

The first obvious thing that jumps out is that, no matter what happens to SF and CAR in 2009, it's almost even money that the 49ers will be able to trade for the #1 overall pick in a package deal. In other words, flipping a coin right now would basically be just as accurate as actually, you know, playing out the 2009 season in terms of predicting whether or not SF can get the #1 pick in a package deal. The second interesting bit of information in the table is that the Niners have a slightly better shot of getting a top-2 pick in the 2010 draft, which guarantees them Bradford or McCoy if those two 2 QBs - for the sake of argument - are projected as the top 2 picks next year. Finally, notice that SF has almost a 75% chance of moving into the top 5 via a package deal. When translated into odds, the 0.34 associated with a top 5 pick means that the odds are actually 3 to 1 in favor of the Niners.

SPECIFIC ODDS COURTESY OF BONZO THE IDIOT MONKEY

For those who weren't around this weekend, Bonzo the Idiot Monkey is Cold Hard Football Facts' resident mock draft guru. Being a close relative to humans, he's able to engage in some low-level reasoning. For instance, based on the information given to him by his handlers, Bonzo was able to deduce that Matt Stafford would be selected by the Lions. In contrast, Dumb Luck got that pick wrong because, being a mere calculator, he wasn't equipped to incorporate that kind of information into his prognostications.

Why am I bringing up Bonzo? Well, it's because Dumb Luck's table omits a couple of basic assumptions that only a reasoning creature could use to his/her advantage:

We already learned from my earlier table that, if a team goes 4-12 or worse, then their expected pick is already in the top 5. Therefore, if either SF or CAR wins fewer than 5 games in 2009, the Niners won't need to package their two 1^st-round picks to move into the top 5. They'll already be there.
The '07 Patriots and '04 Steelers aside, no teams have gone 16-0 or 15-1 in the past 5 seasons. CAR did win 12 games last season. However, just as the old NFL adage says that "it's easier to go from 6 wins to 8 than it is to go from 8 wins to 10," the same can be said of going from 12 wins to 15 or 16. Indeed, since the NFL schedule increased to 16 games, only 4 teams have gone 15-1 (including the 49ers in 1984), and only 1 has gone 16-0 (David Tyree anyone?). Therefore, it's unbelievably unlikely that CAR or SF - especially SF - will win 15 or more games in 2009.

So, being the low-level logician that he is, Bonzo's table is a little bit different from Dumb Luck's because he focused only on summarizing the possible picks associated with reasonable expectations about SF's and CAR's records in 2009. Specifically, here's Bonzo's table, which only looks at the bold, inset 5-to-14-wins box of the chart I presented earlier:

Bonzo's Crib Sheet
Pick	Likelihood	Odds Against
4	33.00%	2.03 to 1
3	15.00%	5.67 to 1
8	10.00%	9.00 to 1
6	7.00%	13.29 to 1
9	6.00%	15.67 to 1
10	6.00%	15.67 to 1
2	5.00%	19.00 to 1
11	5.00%	19.00 to 1
5	4.00%	24.00 to 1
7	4.00%	24.00 to 1
12	4.00%	24.00 to 1
1	1.00%	99.00 to 1
Top 2	6.00%	15.67 to 1
Top 5	58.00%	0.72 to 1
Picks 6-12	42.00%	1.38 to 1

Comparing Bonzo's table to Dumb Luck's, we can see how having reasonable expectations about SF's and CAR's records affects the likelihood that the Niners can acquire a given 2010 1^st-round pick in a package deal. First, the most likely obtainable pick changed from #1 to #4. Furthermore, the Niners went from having an even shot at the #1 pick in Dumb Luck's table, to having a 99-to-1 shot in Bonzo's table, which makes the #1 pick now the least likely of the top 12. This no doubt is due to the fact that almost all record combinations in which SF can get the #1 pick are associated with either SF or CAR already having a pick in the top 5. Second, if as before we assume for the sake of argument that Bradford and McCoy are going to be the top 2 picks in the 2010 draft, then Bonzo (about 16 to 1) is far more pessimistic than Dumb Luck (about even money) with respect to the 49ers getting either QB by virtue of a package deal. Finally, although SF still has a likelier-than-not chance of trading into the top 5, the odds against their ability to do so doubled given Bonzo's reasoning.

BUYER BEWARE

Before wrapping things up, I'd just like to caution those of you out there who might want to consult SF's and CAR's strengths of schedule (SOS) in 2009 when experimenting with possible record combinations. In other words, I would advise against paying any attention to the fact that, based on their opponents' 2008 records, SF has the 7^th-easiest 2009 schedule, while CAR has the 2^nd-most difficult. Here's why: A team's SOS before the season, which is based on its opponents' previous-season records, has no practical relationship with its actual SOS or its actual win-loss record once the season has played out. In other words, just because the Panthers have the 2^nd-most difficult 2009 schedule as of right now, don't think for even one second that (a) their schedule is actually going to be that tough, if tough at all, or (b) their 2009 record is actually going to be any worse than if they had the easiest before-the-season SOS. Basically, it turns out that we can't predict anything about CAR's 2009 SOS or how many games they're going to win in 2009 based on what their 2009 opponents did in 2008.

Just in case you don't believe me, or you don't want to just trust me on this, here is a table that shows the correlations between before-the-season SOS, actual SOS, and actual win-loss record for each of the past 5 seasons:

Year	Pre-SOS vs. Actual SOS	Pre-SOS vs. Actual W-L	Actual SOS vs. Actual W-L
2004	0.10	-0.06	-0.36*
2005	0.13	0.05	-0.66**
2006	0.19	-0.15	-0.59**
2007	0.16	0.15	-0.34
2008	0.41*	0.08	-0.56**
Average	0.20	0.01	-0.50

For the non-stat folks out there, an asterisk (*) in the table means there's a 95% likelihood that the given relationship is real (i.e., it's not a statistical mirage), while two asterisks (**) means there's a 99% likelihood. Also, see here for my previous description of what the direction (+ or -) and value of a correlation mean. If these last two sentences made your brain explode, just look for asterisks and numbers farther away from 0 (provided you can still read despite brain detonation).

As the table shows, 2008 was the only season (of the last 5) in which a team's actual SOS was related at all to its before-the-season SOS. Of course, even this 2008 relationship is pretty useless in terms of prognostication when you consider the fact that a team's before-the-season SOS in 2008 was totally unrelated to how many games it actually won. Furthermore, all those asterisks and bigger numbers in the right-most column, which suggest a consistent statistical relationship between actual SOS and actual win-loss record, are pretty useless as well because we have to wait until after the season to calculate them. In other words, we can't use actual SOS to predict a team's record for the upcoming season because we can't find out their actual SOS until after the season.

Finally, some clever NN readers out there might think it's necessary to adjust these correlations for each team's record in the previous season because (a) the NFL schedule is created, in part, based on where each team finished in the standings; and (b) winning teams one year are more likely to be winning teams the next year. Well, first off, the latter isn't actually the case: The correlation between team wins from one year to the next is not statistically significant. More importantly, though, I tested out this adjustment, and it didn't change the above table in any meaningful way.

BOTTOM LINE

Based on everything that I've said in this article, here's what I think is most important to remember between now and the 2010 draft vis-à-vis the Niners packaging their two 1^st-round picks:

Given all possible combinations of SF and CAR wins during the 2009 season, it's even money that the 49ers will be able to move on up to the #1 pick.
Given all possible combinations of SF and CAR wins during the 2009 season, it's even money that the 49ers will be able to move on up to #1 or #2. Also, the odds are 3-to-1 in favor of the 49ers being able to trade into the top 5.
Assuming that both SF and CAR win between 5 and 14 games during the 2009 season, the most likely pick the 49ers will be able to move on up to is #4.
Assuming that both SF and CAR win between 5 and 14 games during the 2009 season, the odds are 94-to-1 against the 49ers being able to move on up to #1 or #2. Also, the odds are 3-to-2 in favor of the 49ers being able to trade into the top 5.
Don't pay any attention to before-the-season SOS when trying to predict the positions of SF's two 1^st-round picks.

Poll

Assuming they don't already have a top-5 pick by virtue of their 2009 record (or CAR's), should the 49ers try and package their two 1st-rounders in the 2010 draft to select a QB in the top 5?

Yes

242 votes

No

412 votes

654 votes | Poll has closed

So Easy a Monkey Could Do It: Mocking the Mockers

Florida Danny — Sat, 25 Apr 2009 02:42:25 -0000

Hey everyone...just a quick post here before the draft gets started...

We have our own resident draft guru here on NN. His name is briandean. Fooch doesn't pay him anything to drop knowledge on Niners Nation; and that's too bad. It's too bad because ESPN now has two resident draft gurus, Mel Kiper, Jr., and Todd McShay, and they're getting paid a lot of money. Funny thing is...I bet anyone 10-1 on their money that the difference in mock draft accuracy between briandean and the ESPN guys isn't going to even come close to approaching the difference in level of paid compensation. In fact, I'll go ahead and predict right now that briandean's 1^st-round mock accuracy is going to be within 6 correct picks of Mel or Todd's.¹ Here's the simple truth about why:

Kiper and McShay aren't much better than Bonzo the Idiot Monkey.

In true me form, I did a little calculation to get to the bottom of this...

If someone - anyone - was to mock the 1^st round only knowing the names of the 32 players that will be selected, the odds against that person picking 7 or more correctly - simply by dumb luck - are about 20,000 to 1. To put that in perspective, it's the same chance you have of penciling in a (legitimate) 1 one your golf scorecard after any given hole.

Of course, Mel and Todd are getting paid big bucks to know more than just the names of the 32 players. Mel, for example, claims to spend 24 hours a day for 361 days per year - he says he takes 4 days off - gathering information to increase his draft knowledge base. In contrast, Dumb Luck spends exactly 0 hours a day gathering draft knowledge. So the question really is, "If Fooch were to feel generous and pay Dumb Luck $1/yr for mocking the 1^st round each draft, is ESPN getting its money's worth with Mel and Todd?" Well, the answer is obviously no because - and I'm going out on a limb here - Mel and Todd's salaries are way higher than $20,000/yr. In other words, the pick accuracy benefit doesn't even come close to outweighing the "draft guru" cost.

And if you think evaluating Kiper and McShay based on exact picks is a little too harsh, take a look at the gold standard of mock draft evaluation: The Huddle Report's Mock Draft Scores. As their scores show, Mel was tied for 29^th, and Todd was tied for 59^th, among the 94 big-time mock drafters in 2008. And their rankings among 47 mock drafters based on average scores from 2006-2008? Kiper 9^th and McShay 22^nd. Incidentally, Mike Mayock was tied for 75^th in 2008, and is 13^th-best over the past 3 drafts.

So, when you watch the draft today, whether on ESPN or NFL Network, keep in mind that the guys who spend all year researching this stuff are - dollar-for-dollar - not much better at actually, you know, predicting picks than Dumb Luck and Bonzo the Idiot Monkey.

¹ Just in case you were wondering, here's a table showing the odds against Dumb Luck getting a given number of picks (or more) right provided he knows the names of the 32 1^st-rounders:

Number Correct	Odds Against
1 or more	1 to 1.7
2 or more	2.8 to 1
3 or more	11.9 to 1
4 or more	57.6 to 1
5 or more	336.2 to 1
6 or more	2,386.5 to 1
7 or more	20,335.4 to 1
8 or more	204,934.2 to 1
9 or more	2.4 million to 1
10 or more	32.8 billion to 1

So, what, you figure that Mel and Todd make somewhere in the neighborhood of $2M/yr in ESPN salary, right? If that's true, then, based on these odds, Mel needs to outpick Dumb Luck 9-1 in order to give ESPN their money's worth; 10-1 and Mel needs his agent to renegotiate that contract, stat!

Poll

Which of the following mock drafters do you think is going to be the most accurate in 2009?

Mel Kiper, Jr.

32 votes

Todd McShay

16 votes

Mike Mayock

31 votes

briandean

12 votes

Bonzo the Idiot Monkey

41 votes

Dumb Luck

23 votes

155 votes | Poll has closed

A Statistical Look at Drafting QBs: LCF Reduxe

Florida Danny — Fri, 24 Apr 2009 16:41:44 -0000

Back in October, I introduced the Lewin Career Forecast (LCF) to Niners Nation (NN). To put it mildly, the post stirred up some lively debate, and signaled the beginning of a still-raging internecine battle on NN over the appropriateness of football stat analysis. Just to refresh everyone's memory, the LCF, which was developed by David Lewin of Football Outsiders (FO), predicts a QB's NFL performance based on 2 of his college stats: games started (GS) and completion percentage (COMP%). A couple of other ancillary features of the LCF are that (a) it's most appropriate for predicting the NFL performance of 1^st- and 2^nd-round QB picks, and (b) 37 college GS and a 60% college COMP% are the implied statistical benchmarks for success.

Because of its (over)simplicity, the LCF has received a healthy amount of criticism since it was unveiled in Pro Football Prospectus (PFP) 2006. Whether we're talking about statistical modeling or plain ol' common sense, it just can't be right that predicting performance for an NFL QB only requires knowing 2 things about him (3 if you count his draft round). In other words, QB performance is way too complicated to be that simple. Well, at least that's what the critics say. There are other arguments against the LCF, but I'll get into those a little later.

So, I figured that, because the draft is tomorrow, and because the "QB at #10" rumors seem to be accelerating, now would be as good a time as any for me to revisit the LCF and evaluate - from a statistical perspective - just how useful it is to the average NFL fan (Warning: Just because I'm a stats guy, don't assume I'm going to conclude that LCF is über-useful). Therefore, in this article, I'm going to do 4 things:

Discuss the statistical and non-statistical positives and negatives of the LCF
Give a statistical rationale for a solution that addresses the negatives
Detail my solution
Apply my solution to the top QB prospects in the 2009 NFL Draft

After the jump, I'll tackle the LCF...

POSITIVES AND NEGATIVES

In order to explore this topic, we need some more background info about the LCF (Note: I'm going to try as hard as possible to write the next couple of paragraphs in English).

Basically, to come up with the LCF, Lewin first thought of as many college stats as he could that might influence how a QB is going to perform at the professional level. Some of these were performance-based (e.g., COMP%), while others were more trait-like (e.g., height). To measure "performance at the professional level," he chose defense-adjusted points per game above replacement (DPAR/G). DPAR was the FO predecessor to DYAR, which is a stat most of you should be familiar with if you've read my previous articles. After choosing DPAR/G as his performance measure, he got all of the relevant stats for every QB drafted from 1997-2006, and looked for statistically significant relationships between these stats.

Based on the result that GS and COMP% were the stats most related to DPAR/G, and that the DPAR/G relationships for other stats seemed to depend on GS (e.g., a QB who starts more games generally throws for more yards and TDs by definition), he ran an analysis called multiple linear regression (MLR) to come up with a simple equation that predicts DPAR/G from GS and COMP%. As part of this analysis, he ran an MLR for each round of the draft, and found that the 1^st and 2^nd rounds yielded the most accurate predictions. Now, he's able to simply plug in any college QB's GS and COMP% stats, and get a predicted value for DPAR/G, which he calls the LCF.

As for this magical equation he came up with, think of it in terms of fantasy football (FF). To figure out how many points your QB scored in a given week according to standard FF scoring, you just divide his passing yards (PAYDs) by 25, add to that his passing touchdowns (PATDs) multiplied by 4, and then subtract his interceptions (INTs) multiplied by 2. Really, all you're doing here is applying the following equation to the QB's stats: FFPts = (.04 x PAYDs) + (4 x PATDs) - (2 x INTs). The MLR-based equation for LCF is just this sort of thing.

Now, onto the positives and negatives...

The first good thing about the LCF is that the inclusion of GS and COMP% makes perfect intuitive sense. The more GS a college QB has, the more experience he has playing QB. Obviously, if you're going to hire a guy to run your company (or offense), it's better that he/she has more experience than less. It means he/she is more likely to know what the (site decorum) he/she is doing. Similarly, the higher a college QB's COMP%, the better he is at hitting his target. Obviously, the point of the passing game in football - at any level - is to move the ball down the field by, you know, completing the pass to your teammate. If a guy can't hit the broadside of a barn in college after having thrown 1,000s of passes in his life up to that point, then he's probably not going to become a sharpshooter when he goes to the pros.

The second good thing about the LCF is more statistical in nature. When a statistician wants to determine whether someone did an analysis correctly and reported accurate results, they do what's called a replication study. Well, I did one of those for this article. Except for (a) using data only from 1^st- and 2^nd-round picks, and (b) using DYAR/G instead of DPAR/G, I did the same thing that Lewin did to come up with the LCF, i.e., I followed the steps I described a minute ago. To Lewin's credit, I got the same results he did: GS and COMP% had the strongest relationships with DYAR/G, and were the best MLR predictors of DYAR/G. Just for the sake of providing info, here are some of the other things that I found:

The correlation between GS and DYAR/G was .71. The correlation between COMP% and DYAR/G was .35. Correlations range from -1 to 1, with the size of the number indicating how strong the relationship is. Positive correlations mean that one stat goes up as the other goes up, whereas negative correlations mean that one stat goes down as the other goes up. So, for all intents and purposes, these 2 correlations mean that a higher GS is twice as important as a higher COMP% when it comes to predicting higher DYAR/G.
Because some people have made this argument, I looked at two variables related to the NFL situation a QB draftee was entering at the time he was drafted. Whether we're talking about the previous year, the year before that, the year before that, or the average of these 3 years, neither a team's wins nor its offense DVOA when its QB arrived from college were related at all to his future NFL stats (COMP%, PAYDs, PATDs, INTs, TD-INT ratio, DYAR/G).
For sh*ts and giggles, I also looked at whether coming from a BCS school has any impact. It doesn't.
The actual equation for LCF that I arrived at in my replication study was: NFL DYAR/G = -232.16 + (2.75 x college GS) + (250.31 x college COMP%).¹ Plug in some QB numbers and try it for yourself if you wish.

Now for the negatives, of which there are 3...

First, as you can tell from what I just discussed, coming up with the LCF requires a good bit of statistical acumen. In terms of deriving it, the outcome variable, DYAR/G, is a little complicated, and the analysis itself requires statistical software and MLR training. In terms of applying it, you need to have the equation handy and probably a calculator. Because of these issues, LCF is not really practical for the average football fan. For example, if you're having a football argument with a friend about some random QB prospect, it's not really practical for you to keep a copy of Lewin's equation in your wallet and a calculator in your pocket (more likely your pocket protector) in order to rebut your friend's claim that "QB X is going to be awesome in the pros!" Likewise, you're probably going to elicit bewilderment rather than agreement if you base your rebuttal on DYAR/G. And even if you simply wait for Lewin's predictions in PFP or on FO's website, you're probably not going to have the predictions handy when arguing with your friend.

Second, as some people have pointed out, the fact that the LCF is only accurate for 1^st- and 2^nd-rounders poses a problem with respect to applying it. In order for it to actually be a "prediction," you have to know in advance whether or not a given college QB is going to be drafted in the 1^st or 2^nd round. This limits your prediction window to the month or so prior to the draft, in which draft pick projections reach a consensus among teams and pundits like Mel Kiper, Jr. And just so you know, this whole "diminishing accuracy by round" thing is real. When I ran my MLR analysis looking only at 1^st-rounders, the model was about 5% better; which brings me to the last problem.

The third problem with LCF is by far the most concerning. When it comes down to it, statistics is about explaining the variation among things. It answers questions like, "Why does Jane have an IQ 20 points higher than John?" Jane's IQ varies from John's by 20 points, and we want to know why. In the context of the LCF, this example can be rephrased, "Why does Peyton Manning (the 1^st pick in 1998) have an NFL DYAR/G 161 yards higher than Ryan Leaf (the 2^nd pick in 1998)?" More generally, the question is, "Why do certain NFL QBs have a higher DYAR/G than others?" The LCF answers these latter 2 questions by saying, "Because of the differences between their GS and COMP% in college."

When you run an MLR analysis, the results tell you a lot more than the multipliers you need to get your prediction. The most important of these supplementary results is called explained variance, which, in the context of LCF, tells us how much of the DYAR/G variation between top-2-round QBs is explained by their college GS and COMP% stats. In my replication of Lewin's analysis, that explained variance value was 56.2%. In other words, over 40% of the DYAR/G variation between QBs was not explained by GS and COMP%. While 56.2% isn't shabby in MLR, especially given the small sample size (n = 35), there's still a veritable sh*t-ton of variation for which LCF has no answer.

Going back to the previous example, the LCF basically says that it can account for about 90 yards (or 56.2%) of the 161-yard difference between Manning and Leaf, but has no clue about the other 70 or so yards. Indeed, if you plug their college numbers into the equation I gave you earlier, you get a predicted DYAR/G difference of 78.94 (Manning = 48.85; Leaf = -30.09), which is far below their actual 161-yard difference. You might say, "Well, it still accounted for half of the difference." My response would be this: Leaf's legacy would be considerably different if he ended up being only 80 DYAR/G worse than Manning. In fact, his actual DYAR/G would increase from -62.24 (worst among the draftees) to 16.70 (on par with Jake Plummer).

And if you thought that wasn't enough, there's one more variation-related problem with LCF: variation in the multipliers themselves. Recall that what you get out of an MLR analysis is a bunch of numbers you have to multiply your stats by to get your predicted DYAR/G. It turns out that these multipliers, called parameters, are estimates just like your DYAR/G prediction is an estimate. In other words, they're statistics-based educated guesses. Think of political polls for this one. Just like every poll has a margin of error, every one of the MLR parameters has a margin of error. Just like the true value for a candidate's vote percentage is somewhere within that margin, the true value for the MLR parameter is somewhere within that margin. Obviously, the goal of estimation, whether we're talking QB performance or political polls, is to minimize your margin of error as much as possible. Later on, I'll tell you how that's done; though I bet you already have a clue.

So what does this "margin of error" problem mean for LCF? First, it means that Lewin gives you his LCF estimate, but doesn't give you his margin of error, which is the same thing as a pollster reporting their prediction for a candidate's vote percentage, but not telling you their margin of error. Second, and much more importantly, it turns out that the amount of parameter estimate variation for the replication MLR I did was huge. Here's a perfect example that is relevant for this year's draft:

According to scout and pundit consensus, Matt Stafford, Mark Sanchez, and Josh Freeman are going to be the only QBs taken in the first 2 rounds of the 2009 draft. Stafford's stats were 33 GS and 56.9% COMP% (or .569 for the purposes of LCF). Sanchez's stats were 15 GS and .643 COMP%. Freeman's stats were 35 GS and .591 COMP%.
Remember that the MLR parameter estimates for GS and COMP% (from the earlier equation) were 2.75 and 250.31, respectively. When you take each of their margins for error into account, the true GS parameter value is anywhere from 1.81 to 3.68, while the true COMP% parameter value is anywhere from 57.54 to 443.08.
Without going into detail here, know that the first number in the equation (-232.16), which is called the intercept, ranges from -349.26 to -115.07 when you take its margin of error into account.
Let's see what the LCF predictions are when the parameter estimates in the equation are exactly right.
Let's also see what the LCF predictions are when the true parameters for GS, COMP%, and the intercept are all at the low end of their respective ranges; for example, 2 for GS, 60 for COMP%, and -340 for the intercept. In other words, let's use the following "low but just-as-likely" equation for LCF: DYAR/G = -340 + (2 x GS) + (60 x COMP%).
Let's also see what the LCF predictions are when the true parameters for GS, COMP%, and the intercept are all at the high end of their respective ranges; for example, 3.5 for GS, 440 for COMP%, and -120 for the intercept. In other words, let's use the following "high but just-as-likely" equation for LCF: DYAR/G = -120 + (3.5 x GS) + (440 x COMP%).

Below is a table showing the differences between DYAR/G predictions using the 3 different LCF equations (low-end, exact, and high-end):

College QB	Low-End LCF	Exact LCF	High-End LCF
Stafford	-239.86	0.88	245.86
Sanchez	-271.42	-30.22	215.42
Freeman	-234.54	11.88	262.54

So basically, using 3 different-but-equally-likely LCF equations, it's just as likely that all 3 QBs are going to be 4 times worse than Leaf in the NFL as it is that they're going to be about 3 times better than Manning. Here's the same table using the range of parameter estimates in an LCF equation that predicts NFL COMP% instead of DYAR/G:

College QB	Low-End LCF	Exact LCF	High-End LCF
Stafford	7.98%	55.66%	93.72%
Sanchez	5.86%	53.98%	92.44%
Freeman	8.82%	57.37%	96.28%

Not all is lost here, though. If you're arguing with your friend about how good Matt Stafford is going to be as a pro, the odds are great that the argument will end when you say, "I think he's going to have a completion percentage somewhere between 8% and 94%." Well, either the conversation will end or your friend will come back with the proverbial, "No sh*t, Sherlock!" Remember, although the exact LCF is obviously the best estimate when it comes to suppressing this reflex, it's just as likely from a statistical perspective that the low-end LCF and high-end LCF predictions end up being right!

I want to make something very clear here before moving on to my potential solution to these problems. My discussion of LCF negatives are in no way meant to suggest that the LCF is useless. Rather, my purpose here is to make it more useful for the average football fan; basically to remove the mystery that surrounds it. Furthermore, I am not saying that Lewin and FO are trying to be deceptive about variability. On the contrary, FO freely acknowledges the variability inherent in their statistics and predictions. The fact of the matter is that football stat analysis is prone to vast amounts of variability, a reality that I'll explore a little bit later. What's important to understand here is that the variability problem with LCF is a byproduct of football stats themselves, not Lewin's or FO's desire to assert certainty where there's uncertainty.

BOTTOM LINE: Here are the main points you should take away from my discussion of the LCF's strengths and weaknesses:

The LCF is a DYAR/G prediction based on a QB's college GS and COMP%. It's based on a stat-analysis-derived equation that quantifies the relative influence of these 2 factors.
The LCF is most accurate for QBs drafted in the 1^st and 2^nd rounds.
The inclusion of GS and COMP% in the LCF makes intuitive sense from a non-statistical perspective.
The inclusion of GS and COMP% in the LCF is justified from a statistical perspective, as is the exclusion of other seemingly important college stats.
Based on my replication study, a QB's predicted NFL DYAR/G = -232.16 + (2.75 x college GS) + (250.31 x college COMP%).
The LCF has an outcome variable, DYAR/G, that might make the average football fan cross-eyed.
Based on LCF, over 40% of "NFL success" for a QB has nothing to do with his college GS and COMP%.
The margin of error for an LCF prediction is very big.
PFP is not engaged in statistical money laundering with respect to LCF.

SOLUTION IDEA

OK, so the main problems with the LCF - in terms of everyday application - are that it uses a potentially confusing measure of "NFL success," and that its prediction equation yields wildly different results when you account for the margin of error. If this is the case, it seems to me that the way to improve the LCF for average-football-fan consumption involves (a) using a more widely known outcome measure, and (b) side-stepping the prediction equation.

In terms of NFL outcome measures, I've mentioned several already in this article: COMP%, GS, PAYDs, PATDs, INTs, and TD-INT ratio. As college COMP% and GS are the two stats that make up the LCF, it makes sense to try NFL COMP% and GS as replacements for DYAR/G. Also, in the same way that college PAYDs, PATDs, and INTs are dependent on GS, the NFL versions of these stats are also dependent on GS; as is NFL TD-INT ratio. Therefore, GS is a viable candidate both in its own right and as a proxy for PAYDs, PATDs, INTs, and TD-INT ratio.

To evaluate whether NFL COMP% and GS are good replacements for DYAR/G, we need to find out how much of the variance in COMP% and GS - as compared to DYAR/G - is accounted for by college COMP% and GS. Here's a table summarizing the results:

NFL Performance Measure	Performance Variance Explained
DYAR/G	56.2%
COMP%	45.6%
GS	20.9%

So it turns out that college COMP% and GS actually account for less of the variation in NFL COMP% and GS than they do for DYAR/G. This means that NFL COMP% and GS are worse than DYAR/G when it comes to choosing an outcome measure for a modified version of the LCF. NFL GS in particular is woefully bad because, as the table shows, almost 80% of a QB's NFL GS has nothing to do with his college COMP% and GS.

Because NFL GS is so bad, perhaps one of those outcomes we tossed out for being dependent on GS might work better. And here's a thought: What if we used NFL FFPts/G as a way to incorporate these outcomes into one summary measure? Well, that's exactly what I did. So how much of the variation in NFL FFPts/G is accounted for by college COMP% and GS? The answer is 48.8%, which means that -although still not as good as DYAR/G - FFPts/G is actually a more useful measure of NFL success than are COMP% and GS. So the moral of the story here is that, if we accept the cost of losing 8% in explained variation, we can gain the benefit of using a more widely known outcome measure. Going back to my QB argument example, your friend will probably know what 10 FFPts/G means in terms of performance, while he probably won't have a clue what 10 DYAR/G means; so that's definitely a plus.

Now, how can we side-step the prediction equation altogether so that we don't have to worry about the LCF's margin-of-error problem? Well, perhaps we can use some other non-MLR system that gives us a prediction that's at least as accurate as LCF.

IDEAL SOLUTION

My solution here involves using categorization to replace prediction equations. Here's what I mean. Suppose we take the general benchmarks set by the LCF (37 GS and 60% COMP%), and group the 35 QBs who were taken in the first 2 rounds of the draft from 1997-2006 into the following categories:

Group A: QBs that had 37 or more GS and a COMP% of 60% or higher
Group B: QBs that had 37 or more GS, but had a COMP% lower than 60%
Group C: QBs that had less than 37 starts, but had a COMP% of 60% or higher
Group D: QBs that had less than 37 starts and a COMP% lower than 60%

Here's how the group membership plays out:

Group A	Group B	Group C	Group D
Ben Roethlisberger	Cade McNown	Aaron Rodgers	Akili Smith
Byron Leftwich	Carson Palmer	Alex Smith	Charlie Batch
Chad Pennington	Donovan McNabb	David Carr	Jim Druckenmiller
Daunte Culpepper	Jake Plummer	Jason Campbell	Joey Harrington
Drew Brees	Jay Cutler	Kellen Clemens	JP Losman
Eli Manning	Kyle Boller	Rex Grossman	Marques Tuiasosopo
Matt Leinart	Shaun King	Tim Couch	Michael Vick
Peyton Manning		Vince Young	Patrick Ramsey
Phillip Rivers			Quincy Carter
			Ryan Leaf
			Tarvaris Jackson

Now, rather than focusing on an exact prediction of NFL success, let's just predict some general standards of performance for each group. For instance, let's say that QBs in Groups A and B should have 10 or more FFPts/G (with rounding), whereas QBs in Groups C and D should have less than 10 FFPts/G (Aside: 10 FFPts works out to about 200 yards, 1 TD, and 1 INT). And for the sake of comparison, let's also say that QBs in Groups A and B should have 20 or more DYAR/G (with rounding), whereas QBs in Groups C and D should have less than 20 DYAR/G.² Below is a table showing the groups and their respective stats:

QB	Group	GS	COMP%	FFPts/G	DYAR/G	LCF FFPts/G	LCF DYAR/G
Peyton Manning	A	45	62.9%	16.06	99.23	12.47	48.85
Drew Brees	A	37	61.0%	14.25	62.87	10.06	22.13
Daunte Culpepper	A	44	63.9%	13.53	43.73	12.53	48.61
Phillip Rivers	A	51	63.5%	12.84	55.92	14.03	66.83
Ben Roethlisberger	A	38	65.5%	12.01	47.93	11.61	36.14
Chad Pennington	A	51	63.4%	11.48	53.82	14.00	66.58
Eli Manning	A	38	61.1%	11.36	25.89	10.32	25.12
Byron Leftwich	A	38	65.1%	9.72	25.91	11.49	35.14
Matt Leinart	A	39	64.8%	*7.63*	*13.24*	11.64	37.13
Carson Palmer	B	41	59.1%	14.14	67.66	10.43	28.36
Jay Cutler	B	45	57.2%	13.59	63.22	10.80	34.58
Donovan McNabb	B	49	58.4%	13.20	33.86	12.07	48.57
Jake Plummer	B	45	55.4%	10.43	*15.34*	10.27	30.08
Kyle Boller	B	40	48.0%	*7.66*	*-5.21*	6.95	-2.18
Shaun King	B	39	55.5%	*7.14*	*1.47*	8.91	13.85
Cade McNown	B	43	55.5%	*6.02*	*-12.20*	9.84	24.84
Aaron Rodgers	C	22	63.8%	*11.42*	*36.09*	7.42	-12.05
Jason Campbell	C	30	64.6%	*10.66*	*34.14*	9.50	11.92
Tim Couch	C	24	67.1%	9.15	-20.73	8.84	1.70
Rex Grossman	C	32	61.0%	8.57	-6.19	8.91	8.40
David Carr	C	26	62.8%	8.02	-15.65	8.05	-3.57
Vince Young	C	32	61.8%	6.74	0.70	9.14	10.40
Alex Smith	C	23	66.3%	6.29	-47.25	8.38	-3.05
Kellen Clemens	C	32	61.0%	4.30	-9.71	8.91	8.40
Joey Harrington	D	28	55.2%	9.06	-8.60	6.28	-17.10
Michael Vick	D	21	56.5%	8.65	-6.39	5.05	-33.07
Patrick Ramsey	D	32	58.9%	8.35	-4.05	8.29	3.14
Quincy Carter	D	29	56.6%	8.09	0.95	6.93	-10.85
Charlie Batch	D	24	58.0%	7.62	10.24	6.18	-21.08
JP Losman	D	27	57.8%	7.44	-16.45	6.81	-13.34
Tarvaris Jackson	D	36	53.6%	7.27	3.48	7.66	0.86
Ryan Leaf	D	24	54.4%	5.23	-62.24	5.13	-30.09
Akili Smith	D	19	56.6%	3.75	-56.73	4.62	-38.31
Marques Tuiasosopo	D	25	55.4%	1.24	-11.23	5.65	-24.84
Jim Druckenmiller	D	24	53.8%	0.93	-43.00	4.95	-31.59

Stats displayed in bold italics are cases in which the general performance prediction was wrong. With this in mind, here's how accurate you would have been at predicting NFL success if you simply used group membership and general performance standards:

Predicting QBs in Groups A and B to average 10 or more FFLPts/G: 12 of 16 (75.0%)
Predicting QBs in Groups A and B to average 20 or more DYAR/G: 11 of 16 (68.8%)
Predicting QBs in Groups C and D to average less than 10 FFLPts/G: 17 of 19 (89.5%)
Predicting QBs in Groups C and D to average less than 20 DYAR/G: 17 of 19 (89.5%)

Those accuracy rates aren't too shabby in the least. Although you'd be more accurate about the curds (i.e., Groups C and D) than about the whey (e.g., Groups A and B), simply predicting a general standard of performance based on group membership is pretty accurate overall. Now, for the sake of comparison, let's see how you would have done if you applied the stat-based predictions. In other words, let's see how many times the exact prediction of over (and under) 10 FFLPts/G and/or 20 DYAR/G was right. The last 2 columns show the exact predictions for FFPts/G and DYAR/G. Stats displayed in bold are cases in which the exact performance prediction was wrong. Here are the accuracy rates:

Predicting a given QB to average 10 or more FFLPts/G: 13 of 15 (86.7%)
Predicting a given QB to average less than 10 FFLPts/G: 19 of 20 (95.0%)
Predicting a given QB to average 20 or more DYAR/G: 11 of 14 (78.6%)
Predicting a given QB to average less than 20 DYAR/G: 19 of 21 (90.5%)

And what if you take group membership into account when evaluating exact prediction accuracy?

Predicting QBs in Groups A and B to average 10 or more FFLPts/G: 14 of 16 (87.5%)
Predicting QBs in Groups A and B to average 20 or more DYAR/G: 13 of 16 (81.3%)
Predicting QBs in Groups C and D to average less than 10 FFLPts/G: 18 of 19 (94.7%)
Predicting QBs in Groups C and D to average less than 20 DYAR/G: 17 of 19 (89.5%)

Obviously, the stat-based exact predictions are going to be slightly better than the non-stat-based general predictions because - as you'll recall - the margin of error associated with the exact prediction method (i.e., LCF equation) is slightly better than the unknown error associated with the general prediction method (i.e., grouped performance standards). Nevertheless, the point here is that the accuracy cost doesn't outweigh the applied benefit. In other words, what you lose in accuracy, you make for by actually being able to come up with a prediction quickly and articulate that prediction using language your friend understands.

One thing I'll add here before moving on is that my solution complies entirely with FO's acknowledgement that variability - indeed, a lot of it - exists in their statistics and predictions. As they've taken great pains to admit, the aim of their stats is to show that some players (or teams) are better than others; not that their stats are the Da Vinci Code of football performance. In other words, they don't claim that knowing a secret code means being able to exactly predict the future (or describe/explain the past). Therefore, because our aims and variability admissions are similar, the main advantage of my solution is that it addresses the, "WTF is DYAR/G?" problem.

APPLYING THE SOLUTION

So I think I've come up with a much more practical application of the LCF that preserves Lewin's basic finding about the exclusive importance of college GS and COMP%. My goal here was not to rip Lewin to shreds. In fact, as I said earlier, his analysis was spot-on from both a methodological and results perspective. Unfortunately, the LCF just suffers from a serious variation problem.

Basically, Lewin's analysis was a slave to sample size. The reason why there's such a huge margin of error associated with the LCF is because the analysis that produced it was based on data from only 35 QB draft picks. It's actually quite remarkable that he was able to obtain such strong relationships between DYAR/G, GS, and COMP% with such a small sample size. Therefore, given the strength of these relationships, there's no doubt in my mind that the margin of error will shrink as future QB draft picks are added to the data set. Indeed, the main way to reduce variability in MLR predictions (and parameters) is to increase your sample size. There's no statistical magic here: The laws of probability dictate that educated guesses get better and better as the amount of information they're based on increases (i.e., the more they're educated).

Nevertheless - and I hope you're reading this methodrampage - the analysis is what it is. The state of the LCF at this point in time dictates that we have to wait for more data (i.e., wait for the sample size to get big enough). As I said, I don't doubt that future data will vindicate LCF. I just can't get behind it right now as a valid application for the average NFL fan (or NN member).

With that said, let's apply my more practical application to this year's draft. As I mentioned earlier, Stafford, Sanchez, and Freeman are, by consensus, the QBs who will be selected in the first 2 rounds. Here are their stats, group membership, and predicted performances:

QB	Group	GS	COMP%	10+ FFPts/G	20+ DYAR/G	LCF FFPts/G	LCF DYAR/G
Matt Stafford	D	33	56.9%	No	No	7.94	0.88
Mark Sanchez	C	15	64.3%	No	No	5.94	-30.22
Josh Freeman	D	35	59.1%	No	No	9.04	11.88

Based on my general standards of performance, these 3 QBs are likely to fail with respect to having 10+ FFPts/G during their NFL careers. If this doesn't give the Lions pause about taking Stafford at #1, I don't know what will. Likewise, if this doesn't give the 49ers, NN members, and pundits pause about taking Sanchez at #10, I don't know what will. Serviceable QBs perhaps; just not top-10 picks. And if we want to quantify just how unlikely it would be for Stafford or Sanchez to end up average 10+ FFPts/G in their NFL careers, all we have to do is go back to the accuracy rates I presented earlier: 17 of the 19 Group C and D QBs from 1997-2006 averaged less than 10 FFPts/G during their careers. This means that, based on previous history, there's only about a 10% chance that Stafford and Sanchez buck the trend. Notice that I'm not saying a 0% chance; it's just highly unlikely. If it happens, it doesn't mean the stats were wrong. It just means these QBs defied the odds, and there's nothing inherently wrong with that (esp. for their bank accounts).

Although I obviously prefer the general predictions, if we compare the slightly more-accurate exact predictions in this table with the earlier one detailing prior draft picks, we get some sobering results. Stafford is closest to Tarvaris Jackson and David Carr in predicted FFPts/G, and closest to Jackson and Tim Couch in predicted DYAR/G. Sanchez is closest to Marques Tuiasosopo and Charlie Batch in predicted FFPts/G, and closest to - wait for it - Ryan Leaf and Jim Druckenmiller in predicted DYAR/G. I'm going to pause for a few seconds and let that sink in.................................................................................OK, done vomiting; back to the show. Finally, Freeman is closest to Vince Young and Shaun King in predicted FFPts/G, and closest to Jason Campbell and Vince Young in predicted DYAR/G.

So, based on exact predictions, the QBs should be ranked Freeman, Stafford, Sanchez. However, as is the major take-home of this article, we can't be so trusting of the exact predictions. The better thing to do - from a practical perspective - is simply to say that none of the 3 QBs are going to meet 2 general standards of good NFL performance. In this way, we can replace the exact rankings with general groupings that are just as accurate overall.

And if you want to really win that argument with your friend, then instead of saying, "Matt Stafford is going to have a completion percentage between 8% and 94%," or worse yet, "Matt Stafford is going to have a DYAR/G between -240 and 245," say, "I bet you Matt Stafford doesn't score more than 10 fantasy football points per game during his career." If your friend took that bet, you'd be a winner almost 90% of the time. Feel free to donate those proceeds to NN. Fooch needs the money (just kidding, Fooch).

¹ Let me stress that this is the LCF equation that I came up with based on my replication study. It's probably not the same exact equation used by Lewin and FO in their publications because Lewin has no doubt refined the equation since 2006 as more data has come in. Also, unlike me, he has full access to all FO stats that aren't on the website, along with all of their game charts, etc. The point is that if you want Lewin's actual LCF predictions, you should rely on Pro Football Prospectus and FO's website, rather than my replication-derived equation. And if you're wondering whether this is a copyright CYA on my part, it is.

² I used 10 FFPts/G and 20 DYAR/G as general standards because (a) that's what you get when you plug 37 GS and 60% COMP% into their respective equations, and (b) it makes the group membership accuracy as close as possible to the exact prediction accuracy.

**DYAR and DVOA statistics used to produce this article were obtained from Football Outsiders.

Poll

Of Matt Stafford, Mark Sanchez, and Josh Freeman, which QB do you think is going to average at least 10 fantasy football points per game during his NFL career (that's about 200 yards, 1 TD, and 1 INT or better per game)?

All of them

20 votes

Stafford and Sanchez only

44 votes

Stafford and Freeman only

7 votes

Sanchez and Freeman only

12 votes

Stafford only

22 votes

Sanchez only

81 votes

Freeman only

26 votes

None of them

84 votes

296 votes | Poll has closed

The Art (and Science) of Drafting: IV. 49er vs. Patriot Regimes in the Salary Cap Era

Florida Danny — Wed, 22 Apr 2009 22:23:55 -0000

In Part 3 of this series, I compared the 49ers' overall draft strategy in the Salary Cap Era to that of the Patriots. Obviously, the problem with my analysis in Part 3 is the same one we encountered in Part 1. Namely, evaluating 15 years worth of picks, whether for SF or NE, and comparing two teams with respect to those picks, is akin to putting apples and oranges (i.e., different regimes and team contexts) into one basket and then comparing that basket to one containing bananas and grapes. So Part 3, in reality, was just a way of uncovering general differences between the Niners and Patriots that might explain their overall differences in winning percentage and playoff success.

The better analysis is to look at specific regimes under similar contexts. Because we're technically still in the McNolan regime until the 2009 draft, it's probably of most interest for you (and I) to see how McNolan compares to specific Patriot regimes. Beginning with the 1994 draft, there have been 3 such regimes:

Bill Parcells, 1994-1996
Bobby Grier, 1997-1999
Scott Pioli, 2000-2008

As I mentioned during the introduction to Part 3, it turns out that NE has had overall team contexts similar to those of SF. All that's different are the dates on the calendar. Specifically, Parcells was the Bill Walsh I of NE, transforming a hapless franchise into a consistent contender. Grier was a combination of Policy and Donahue in that he unwound the pristine ball of yarn spun by Parcells. Finally, Pioli was a mixture of Walsh II and McNolan. He was charged with returning a once-proud franchise to its winning ways.

So the question is, "Which of the 3 NE regimes has McNolan been most similar to regarding draft strategy?"

But before I move on, I need to make one additional regime distinction. Further inspection of the Pioli regime makes you realize that it has actually encompassed two different team contexts, i.e., it's 2 different regimes rolled into 1. Pioli's drafts from 2000-2003 were of the dynasty-building variety, whereas his drafts from 2004-2008 were of the dynasty-sustaining variety. As my discussion of the Policy and Donahue regimes vis-à-vis the Walsh regime suggested, rebuilding drafts look a lot different than sustaining drafts. So when attempting to compare McNolan to the NE regimes, we have to consider Pioli as having had 2 different regimes (Pioli I & Pioli II) from 2000-2008. In other words, my question above revises to, "Which of the 4 NE regimes has McNolan been most similar to regarding draft strategy?"

To refresh your memory, here is the McNolan draft strategy I detailed in Part 2d of this series:

Stockpile picks when your team sucks, but trade away picks when (you think) your team is good.
Draft for need on Day 1, especially when your team sucks.
Use Day 2 to take a lot of bites at the apple.
BCS or die!
We're Goldilocks (aka Vanilla McNolan).

After the jump, I'll compare McNolan's draft strategy from 2005-2008 to 1 of the 4 Patriot regimes of the Patriots. Keep reading to find out which one. The lucky winner might surprise you...

OVERALL REGIME COMPARISON

My preparation for this article put me in one of those pesky scientific situations wherein you expect one result and get an entirely different one. Anyone who does research of any kind always has to guard against what's called "confirmation bias," which, as the name implies, is the tendency to cherry-pick evidence or construct your research in a way that confirms what you were hoping to find in the first place. Plenty of researchers fall into this trap - especially when money and professional prestige are at stake - which actually is antithetical to the scientific method. The whole point of top-down research is to develop a hypothesis and try to prove it wrong, not right.

With that said, I was fully expecting that, given its mediocre record since 2005, McNolan's drafts would be somewhere in the middle between the least successful Patriot regime (Grier) and the most successful (Pioli). Also, given the disparity in team success despite similar contexts, I was expecting there to be a vast difference between the McNolan and Pioli I drafts. Needless to say, this wasn't the case.

Here is a table showing the differences in various draft categories between McNolan and the other regimes that have drafted for SF and NE in the Salary Cap Era:

	DIFFERENCES WITH MCNOLAN
CATEGORY	POLICY	WALSH II	DONAHUE	PARCELLS	GRIER	PIOLI I	PIOLI II
Picks per Round
Picks per Day
Picks per Position
Picks per Unit
Picks per Conference	✓	✓✓		✓	✓
BCS Rate	✓	✓✓✓✓		✓			✓✓
Starter Rate
Pro Bowl Rate	✓✓			✓
Picks per Unit on Day 1		✓✓✓		✓
Picks per Unit on Day 2				✓	✓
BCS Rate on Day 1					✓
BCS Rate on Day 2	✓✓✓	✓✓✓✓		✓✓	✓✓✓		✓✓
Starter Rate on Day 1
Starter Rate on Day 2			✓✓✓		✓
Pro Bowl Rate on Day 1
Pro Bowl Rate on Day 2	✓✓✓					✓
BCS Rate on OFF		✓✓					✓
BCS Rate on DEF	✓✓	✓✓✓
Starter Rate on OFF
Starter Rate on DEF		✓		✓✓
Pro Bowl Rate on OFF
Pro Bowl Rate on DEF	✓			✓
Starter Rate from Non-BCS				✓	✓✓
Starter Rate from BCS		✓✓✓		✓
Pro Bowl Rate from Non-BCS	✓
Pro Bowl Rate from BCS				✓✓
✓	4	1	0	8	4	1	1
✓✓	2	2	0	3	1	0	2
✓✓✓	2	3	1	0	1	0	0
✓✓✓✓	0	2	0	0	0	0	0
Total ✓s	14	22	3	14	9	1	5

Each check mark (✓) indicates a statistical difference between McNolan and a given regime with respect to a given category. The number of check marks indicates the size of the statistical difference:

✓ = There's an 80% likelihood that the regime's drafts were statistically different.
✓✓ = There's a 90% likelihood that the regime's drafts were statistically different.
✓✓✓ = There's a 95% likelihood that the regime's drafts were statistically different.
✓✓✓✓ = There's a 99% likelihood that the regime's drafts were statistically different.

Looking at the totals for the 49er regimes, it's pretty depressing that (a) McNolan has been least similar to Walsh II, (b) McNolan has been most similar to Donahue, and (c) the only statistical difference between McNolan and Donahue is that Donahue was actually better at finding starters on Day 2 (Donahue = 29.6%; McNolan = 8.6%). Stewardess, can you please get me a barf bag?

Looking at the totals for the Patriot regimes, however, paints an entirely different picture. In this case, it turns out that McNolan has actually been most similar to the rebuilding half of the Pioli regime (i.e., Pioli I); to an even larger extent than McNolan's similarity to Donahue. Of the 26 categories, the only area in which McNolan has differed from Pioli I has been the latter's relative effectiveness in finding Pro Bowlers on Day 2, which we all know by now is a little skewed because of the better team having a built-in advantage for getting players selected to the Pro Bowl.

Some of you might notice that there seems to be an "era trend." The Policy regime and Parcells were equally different from McNolan, and they happened to be drafting during the same period of time (the beginning of the Salary Cap and Free Agency Eras). Also, McNolan has been most similar to Donahue and Pioli (both I and II), who happen to be the closest in proximity with respect to draft years. However, I'd caution against this interpretation because, as the table shows, Walsh II and Pioli I were polar opposites with respect to their McNolan differences despite overlapping draft years (Walsh = 1999-2001; Pioli I = 2000-2003) and similar draft contexts (i.e., rebuilding). And if you're wondering whether this discrepancy is due to the fact that Walsh II focused on rebuilding the DEF, whereas Pioli I and McNolan focused on rebuilding the OFF, just look at the table categories involving team units: only 4 of Walsh II's 22 check marks are due to unit differences with McNolan. This means that the unit-adjusted difference for Walsh II is 18 check marks, which is still the largest of all regimes, and dwarfs Pioli II's 1-check-mark McNolan difference.

So, the moral of the table is that, despite my (and probably your) expectations to the contrary, McNolan's drafts have actually been most similar to Pioli I's. Given this result, I'll be using the rest of the article to compare McNolan's draft strategy to Pioli's. Odds are, of course, that their draft strategies are quite similar. However, there might be specific areas in which the 2 regimes have differed, with these differences shedding light on the answer to our now-more-important question, "Why was Pioli I so much more successful than McNolan despite being so similar in the draft?" I'm sure you've already developed an opinion or two.

PIOLI I IN CONTEXT

When Bill Belichick hired Pioli in time for the 2000 draft, the Patriots had seen a Donahue-esque decline in team wins during the previous 3 seasons. Although they didn't exactly nosedive like Donahue's, the Patriot win totals during Grier's tenure nevertheless got worse each season (10 in 1997, 9 in 1998, and 8 in 1999). Essentially, just like Donahue's squandering of Walsh II's success, Grier squandered Parcells' success. And just like McNolan was charged with cleaning up Donahue's mess, Pioli I was charged with cleaning up Grier's mess.

I've already told you that Pioli I focused primarily on rebuilding the NE OFF - and with good reason. Below are the specific contexts of each NE draft during the Pioli I regime:

2000 Draft - Key Losses: RB Terry Allen (free agency), WR Shawn Jefferson (free agency), TE Ben Coates (free agency), OL Heath Irwin (free agency), OL Zefross Moss (released), and DB Steve Israel (free agency). Weak 1999 Stats: 22^nd in OFF DVOA.¹ Lingering Issues: Still hadn't found a worthy replacement for RB Robert Edwards; major injury to LB Ted Johnson. Needs: RB, WR, TE, OL, DB, LB depth.

2001 Draft - Key Losses: FB Tony Carter (free agency), OL Bruce Armstrong (retirement), DL Chad Eaton (free agency), and LB Chris Slade (free agency). Weak 2000 Stats: 20^th in OFF DVOA; 20^th in DEF DVOA. Lingering Issues: Still hadn't found worthy replacements for RB Robert Edwards and TE Ben Coates; Ted Johnson missed 3 games with minor injury; DB Otis Smith now 35 years old. Needs: RB, FB, TE, OL, DL, LB, DB depth.

2002 Draft - Key Losses: QB Drew Bledsoe (trade), WR Terry Glenn (trade), TE Rod Rutledge (free agency) and DL Brandon Mitchell (free agency). Weak 2001 Stats: -3.7% (14^th) OFF DVOA. Lingering Issues: Still hadn't found worthy replacements for RB Robert Edwards and TE Ben Coates; DL Willie McGinest missed 5 games with minor injury; Otis Smith now 36 years old. Needs: RB, WR, TE, DL, QB depth, DB depth.

2003 Draft - Key Losses: FB Marc Edwards (free agency) and DB Tebucky Jones (trade). Weak 2002 Stats: None. Lingering Issues: Major injury to Otis Smith, who is now 37 years old; DL Anthony Pleasant and LB Roman Phifer now 34 years old. Needs: FB, DB, DL depth, LB depth.

In this case, the specific needs accurately reflect the general: NE primarily needed to rebuild the OFF. However, looking at these draft contexts, our first difference emerges. Namely, Pioli I had to replace 4 starters per draft, whereas McNolan had to replace about 7 starters per draft. Media accounts of both regimes (here and here) invoked the proverbial "out with the old, in with the new" theme. Looking back, it's more accurate to say that McNolan had much more old and new to take out and bring in, respectively.

So how did Pioli I replace their 16 departed starters? Here's how:

RB Kevin Faulk (1999 Grier pick) for Allen
WR Troy Brown (1993 Parcells pick) for Jefferson
Rutledge (1998 Grier pick) for Coates
OL Joe Andruzzi (2000 Pioli I free agent signing) for Irwin
OL Grant Williams (2000 Pioli I free agent signing) for Moss
Otis Smith (2000 Pioli I free agent signing) for Israel
Marc Edwards (2001 Pioli free agent signing) for Carter
OL Matt Light (2001 Pioli pick) for Armstrong
DL Anthony Pleasant (2000 Pioli I free agent signing) and DL Bobby Hamilton (2000 Pioli I free agent signing) for Eaton (after switching to a 4-3 defense)
LB Mike Vrabel (2001 Pioli I free agent signing) for Slade
QB Tom Brady (2000 Pioli I pick) for Bledsoe
WR David Patten (2001 Pioli I free agent signing) for Glenn
TE Christian Fauria (2002 Pioli I free agent signing) for Rutledge
Hamilton for Mitchell (after moving him from DE to DT)
TE Daniel Graham (2002 Pioli I pick) for Marc Edwards (after switching to a 2-TE offense)
DB Eugene Wilson (2003 Pioli I pick) for Tebucky Jones

Adding it all up, Pioli I rebuilt the starting lineup with 9 traditional free agents, 4 draft picks, and 3 players originally acquired by previous regimes. As compared to McNolan's first two drafts, which required a similar 15 replacements, Pioli I used free agency nearly twice as much, and had a holdover rate (20.0%) that was half as large. Therefore, we've found a second difference between the two regimes: Pioli I liked free agency and cleaned house more.

Now, let's move on to the point-by-point draft strategy comparison. Again, although McNolan and Pioli I drafts are similar, there are nevertheless important differences.

STRATEGY #1

Stockpile picks when your team sucks,

but trade away picks when (you think) your team is good.

As I reported in Part 2d, McNolan has had 35 picks in its 4 drafts. In comparison, Pioli I had 36 picks in its 4 drafts; very similar indeed. So, the way to evaluate this strategy is to compare the two regimes' draft picks by year after matching up the number of draft picks at identical junctures during their respective regimes. The table below shows McNolan's and Pioli's pick totals during their first, second, third, and fourth drafts:

Remember that Pioli I drafts followed seasons with 5 wins (1999), 8 wins (2000), 11 wins (2001), and 9 wins (2002). Regarding number of draft picks, Pioli I's totals did seem to ebb and flow with team wins. Indeed, Pioli I's number of draft picks was lowest in Year 3, which followed an 11-5, Super-Bowl-winning 2001 season. Where a difference emerges relates to my placement of "(you think)" in the McNolan strategy. In true Vanilla McNolan form, SF had only 6 picks in Year 4 presumably because they thought the disastrous 2007 season was a mere aberration; the Niners were actually good in McNolan's mind, only requiring a new OC and a few draft picks to return to their 7-9 "glory" of 2006.

Contrast this with the much-more-realistic Pioli I approach. That regime only got rid of picks after winning the Super Bowl. Furthermore, they returned to the "stockpile picks" approach after a winning season. In other words, what I see here is Pioli I having a much higher threshold than McNolan for "your team is good." It seems that McNolan viewed a 5-11 season as only requiring some roster- and assistant-coach-tweaking, whereas Pioli I responded to a 9-7 season by amassing 10 picks in the 2003 draft. I guess the easiest way to put this is that, to Pioli I, 9 wins meant "we suck," but, to McNolan, 5 wins means "we're good."

STRATEGY #2

Draft for need on Day 1, especially when your team sucks.

OK, so I just established that "when your team sucks" meant "less than 10 wins" to Pioli I. Therefore, let's look at Day 1 of NE's 2000, 2001, and 2003 drafts. Referring back to the team needs I detailed earlier, the 2000 draft required Pioli to fix the OFF; specifically, RB, WR, TE, and OL. Their Day 1 picks? An OL (Round 2) and a RB (Round 3). In fact, their first 4 picks were an OL, a RB, an OL (Round 4), and a TE (Round 5). Not to mention that they stumbled upon a certain unassuming, jelly-legged QB from the University of Michigan in Round 6. Of course, that was luck more than anything attributable to Pioli I, but it's still worth mentioning. Anyway, addressing needs on Day 1 in 2000? Check.

In 2001, most of the needs were still on OFF. The positions they took on Day 1 of that draft were DL (Round 1), OL (Round 2), and DB (Round 3). So, did Pioli I address needs on Day 1 in 2001? No. One thing I will say though is that, if you're not going to address needs on Day 1, it helps to draft a Pro Bowl DE (Richard Seymour) and a Pro Bowl LT (Light) with your top two picks, and then snag your 2002 starter at RT (Kenyatta Jones) with your first pick on Day 2.

In 2003, the needs shifted to DEF because of advancing age and increasing proneness to injury. Specifically, NE needed to replace a starting DB, and needed depth at DL and LB. What did they take during Day 1 of the 2003 draft? Round 1: a DL. Round 2: a DB. Round 3: a WR. And just for good measure, they took another DL in Round 4. Addressing needs? Check.

Last question: "Did Pioli I also draft for need on Day 1 of the 2002 draft, i.e., when NE was good?" Even in their Super-Bowl-winning season of 2001, NE was still a below-average offensive team (See above OFF DVOA). At the very least, their 14^th-ranked OFF was worse than their above-average DEF (DVOA = -6.3%). Therefore, their needs were on OFF. So what did Pioli I do on Day 1 in 2002? He took a TE in Round 1 and a WR in Round 2, and then, for good measure, took a QB with NE's first pick of Day 2. Addressing needs? Check.

So, taken together, Pioli I seems to have been very similar to McNolan regarding the draft strategy of interest in this section.

STRATEGY #3

Use Day 2 to take a lot of bites at the apple.

My identification of this McNolan strategy in Part 4 was based on its tendency to take a lot of WRs, DLs, and DBs on Day 2. In addition, I argued that their reliance on Day 2 talent at these positions might be one reason why they haven't found many good ones. Hey, but at least they're trying.

From the table showing regime differences that I presented at the beginning of this article, you can already surmise that Pioli I didn't differ from McNolan with respect to team units on Day 2. Therefore, there's no general trend. It's still useful to look at the specific position-by-position draft day totals, as well as the actual unit-by-unit results. Here's the relevant table:

Pos	Day 1	Day 2	Total
QB	0	3	3
RB	1	2	3
FB	0	0	0
WR	2	1	3
TE	1	4	5
OL	2	3	5
DL	2	5	7
LB	0	3	3
DB	2	4	6
K	0	1	1
P	0	0	0
OFF Total	6	13	19
DEF Total	4	12	16
ST Total	0	1	1
Total	10	26	36

From this table, it's clear that Pioli I didn't have a unit bias between draft days: he split his picks 60/40 between OFF and DEF on Day 1, and basically split them 50/50 on Day 2. Nevertheless, there are two (albeit shaky) pieces of evidence of a Day 2 "bites at the apple" strategy having been used by Pioli I. First, they sure had a lot of picks on Day 2, meaning that they were taking a lot of bites at the apple all over the field. Of course, it makes you wonder how NE won a championship during Pioli I with all that Day 2 talent on the roster. Oh yeah, I forgot. Two words: free agency. OK, well, 5 words: free agency and Tom Brady. Second, you might recall from Part 2d that McNolan has had 21 picks on Day 2 in their 4 drafts, i.e., 5 fewer than Piol I. So how did Pioli I use those extra picks? How about at those needs on OFF? Specifically, Pioli I used 4 of the 5 extra Day 2 picks on OFF (and the 5^th on a K), while taking the exact same number of DEF picks on Day 2 as McNolan.

So, in general, I'd say the differences between McNolan and Pioli I are pretty subtle with respect to this strategy, if they exist at all. One thing I will add, though, is that, unlike McNolan, Pioli I found 3 Pro Bowlers at need positions on Day 2: Brady, OL Dan Koppen, and DB Asante Samuel. Indeed, the table of regime differences shows that Day 2 Pro Bowl rate was the only category in which a regime difference actually emerged between McNolan and Pioli I. In this sense, I suppose it was a good thing that NE had 15 Day 2 picks - almost 2 extra per round than normal - in the 2000 and 2003 drafts.

STRATEGY #4

BCS or die!

This one's a pretty easy nut to crack. Pioli I used 28 of his 36 picks on players from BCS conferences, which is pretty similar to McNolan using 31 of 35 on such players. The only slight difference between the two regimes - certainly not large enough to be statistically significant - has been with respect to draft days. Whereas their Day 1 BCS rates were pretty similar, Pioli's Day 2 BCS rate (23.1%) was more than double McNolan's (9.5%). Again, this difference isn't statistically significant, but it is a practically important difference nevertheless.

STRATEGY #5

We're Goldilocks (aka Vanilla McNolan).

As this strategy is more about attitude than statistics, it's not surprising that this is where McNolan and Pioli I seem to differ the most. Remember that my Goldilocks argument was based on the fact that McNolan has had perennial needs that it continues to leave unaddressed (i.e., pass rusher and QB). Essentially, once McNolan thinks it has addressed a position in the draft, say, by taking Alex Smith with their 1^st pick, it doesn't feel an urgent requirement to reassess whether, perhaps, they adequately addressed the need. Pioli I was the exact opposite in this regard.

The dirty secret about NE's 2000 draft is that it was pretty (site decorum) bad. Their 2^nd-round pick, OL Adrian Klemm was a certified bust. Their 3^rd-round pick, Redmond, was basically a strikeout swinging. Their 4^th-round pick, OL Greg Randall, only ended up being a regular starter for 1 season. Of their remaining 7 picks, all Pioli I got out of them in terms of starters and/or Pro Bowlers was Brady; granted, he certainly makes up for the rest of the Day 2 garbage. Essentially, only the Policy regime's 1995 and 1997 drafts are in the same solar system in terms of being bad from top-to-bottom.

The difference, though, is that Pioli I realized really quickly the enormity of his 2000 disaster. After failing with his OL picks in 2000, he took another 2 OL with his first 4 picks in 2001. After failing with his DL picks in 2000, he took a DL with his first pick in 2001. This trend continued in subsequent drafts, as Pioli I took a TE with his first pick in 2002 after missing on 2 TEs in the previous 2 drafts. You get the idea. The point here is that there seems to have been an honest reassessment of previous draft failures at need positions. In essence, you don't just wait for a pick to be what you thought he would be. Rather, you quickly say, "he sucks," and address the need again immediately.

The second example of Pioli I not being "vanilla" is something I mentioned earlier. Namely, he didn't settle for mediocrity after NE went 9-7 in 2002. Rather, that 2003 draft ended up being what took the Patriots to the next level, allowing them to escape the mire of mediocrity. The proof of the pudding is in the eating. With their first 6 picks, NE found 4 starters (DL Ty Warren, Eugene Wilson, Samuel, and Koppen), the latter 2 of which made the Pro Bowl with NE. And what did the Patriots do in 2003 and 2004? They won two championships, of course. So the point here is that, whereas McNolan looked at a 5-win team and made a few changes to get back to 7-9, Pioli I looked at a 9-7 team and made a whole host of changes to get back to the Super Bowl. If this isn't the difference between vanilla and mint chocolate chip, I don't know what is.

Here's one last thing about this. Although I'm focusing on Pioli I here, his commitment to excellence (rather than mediocrity) is an identifiable trait because of how consistently he's acted on it. Specifically, not settling for mediocrity continued well into the Pioli II regime of 2004-2008. Here are a few examples:

He amassed 10 picks in the 2006 draft after a Divisional Round exit in 2005.
Despite being minutes away from reaching another Super Bowl in 2006, he traded for WR Randy Moss and signed WR Wes Welker after realizing that NE's mediocre WRs were holding them back.
He realized that a certain Bay Area team was not as good as they thought they were, and traded NE's late-1^st-round pick in 2007 in exchange for that team's earlier 4^th-round pick and sure-to-be-earlier 1^st-round pick in 2008. All this trade produced was NE having the #7 pick in the draft coming off of an undefeated regular season.

This last example was the difference between vanilla and mint chocolate chip in a nut shell.

GENERAL ANALYSIS AND CONCLUSION

To wrap things up, below is a table summarizing ways in which Pioli I's draft strategy was similar to and different from McNolan's:

MCNOLAN STRATEGY	PIOLI I SIMILAR?
Stockpile picks when your team sucks, but trade away picks when (you think) your team is good.
Draft for need on Day 1, especially when your team sucks.
Use Day 2 to take a lot of bites at the apple.
BCS or die!
We're Goldilocks (aka Vanilla McNolan).	X

As I said during the introduction, what I found out in my regime comparisons was not what I expected. The first thing that surprised me was that McNolan and Pioli I have been pretty similar. More accurately, though, I was shocked that McNolan's drafts have been most similar to a regime that won 10 more games and 1 more Super Bowl in the same context and period of time. This revelation implied an unexpected finding that was even more shocking to me. Namely, that NE's success during Pioli - and Pioli II for that matter - was minimally attributable to their draft success. Rather, from a personnel perspective, NE won in 2001 for 2 reasons:

Tom Brady
Free agency

Although this is certainly arguable, I'm sure the vast majority of you - and a plurality of Patriot fans - would agree that there is no Patriot dynasty without NE discovering what NFL Network deemed the greatest "diamond in the rough" in NFL draft history. It's pretty clear that their other draft picks in 2000 were pretty useless, so it stands to reason that, without Brady, NE's upset over the Rams in Super Bowl XXXVI never happens (p.s. I'm ignoring the Tuck Rule here). Same goes for all those free agent acquisitions following the 1999 and 2000 seasons.

Where the draft has helped NE the most has been in sustaining their success past 2001. Without that 2003 draft supplementing a few free agent acquisitions (e.g., DB Rodney Harrison), I'm pretty sure the Patriots don't win rings in 2003 and 2004.

The bottom line is that, whether through free agency or the draft, NE hasn't sat on its laurels and accepted mediocrity. Basically, 9-7 has meant "we suck." That is the lesson that Donahue never learned, and it seems to be the modus operandi for McNolan thus far. Simply put, Pioli I was mint chocolate chip, not vanilla.

So, based on everything I've discussed in this article, here's the way forward for our beloved 49ers if they truly want to emulate NE's recent success:

Solidify the QB position once and for all. If that means hiring a shaman to do a good-luck "diamond in the rough" dance, so be it.
Use free agency a little more, and perhaps replace Director of Pro Personnel Tom Gamble, who only seems to target false hopes (e.g., WR Bryant Johnson, OL Jonas Jennings, DL Justin Smith, and LB Tully Banta-Cain) and old fogies (e.g., FB Moran Norris, WR Isaac Bruce, OL Larry Allen, OL Marvel Smith, LB Takeo Spikes, DB Walt Harris, and DB Michael Lewis).
Quit being vanilla! Honestly reassess your failures, and correct them immediately. Don't think a 5-win team only requires a little nibbling around the edges. Hell, as it relates to this year, the same goes for a 7-win team.

That's it for the series. I hope you enjoyed it, and thanks for sticking around through 7 (lengthy) parts. At the very least, I hope you learned a couple of things about 49er draft history that can be assimilated into your opinions of current and past 49er regimes.

¹ DVOA statistics used to produce this article were obtained from Football Outsiders.

Poll

Assuming that McCloughan, Singletary, and Jed will be around for the foreseeable future, which of the following adjustments is MOST necessary for the Niners to become a consistent winner ala the Patriots?

Solidifying the QB position

97 votes

Getting better at free agency

44 votes

Stop living in la-la-land thinking this mediocre team is actually good

73 votes

214 votes | Poll has closed

The Art (and Science) of Drafting: III. 49ers vs. Patriots in the Salary Cap Era

Florida Danny — Mon, 20 Apr 2009 18:08:11 -0000

Way back in Part 1 of this series - which seems like forever ago - I broke down the 49ers' draft picks from 1994-2008 by round, position, and conference. Just to refresh everyone's memory, here was my description of the Niners' overall draft strategy:

When we're good, let's trade away picks. When we're bad, let's acquire more picks.
Trade up into the 1^st round using 2^nd-round picks. Trade down into the 6^th and 7^th rounds using 5^th-round picks.
Take players from BCS conferences.
Dominate the neighborhood (i.e., take Pac-10 players).
Wait until Day 2 to draft QBs, RBs, and WRs.
Grab TEs, LBs, and DBs in the late rounds for depth and special teams.
Don't pull an Al Davis (i.e., Ks and Ps are not draft-worthy).

Today, I'll conduct a point-by-point comparison of this strategy to that of the New England Patriots. Why the Patriots, you ask? Well, the goal here was to find a team that, unlike the 49ers, has actually won a lot during the past 15 years, has done so consistently throughout those 15 years, and has done so in similar contextual circumstances. The specific criteria for identifying a team with which to compare were (a) having a top-3 winning percentage in the Salary Cap Era, (b) having won multiple championships in the Salary Cap Era, (c) having had 5 or more years between Super Bowl appearances in the Salary Cap Era, and (d) having regime contexts that were similar to the 49ers.

Given these criteria, NE is the only team that qualifies. Regarding criterion (a), the top 3 teams in winning percentage since 1994 are NE (65.0%), PIT (62.9%), and GB (62.1%). Regarding criterion (b), the only teams that have won multiple Super Bowls since 1994 are NE (3), PIT (2), and DEN (2). That narrows it down to NE and PIT. Regarding criterion (c), the only teams to have played in multiple Super Bowls separated by 5 years or more are NE (1996 and 2001), PIT (1995 and 2005), and NYG (2000 and 2007). So it comes down to criterion (d), which favors NE. That's because, whereas PIT has had a pretty stable organizational hierarchy (2 draft regimes, 2 coaches, 1 owner in 15 years), NE has had management turmoil (3 draft regimes, 3 coaches, and 2 owners in 15 years) similar to the 49ers. Not to mention that NE also had a Donahue/Erickson-esque "Does this team have a clue?" period from 1997-1999 when Bobby Grier was fully in charge of the draft (i.e., finally free of Bill Parcells' watchful eye, ala Donahue free from Walsh) and Pete Carroll was ineptly manning the sidelines (ala Erickson). So, taken together, NE is the best comparison because they've had a similar team context during the Salary Cap Era, yet actually have consistently won despite it, wholly unlike the 49ers.

Alright, so without further ado, let's start the draft strategy comparison...

After the jump, I'll compare each aspect of the 49ers' draft strategy in the Salary Cap Era to that of the Patriots...

DRAFT RESULTS

Before I really get into the meat of the matter, I think it might be useful to preface this article by discussing how NE has compared to SF in terms of their starter and Pro Bowl rates.

Starter Rates			Pro Bowl Rates
Category	SF	NE	Category	SF	NE
Overall	42.4%	35.1%	Overall	9.3%	10.5%
Round 1	72.2%	100.0%	Round 1	16.7%	41.2%
Round 2	84.6%	53.3%	Round 2	7.7%	13.3%
Round 3	44.4%	33.3%	Round 3	11.1%	5.6%
Round 4	31.3%	45.5%	Round 4	6.3%	9.1%
Round 5	7.1%	13.3%	Round 5	0.0%	6.7%
Round 6	38.9%	8.7%	Round 6	16.7%	4.3%
Round 7	23.8%	8.3%	Round 7	4.8%	0.0%
Day 1	65.3%	62.0%	Day 1	12.2%	20.0%
Day 2	26.1%	19.0%	Day 2	7.8%	4.8%
QB	28.6%	28.6%	QB	0.0%	14.3%
RB	33.3%	25.0%	RB	16.7%	6.3%
FB	100.0%	--	FB	33.3%	--
WR	33.3%	25.0%	WR	5.6%	8.3%
TE	44.4%	25.0%	TE	11.1%	0.0%
OL	64.3%	41.7%	OL	7.1%	16.7%
DL	31.8%	33.3%	DL	4.5%	11.1%
LB	40.0%	29.4%	LB	20.0%	5.9%
DB	38.1%	48.0%	DB	4.8%	12.0%
OFF	45.6%	31.0%	OFF	8.8%	9.9%
DEF	36.2%	38.3%	DEF	8.6%	10.0%
BCS	43.6%	38.6%	BCS	8.5%	10.9%
Non-BCS	37.5%	24.2%	Non-BCS	12.5%	9.1%

First, some housecleaning... Remember back to my Walsh Era (Part 2b) discussion of starter and Pro Bowl rates for a moment. You'll recall that I attempted to adjust these rates for context because draft picks on better teams are more likely to make the Pro Bowl, but less likely to crack the starting lineup. The crude method I used for this "likelihood adjustment" was the difference in winning percentage. Applying that method here, I noted earlier that NE has had a 65.0% winning percentage in the Salary Cap Era. The 49ers' winning percentage during that span was 52.1%. Therefore, if we apply my crude method of adjusting for context, a NE draft pick has been 12.9% more likely to make the Pro Bowl, but 12.9% less likely to start.

However, there's one more adjustment we need to make. Namely, it's also 7 times more likely in general for a player to be an OFF or DEF starter than it is for that player to be an OFF or DEF Pro Bowler. Trust me on this one, I did the math. So, we actually have to divide that crude 12.9% Pro Bowl advantage for NE by 7 to eliminate a draft pick having dumb luck, and thereby get the true advantage due to that draft pick playing for NE (i.e., the better team). If your eyes just glazed over, just know this for the purposes of our discussion: NE draft picks have had a 1.8% Pro Bowl advantage over SF draft picks since 1994.

OK. So as it relates to reading the table, I've displayed in bold the better percentage for each category after adjusting it in the manner I just described and adding in a little margin for error. Specifically, if a starter rate is in bold, it means it's the better rate based on a SF advantage of +10% to +12%. If a Pro Bowl rate is in bold, it means it's the better rate based on a NE advantage of +1% to +3%. Again, I realize this is a crude method, but it's what I'm going to go with for the sake of simplicity. OK, back to the show...

First question: "Has NE been better overall at drafting than SF?" The answer is "probably not." We get this answer from the differences in their overall starter and Pro Bowl rates, which are within their respective margins of error.

Maybe, then, NE has been better at drafting in specific categories rather than overall. Indeed, looking at specific categories starts to paint the picture of how NE has been better than SF in the draft. What stands out to me is that NE has outshined SF in the 1^st round, at QB, and at DB. How I interpret this is that the Patriots have been light years better than SF in the most important round of the draft and at picking the most important position skill positions on the field. Indeed, NE has been a perfect 17 for 17 at finding starters with their 1^st round picks, with 7 of these 17 picks making the Pro Bowl as a Patriot. In contrast, the SF has been 13 for 18 at finding starters in the 1^st round, with only 3 of their 18 1^st-round picks making the Pro Bowl as a Niner. At QB, the Patriots obviously hit with Tom Brady (and Matt Cassell for that matter), whereas all the 49ers have to show for their QB picks is Alex Smith and Tim Rattay. Finally, at DB, 12 of the 25 NE picks have become Patriot starters; 3 of whom have made the Pro Bowl as a Patriot. This compares to 11 of 23 SF picks having become 49er starters; only 1 of whom (Lance Schulters) made the Pro Bowl as a Niner. So the moral of the story is that, if you want to draft like a consistent NFL winner, be great in the 1^st round, and hit home runs with your QB and DB picks.

Now for the point-by-point strategy comparison...

STRATEGY #1

When we're good, let's trade away picks.

When we're bad let's acquire more picks.

As I described in Part 1, the 49ers have had 118 draft picks since 1994. The Patriots, in contrast, have had 134. That's a difference of 1 pick per season, which, although not statistically significant, is nevertheless practically important. I mean, that's still 1 more pick per season. It's also practically important because the average round for each team's picks was 4.13 for SF and 4.25 for NE, which means that we're talking about 16 more 4^th round picks, on average, for NE than SF.

Just intuitively, if the Patriots have had the highest winning percentage since 1994, and they've had 16 more picks than the 2-wins-less-per-year Niners, then it stands to reason that the Patriots haven't been trading away picks when they're good. However, let's look a little deeper, and a picture says a thousand words. Below is a graph showing the number of 49er wins each season along with the number of picks they had in the draft immediately following that season:

To fully appreciate this, just pay attention to the two trendlines. As you can see, there's a pretty strong relationship between 49er wins and the number of picks they had in the subsequent draft. As the number of wins increases, the number of picks decreases, and vice versa. Now, here's the same graph for NE:

Again, the trendlines tell the story. For the Patriots, their wins and picks ebbed and flowed together until around 1999, at which point their number of picks stayed relatively the same as the number of wins increased. One more thing you'll notice when you compare the two graphs is that, whereas the Patriots' number of picks have generally ranged from 7 to 10 each season, the 49ers' number of picks has bounced around all over the place. Therefore, based on these two differences, we'd have to conclude that the Patriots have not subscribed to the first aspect of Niner draft strategy.

STRATEGY #2

Trade up into the 1^st round using 2^nd-round picks.

Trade down into the 6^th and 7^th rounds using 5^th-round picks.

Suffice it to say that going back and looking at every draft pick trade since 1994 is a bit of a tall order seeing as how (a) the internet didn't hit the mainstream until 1999, (b) I'm not going to sort through microfiche slides at my local library, and (c) I'm not going to call the 49ers' and Patriots' front office and wait weeks, months, or years for them to get back to me with the info. No big deal, though. I needed to rephrase this strategy as "The 2^nd and 5^th rounds aren't that important to us," anyway. After all, the reason for coming up with it was because the Niners had peculiarly low pick totals in those two rounds. So, the best way to see if the Patriots are any different is to compare the round-by-round totals:

A few things are worthy of noting in this graph. First, given that there have been 15 drafts since 1994, NE has averaged no less than 1 pick per round per draft. In other words, they've had a pick total in every round that's greater than or equal to the 15 that have been allotted to them by the NFL (before Spygate). This is true of the 49ers for every round except - you guessed it - the 2^nd and 5^th. Second, if we were to rank the 14 total rounds (7 for SF and 7 for NE) by number of picks since 1994, the lowest 4 would be SF's 2^nd round, SF's 5^th round, NE's 2^nd round, and NE's 5^th round. Third, for both teams, look at how the 2^nd round compares to the 1^st and 3^rd, and how the 5^th round compares to the 4^th and 6^th. You'll notice that both teams have a dip in the 2^nd and 5^th rounds as compared to the rounds that are on either side.

Fourth, check out that 4^th round total for NE. They've had nearly as many picks in the 4^th round as they have in the compensatory-pick-heavy 6^th and 7^th rounds. Now compare that with the 49ers' middling 4^th round pick total of 16, and you begin to think to yourself, "NE must really like something about that 4^th round." Well, the possibilities are endless, but it might be useful to see what positions NE is taking in the 4^th round.

Here's a table breaking down SF's and NE's 4^th-round picks by team unit:

Unit	SF	NE	Total
OFF	4	12	16
DEF	11	9	20
ST	1	1	2
Total	16	22	38

Whereas the 49ers have taken far more DEF players in the 4^th round, NE has taken more OFF players. This, by the way, is a statistically significant finding: whether or not a 4^th-round pick plays OFF or DEF is dependent on whether SF or NE selected him.

And what kind of success have the two teams had with their 4^th-round picks? All we have to do here is refer back to the starter and Pro Bowl rate table I displayed earlier. As you can see, NE has been far better than SF at finding starters in Round 4. So perhaps one reason the Patriots stockpile 4^th-round picks is because they're really good at finding starters in that round. I know you're wondering, "Are they conscious of this success?" Well, they've acquired at least 1 extra 4^th-round pick in 5 of the last 8 drafts after having found starters with 4 of the 9 4^th-round picks they selected in their first 7 drafts of the Salary Cap Era. Remember, we're talking about the 4^th round here, so 44.4% is pretty damn good. Or maybe it's just some numerology thing involving 4s. You be the judge.

STRATEGY #3

Take players from BCS conferences.

This one's pretty easy to figure out. Here's a table showing the number of SF and NE picks that have come from BCS and non-BCS conference:

BCS?	SF	NE	Total
Yes	94	101	195
No	24	33	57
Total	118	134	252

As this table shows, the totals are pretty similar (as are the corresponding percentages). Furthermore, this conclusion stays the same even if we break the 49ers' and Patriots' picks down by draft day. Specifically, SF's BCS rate on Day 1 has been 89.8% since 1994, whereas NE's has been 86.0%. Similarly, the Niners' BCS rate on Day 2 has been 72.5%, whereas the Pats' has been 69.0%.

The only discernible difference in BCS rates between the two teams has had to do with the team unit on which the draft picks play(ed). Here's the relevant table:

BCS Rate
Unit	SF	NE
OFF	78.9%	67.6%
DEF	79.3%	87.6%
ST	100.0%	33.3%

It seems that NE has had more of a BCS bias with DEF picks than with OFF picks. In contrast, SF has had more of a BCS bias than NE with respect to OFF draft picks. In fact, the latter difference between the two teams is statistically significant: SF has been more likely than NE to take BCS players at OFF positions. Taking everything in the section together, we can conclude that, except for the SF's larger BCS bias on OFF, the 49ers and Patriots have been pretty similar when it comes to favoring players from BCS conferences.

STRATEGY #5

Wait until Day 2 to draft QBs, RBs, and WRs.

Here are two tables that show the 49ers' and Patriots' QB, RB, and WR picks by draft day:

SF
Pos	Day 1	Day 2	Total
QB	3	4	7
RB	2	4	6
WR	7	11	18
Total	12	19	31

NE
Pos	Day 1	Day 2	Total
QB	1	6	7
RB	7	9	16
WR	6	6	12
Total	14	21	35

By far, the difference between the two teams has been at QB. When compared to the 49ers, the Patriots have had a considerable Day 2 bias at this position. With respect to RB, SF and NE have been pretty similar between draft days. Finally, at WR, the Patriots haven't seemed to have had a Niner-esque Day 2 bias.

However, where there does seem to be a huge difference is in the total number of picks at RB and WR. Specifically, as the table shows, NE has taken a staggering 16 RBs since 1994, whereas the 49ers have only taken 6. In contrast, SF has taken 18 WRs, whereas the Patriots have only taken 12.

To appreciate just how aberrant these totals are, all we need is a simple knowledge of probability. Between the two teams, 8.7% of their 252 picks since 1994 have been at RB. This means we would expect SF to have used 8.7% of their 118 picks, or about 10, at this position. Using this same calculation for NE, we would expect the Patriots to have used about 12 of their picks at RB. Essentially, the Niners have taken 4 fewer RBs than expected, whereas the Patriots have taken 6 more RBs than expected. For WR, the expected totals are 14 for SF and 16 for NE, which means that SF has taken 4 more - and NE 4 less - than expected. It turns out that the differences between the two teams' expected and actual pick counts at RB and WR are statistically significant.

STRATEGY #6

Grab TEs, LBs, and DBs in the late rounds for depth and special teams.

Similar tables can be used to determine any differences between the two teams with respect to Day 2 picks at TE, LB, and DB. Here they are:

SF
Pos	Day 1	Day 2	Total
TE	2	7	9
LB	7	8	15
DB	7	14	21
Total	16	29	45

NE
Pos	Day 1	Day 2	Total
TE	4	8	12
LB	6	11	17
DB	13	12	25
Total	23	31	54

From these tables, we can conclude that (a) both teams have had a Day 2 bias at TE, (b) NE has had more of a Day 2 bias at LB, and (c) SF has had more of a Day 2 bias at DB. In other words, the teams have a 1-1-1 record against each other with respect to their propensity for drafting Day 2 TEs, LBs, and DBs. I don't know about you, but that seems pretty even to me.

STRATEGY #7

Don't pull an Al Davis (i.e., Ks and Ps are not draft-worthy).

Here, it's even Stevens. NE has drafted 2 Ks and 1 P since 1994. SF has done the same.

ADDITIONAL SF-NE COMPARISONS

Before I wrap this up, here are some additional draft-related differences between SF and NE that I think are interesting, but that didn't really fit into the point-by-point strategy breakdown. I figured that, because I did the stats work, I might as well share the full fruit of my labor with you. I'd be more than happy to discuss these further in the comments section if you wish. First, the strongest (i.e., statistically significant) differences:

SF has picked more 3^rd-round WRs than NE (5-0). SF has picked more 2^nd-round Big East players than NE (4-0).
SF has picked more Pac-10 OLs (6-1) and OFF players (13-5) than NE.
SF has picked more 2^nd-round Big East players than NE (4-0).

Next, the stronger (i.e., almost statistically significant) differences:

NE has picked more 2^nd-round WRs (5-0), 3^rd-round DBs (6-1), and 6^th-round OLs (3-0) than SF. SF has picked more 6^th-round WRs (5-2) than NE.
NE has picked more Day 1 RBs (7-2) and Day 2 OLs (18-6) than SF. SF has picked more Day 2 WRs (11-6) than NE.
NE has picked more Big 12 (25-13), MAC (5-1), and MEAC players (3-0) than SF.
SF has picked more Big 12 players in the 4^th round (4-1) and SEC players in the 5^th round (4-0) than NE. NE has picked more Conference USA players in the 4^th round (4-1) and Big 12 players in the 5^th round (5-1) than SF.
SF has picked more Day 1 Pac-10 players than NE (10-4). NE has picked more Day 2 MAC players than SF (4-0).
NE has picked more Conference USA OFF players (4-0), Big 12 DLs (6-2), and Big 12 DEF players (15-8) than SF. SF has picked more ACC OFF players than NE (6-2).

Finally, the strong (i.e., more trend-like than statistically significant) differences:

NE has picked more OLs than SF (24-14).
NE has picked more 4^th-round OFF players (12-4) and Day 1 DBs (13-7) than SF.
NE has picked more MWC players than SF (6-2).
NE has picked more 4^th-round Big East players (4-1), Day 1 Big 12 players (10-5), and Day 2 MWC players (5-1) than SF. SF has picked more Day 2 SEC players than NE (12-8).
NE has picked more ACC DEF players than SF (7-3).

SUMMARY AND CONCLUSION

To wrap things up, below is a table summarizing ways in which NE's overall draft strategy since 1994 has been similar to and different from the SF draft strategy I detailed in Part 1 of this series:

STRATEGY	NE SIMILAR?
Trade away picks when good. Stockpile picks when bad.	X
The 2^nd and 5^th rounds aren't that important to us.
Take players from BCS conferences.
Dominate the neighborhood.	X
Wait until Day 2 to draft QBs, RBs, and WRs.
Grab TEs, LBs, and DBs in the late rounds for depth and ST.
Don't pull an Al Davis.

So, NE has differed entirely on 2 of the 7 SF draft strategies. First, the Patriots have valued the draft more consistently than the Niners because (a) NE has had an average of 1 more pick per draft than SF, and (b) the relationship between NE's previous year record and current draft number of picks is much weaker than SF's. Second, the Patriots haven't had a localized regional bias towards their draft picks because (a) the Big East only ranks 4^th in number of NE picks since 1994, and (b) NE's Big East rate is much lower than SF's Pac-10 rate.

NE has differed, in part, on 2 other SF draft strategies. First, although NE has exhibited an overall and draft day BCS rate that's similar to SF, their BCS bias decreases considerably for OFF picks; to the point that it differs significantly from that of the 49ers. Second, although NE is similar to SF in terms of their QB, RB, and WR rate on Day 2, the Patriots differ from SF to a statistically significant extent with respect to the total number of RBs and WRs they've taken since 1994. Essentially, it's true that NE generally prefers selecting players from these three positions on Day 2; but it's also true that NE generally prefers selecting RBs regardless of the draft day. In other words, they take a lot more bites at the RB apple than SF.

Considering the similarities and differences in draft strategies, and putting them into the context of actual draft results (i.e., starter and Pro Bowl rates), it's certainly arguable whether NE has drafted better than SF over the past 15 years. And it's even shakier to conclude that the reason for NE being much better than SF since 1994 is an overall superiority in the draft.

Nevertheless, there are three specific areas that I think have contributed to the Salary Cap Era disparity in wins and championships between the 49ers and Patriots. First, the Patriots seem to have been more committed to the draft as a source of talent replenishment than the Niners: They've had more picks, they've had more round-to-round pick total consistency, and their pick totals haven't fluctuated with regular season wins and losses. It's true that NE hasn't necessarily been better than SF in terms of starter and Pro Bowl rates, but they have been better in terms of starter and Pro Bowl totals, and that's because they've had more bites at the apple (aka picks). In other words, everyone knows that the draft is a crap shoot. The question is, "How many rolls of the dice do you want to have?"

The second area in which I think that the Patriots' drafts have contributed to their success has to do with the Walshian quality of thinking outside the box. Specifically, the Patriots haven't seemed to go into the draft with prejudices and preconceptions about where NFL-caliber talent can be found. Rather, they seem to think that OFF players can be found outside the BCS, they seem to think talent in general can be found in unheralded conferences (e.g., the MAC, MEAC, SIAC, and SWAC), they seem to think that the first round of Day 2 (i.e., the 4^th round) is just as important - if not more important - than the first round of Day 1, and they seem to draft OLs when everyone else is taking special teamers.

The final contribution that NE's drafts have made toward their overall success has to do with what I'll call, "the double whammy." The double whammy has everything to do with getting value out of your picks. To understand it, all we have to do is use fantasy football draft success (or failure) as a proxy. The two make-or-break points in a fantasy football draft (Whammy 1 and Whammy 2) are the first round and the middle rounds. When you miss spectacularly on a 1^st-round pick in fantasy football, you're basically in catch-up mode for the rest of the season. While other teams' 1^st-round picks are off having their predictably awesome seasons, yours has taken a one-way trip to Steven Jackson-land circa 2008. And what's worse is that you're never going to get equal trade value once your 1^st-rounder has been declared a bust for the season.

In the middle rounds, it's just as important to hit on players; for two reasons. First, if you do happen to take a bust with your 1^st-round pick, you're replacing that bust with a middle-round talent rather than a late-round talent. Second, and more importantly, the players available in the middle rounds can be classified as two types: (a) early-round talents at important positions whose stocks have fallen for one reason or another, and (b) the best players available at relatively unimportant positions (e.g., TE, K, DEF). The successful fantasy teams are those that get value here. While everyone else is taking their starters at unimportant positions, it's a perfect time to take the Jamal Andersons of the NFL circa 1998, i.e., those players who aren't total sleepers (e.g., Matt Cassel), yet nevertheless have been overlooked by everyone else. It's all about value. If you're able to find a player in the 6^th round who blows up, you've now just put your team light years ahead of where it would be if you had instead taken the top-rated TE. The point here is that, contrary to popular opinion, it's the middle rounds - not the late rounds - in which you get the most bang for your "sleeper" buck.

Bringing this back to the SF-NE comparison, the Patriots have been perfect in the 1^st round. As I said earlier, they've taken 17 starters and 7 Pro Bowlers with their 17 1^st-round picks since 1994. That's getting unbelievable value in the 1^st round. To boot, by not missing in the 1^st round, or screwing around with the 2-or-3-years-away Kentwan Balmer types, the Patriots haven't had to play catch up with the talent on their roster. In the middle rounds, NE has taken a lot of players, with considerable success, in the 4^th round. This means that the sleepers they hit on are 4^th-round (or higher) talents rather than 6^th- or 7^th-round talents, and that they're not replacing injured starters with 6^th- and 7^th-rounders - Matt Cassel notwithstanding. Contrast this with the 49ers, who have swung and missed in the 1^st round several times, and have been better at finding starters in the 6^th round than they have in the 4^th. What this means is that the Niners have been populating their starting lineup - whether by choice or by injury - with inferior college talent. They haven't been getting optimal value at the two most important junctures in the draft. In other words, they haven't succeeded in the double whammy.

That's it for now. On Wednesday, I'll compare specific SF and NE regimes, focusing primarily on McNolan. In other words, to which regimes are they most similar and different? TO BE CONTINUED...

Poll

How much of the credit for New England's overall success since 1994 would you assign to their use of the draft?

All of the credit

23 votes

Most of the credit

182 votes

Some of the credit

166 votes

None of the credit

9 votes

380 votes | Poll has closed

The Art (and Science) of Drafting: IId. The McNolan Era (2005-2008)

Florida Danny — Wed, 15 Apr 2009 22:33:08 -0000

AUTHOR'S NOTE: Sorry about the delay in getting this one posted. My computer crashed yesterday because a genius (aka me) let the battery run out, so I had to wait until it fully recharged (aka today) to continue writing up the piece.

Yesterday, in Part 2c of my review of 49er draft history, I broke down the Donahue regime's picks from 2002-2004 by round, position, and conference. Just to refresh everyone's memory, here was my description of Donahue's draft strategy:

Stockpile picks, especially when your team sucks.
Draft for need on Day 1, especially when your team sucks.
I <3 the Pac-10!
BCS or die!

In comparison with the Walsh II and Policy regimes, Donahue didn't make the Policy regime mistake of trading away picks when the team was good (2002), he used Walsh II's strategy of drafting for need on Day 1 when rebuilding, he brought back the Policy regime's affinity for Pac-10 players, and he surpassed both Walsh II and the Policy regime in percentage of picks that came from BCS conferences

In this article, I'll be ending the regime breakdowns with a look at the McNolan era (2005-2008).

After the jump, I'll put the McNolan drafts in context; break the results down by position, draft day, and conference; sum up the regimes; and provide the idiot's guide to McNolan draft strategy...

Here again is the link to my Excel spreadsheet of 49er draft picks in the Salary Cap Era. I realized that some of you probably don't have Excel 2007, so I converted it to be compatible with older versions (and fixed the link in previous posts as well):

49er Draft Picks 1994-2008

IN CONTEXT

In Part 2c, I told you how Donahue didn't hit the home run he needed to with his rebuilding project in the 2004 draft. One additional piece of evidence for this is the fact that, by the very next draft, Donahue's replacements had to rebuild the roster once more through the draft. Therefore, Donahue's 2004 fire sale and draft has to rank among the shortest rebuilding phases of a franchise in the history of sports (and not in a good way). Now, I'm sure you don't need some long-winded description of what the team has done record-wise since 2004, so I'll just get on with the specific needs in each of McNolan's drafts:

2005 Draft - Key Losses: WR Cedrick Wilson (free agency), OL Kyle Kosier (free agency), OL Jeremy Newberry (career-threatening injury), OL Scott Gragg (free agency), DL John Engelberger (trade), DB Ronnie Heard (free agency), K Todd Peterson (free agency). Weak 2004 Stats: 30^th in OFF DVOA; 31^st in DEF DVOA; 19^th in ST DVOA.¹ Lingering Issues: Still hadn't found worthy replacements for QB Jeff Garcia, RB Garrison Hearst, WR Terrell Owens, and DB Zack Bronson; major injuries to QB Tim Rattay, LB Andre Carter, DB Ahmed Plummer. Needs: QB, RB, WR, OL, DL, DB, K, LB depth.

2006 Draft - Key Losses: QB Tim Rattay (trade), RB Kevan Barlow (trade), FB Fred Beasley (free agency), WR Brandon Lloyd (trade), WR Johnnie Morton (released), LB Julian Peterson (free agency), LB Andre Carter (free agency), and DB Ahmed Plummer (career-threatening injury). Weak 2005 Stats: 32^nd in OFF DVOA; 31^st in DEF DVOA. Lingering Issues: Still hadn't found worthy replacements for Garcia, Hearst, Owens, and Bronson; major injury to TE Eric Johnson. Needs: FB, WR, LB, DB, QB depth, RB depth, TE depth.

2007 Draft - Key Losses: WR Antonio Bryant (released), TE Eric Johnson (free agency), OL Jeremy Newberry (free agency), and DB Tony Parrish (released). Weak 2006 Stats: 22^nd in OFF DVOA; 28^th in DEF DVOA; 18^th in ST DVOA. Lingering Issues: Still hadn't found worthy replacements for Owens and Bronson; OL Larry Allen and DL Bryant Young now in their mid-30s. Needs: WR, DB, ST, TE depth, OL depth, DL depth.

2008 Draft - Key Losses: WR Darrell Jackson (free agency), OL Larry Allen (retirement), OL Justin Smiley (free agency), DL Bryant Young (retirement), DL Marques Douglas (free agency), and LB Derek Smith (free agency). Weak 2007 Stats: 32^nd in OFF DVOA; 28^th in DEF DVOA. Lingering Issues: Still hadn't found worthy replacements for Owens and Bryant; major injuries to QB Alex Smith, OL Jonas Jennings, and LB Manny Lawson. Needs: WR, OL, DL, LB, QB depth.

In preparation for its drafts, McNolan has had to deal with about 7 key losses per offseason, which is fewer than Walsh II (8), but more than Donahue (4) and the Policy regime (4). Interestingly, the course of McNolan's tenure has seen a mixture of his 3 predecessors' draft contexts. The 2005 and 2006 offseasons were Walshian, with about 8 key self-induced, salary-cap-caused losses per season that generally involved kicking dead weight to the curb or trading it for draft picks. The 2007 offseason was Donahue-esque, with only 4 key losses and a team feeling good about itself after a promising season. Finally, 2008 was Policy-like in that there were two retirements to deal with. Overall, I'd say the most similar previous context to McNolan's was the Walsh II era.

I know this is totally off-topic, but, Cedrick Wilson may have made the greatest career decision in the history of the NFL by signing with PIT after the 2004 season. If he stays in SF, he's stuck on a 6-win team for the foreseeable future (at least until he roughed up his ex-girlfriend) catching passes from Rattay, Smith, Ken Dorsey, Cody Pickett, Trent Dilfer, Shaun Hill, and Chris Weinke. Instead, he goes to PIT and wins a ring his first season with the team catching passes from Ben Roethlisberger. Seriously, whatever it was that made him leave the Niners, he needs to bottle and sell it. Can you guys think of any other player in recent years who's lucked out like that?

Anyway, back to the regularly scheduled program. For the sake of comparison, here's how McNolan rebuilt the starting lineup in 2005 and 2006:

Morton (2005 McNolan free agent signing) for Wilson
Snyder (2005 McNolan draft pick) for Kosier
OL Eric Heitmann (2002 Donahue draft pick) for Newberry
OL Kwame Harris (2003 Donahue draft pick) for Gragg
Douglas (2005 McNolan free agent signing) for Engelberger (after moving Young to 3-4 DE)
DB Mike Adams (2003 Donahue undrafted free agent signing) for Heard
K Joe Nedney (2005 McNolan free agent signing) for Todd Peterson
Smith (2005 McNolan draft pick) for Rattay
RB Frank Gore (2005 McNolan draft pick) for Barlow
FB Moran Norris (2006 McNolan free agent signing) for Beasley
Bryant (2006 McNolan free agent signing) for Lloyd
WR Arnaz Battle (2003 Donahue draft pick) for Morton
Lawson (2006 McNolan draft pick) for Julian Peterson
LB Brandon Moore (2002 Donahue undrafted free agent signing) for Carter (after moving him to 4-3 ROLB)
DB Shawntae Spencer (2004 Donahue draft pick) for Plummer (after moving him to LCB)

Adding it all up, McNolan rebuilt the starting lineup with 5 traditional free agents, 4 draft picks, and 6 players originally acquired by Donahue. Comparing this to the previous regimes, McNolan used free agency more than Walsh II and Donahue, but less than Policy; and they had a "holdover rate" of 33.3%, which was smack dab in between Walsh II (16.7%) and Donahue (50.0%).

So there's the context. Hopefully you sense a pattern emerging here: the contexts of McNolan drafts have been a cocktail of issues and solutions having the ingredients of each predecessor; 2 ½ parts Walsh II, 1 part Donahue, and ½ part Policy.

PICKS, PICKS, AND MORE PICKS

From 2005-2008, the 49ers made 35 selections in the NFL draft, which nearly equalled the Walsh II and Donahue rate of 9 picks per season. In 2005 and 2006, the main rebuilding years, McNolan stockpiled picks via all those trades I mentioned earlier; just like Walsh II and Donahue. The past 2 drafts, however, McNolan has tended to trade away picks (you're welcome, NE) rather than stockpile them; just like the Policy regime. Once again, the trend here is for McNolan to be a mixture of all 3 predecessor's strengths and weaknesses.

In terms of McNolan's overall Pro Bowl rate, a Donahue-esque and Walshian 5.7% of its picks have become 49er Pro Bowlers (Gore and LB Patrick Willis). However, its starter rate of 34.3% - which includes OL Chilo Rachal's midseason and DB Dashon Goldson's offseason ascensions - is the worst of the 4 regimes (Policy regime = 41.4%; Walsh II = 48.1%; Donahue = 48.1%). When you remember the differences in context between regimes, you realize that McNolan has been way worse at picking starters than the context-similar Walsh II and Donahue regimes, and even worse than the context-different Policy regime, whose superior teams made it harder for draft picks to crack the starting lineup (i.e., the Policy regime's starter rate is artificially low due to context).

Bottom line - Here's what I think are the main things to take away from this section:

McNolan has continued the Walsh II and Donahue standard of 9 draft picks per season.
McNolan has stockpiled picks when the team sucked, and traded away picks when (they thought) the team was good.
McNolan has been as bad at picking Pro Bowlers as the "unlikely to have Pro Bowlers because the team sucked" Walsh II and Donahue regimes, and even worse at picking starters than the "unlikely to have draft selections start because the team was good" Policy regime.
Overall, McNolan seems to be channeling the good and bad of its 3 predecessors.

DAY ‘N' NIGHT

McNolan has taken exactly 40% of its picks on Day 1 of the draft, and 60% on Day 2. That's almost identical to Walsh II's 41/59 split, which, as I said in Part 2c, was smack dab between the Policy (52/48) and Donahue (33/67) regimes.

Looking at the relative success of its picks by draft day, we find that 64.3% of McNolan's Day 1 picks have become 49er starters, which is, again, right in the middle of the 4 regimes; not as good as Walsh II's (81.8%, which I incorrectly reported as 64.3% in Part 2b), but basically on par with the Policy regime's (60.0%) and Donahue's (55.6%). Remember, though that Walsh II and Donahue are the better comparisons in terms of team context. On Day 2, an abysmal 14.3% of McNolan picks have become 49er starters, easily the worst of the 4 regimes (again, even worse than the starter-unlikely Policy regime). So while McNolan has had pretty standard success on Day 1 in terms of starters, it's been horrible at picking Day 2 starters.

The best example of this dichotomy between Day 1 and Day 2 results is McNolan's initial rebuilding draft in 2005. Here it is:

Rd	Player	Pos	Starter?	Pro Bowl?
1	Alex Smith	QB	Yes	No
2	David Baas	OL	Yes	No
3	Frank Gore	RB	Yes	Yes
3	Adam Snyder	OL	Yes	No
5	Ronald Fields	DL	No	No
5	Rasheed Marshall	WR	No	No
6	Derrick Johnson	DB	No	No
7	Daven Holly	DB	No	No
7	Marcus Maxwell	WR	No	No
7	Patrick Estes	TE	No	No
7	Billy Bajema	TE	Yes	No

Day 1? 4 picks, 4 starters, 1 Pro Bowler. Day 2? 7 picks, 1 starter, 0 Pro Bowlers. Enough said.

With regard to Pro Bowlers, however, the preferred drink is once again "Regime Cocktail." On Day 1, 14.3% of McNolan's picks have become 49er Pro Bowlers, which is slightly worse than the Pro-Bowl-likely Policy regime (20.0%), considerably better than Donahue (0.0%), and slightly better than Walsh II (9.1%). OK, correction: McNolan drinks Regime Cocktail only on Day 1. That's because, it has yet to pick a Day 2 Pro Bowler. Of course, it's not like the other two Pro-Bowl-unlikely regimes (Walsh II = 6.3%; Donahue = 5.3%) were much better at it.

Bottom line - Here's what I think are the main things to take away from this section:

McNolan has had a Walshian 40/60 split of picks between Day 1 and Day 2.
McNolan has been worse than Walsh II, but better than Donahue, at picking Day 1 starters.
McNolan has been far less successful than the other 3 regimes at picking Day 2 starters.
McNolan has been more successful than Walsh II and Donahue at picking Day 1 Pro Bowlers, but slightly worse than these two regimes at picking Day 2 Pro Bowlers.

ASSUME THE POSITION

Here's how McNolan has used its 35 picks by position, unit, and draft day:

Pos	Day 1	Day 2	Total
QB	1	0	1
RB	1	1	2
FB	0	0	0
WR	2	5	7
TE	1	2	3
OL	4	1	5
DL	2	5	7
LB	2	1	3
DB	1	6	7
K	0	0	0
P	0	0	0
OFF Total	9	9	18
DEF Total	5	12	17
ST Total	0	0	0
Total	14	21	35

Overall, McNolan has basically split their picks up evenly between units, which is more similar to what the Policy regime and Donahue did than Walsh II's need-induced DEF preference. Also, like Walsh, McNolan hasn't taken any ST players in the draft. Of course, that may have something to do with having Nedney and ROBO-PUNTER on the roster. What's most interesting here, however, is that, although OFF has received equal attention on Days 1 and 2, 70.6% of McNolan's DEF picks have come on Day 2.

The even split of OFF picks between draft days is a little misleading because McNolan was actually Walshian in 2005 and 2006 when it had to replace most of the OFF. Going back to that 2005 draft that I showed earlier, you'll notice that all 4 Day 1 picks were on OFF. This continued in 2006 when they selected TE Vernon Davis in the 1^st round after Johnson suffered a major injury the previous season. Essentially, McNolan followed the Walsh II example of spending Day 1 of rebuilding drafts addressing the most glaring positional needs. This is really eye-opening when you consider that Scot McCloughan claims to be a devout believer in "best player available." Is he full of sh*t or is the focus on need in Day 1 of the 2005 and 2006 drafts just a coincidence? I'll leave you to decide.

DEF, however, is a different animal altogether. From the table, it's clear that the draft day disparity on DEF has been due to an inordinate amount of Day 2 picks being spent on DLs and DBs. The 5 Day 2 DLs were, in order of selection, Ronald Fields, Parys Haralson, Melvin Oliver, Jay Moore, and Joe Cohen; the 6 Day 2 DBs were Derrick Johnson, Daven Holly, Marcus Hudson, Vickiel Vaughn, Goldson, and Tarell Brown. Looking at these lists, you could say that McNolan has gotten worse over time picking Day 2 DLs, whereas it's gotten better over time picking Day 2 DBs. Either way, aside from Haralson and Goldson, that's a pretty mediocre-at-best bunch. So, if you want to know why the DEF has sucked for so long (before Singletary pulled their heads out of their asses), look no further than McNolan mostly waiting until Day 2 to pick DEF players. Couple this with McNolan's aforementioned horrible starter rate on Day 2, and you get one wet, stinking heap of also-ran-ity.

With 71.4% of picks taken on Day 2, WR is another position at which McNolan has focused little Day 1 attention. The two Day 1 WRs were Brandon Williams and Jason Hill, both 3^rd-rounders.

Hopefully, you're starting to see a trend here. McNolan gets credit for using Walsh II's "pick for need on Day 1" and "bites at the apple" strategies. However, it's made the mistake of waiting until Day 2 to take most of their bites at DL, DB, and WR. Perhaps, then, one reason why the Niners haven't been able to get over the mediocre 7-9 hump since 2005 is because their WR, DL, and DB corps are populated with Day 2 picks.

To drive it home, here's one last point about McNolan's (lack of) success wtith DEF picks. On OFF, McNolan's starter rate (44.4%) has been similar to that of the Policy regime (43.8%), Walsh II (50.0%), and Donahue (46.2%). However, the DEF starter rates for the previous 3 regimes were 47.1% for Walsh II, 41.7% for Donahue, and 33.3% for the Policy regime. McNolan's? 23.5%!

All is not lost, however. Unlike Donahue, McNolan seems to have actually learned something about positional picks as its tenure has progressed. Namely, 2 of their 3 picks in 2008 were DEF players. Whether or not DL Kentwan Balmer and DB Reggie Smith pan out in the long run, at least McNolan seems to have noticed that their previous Day 1 picks on DEF have become integral parts of the starting unit. Don't believe me? Here they are: Lawson, Willis, and DL Ray McDonald. So, their results - 2 starters, 1 Pro Bowler, and an emerging pass-rush specialist - were not shabby at all. This suggests that (a) when McNolan commits Day 1 resources to DEF, it's been pretty successful; and (b) the regime seems to have figured this out. So, at the very least, kudos to them for that.

Bottom line - Here's what I think are the main things to take away from this section:

Like Walsh II, McNolan has focused its Day 1 picks on the most glaring positional needs (when it had to).
McNolan has employed Walsh II's "bites at the apple" strategy at WR, DL, and DB. However, they've primarily waited until Day 2 to take their bites.
Given their propensity for waiting until Day 2 for DEF picks, it's not surprising that McNolan has been horrible at drafting DEF starters.
Based on 2008, McNolan seems to have learned from their "wait until Day 2 for DEF" mistake.

CONFERENCE ROOM

So far, I've basically told a story about how the McNolan regime's drafts have been a little bit Walsh II, a little bit Donahue, and a little bit Policy; how they've used the good strategies as well as the bad. Basically, not too hot, not too cold; I'll call it "Goldilocks" if you will. The conference breakdown is where that tale ends. There's one regime's lead in particular that McNolan has been happy to follow, and I'm sure you'll easily be able to figure it out. Here's how McNolan's 35 picks shake out by conference: 7 ACC picks, 7 SEC picks, 6 Big 12 picks, 6 Pac-10 picks, 4 Big 10 picks, 1 Big East pick, 1 Conference USA pick, 1 MAC pick, 1 MAII pick, and 1 MWC pick.

BCS anyone? McNolan has taken a Donahue-esque 88.6% of its picks from BCS conferences. I already detailed in Part 2c what I think of this strategy, so you know by now that I don't like it. Now, Scot McCloughan has argued that he takes BCS players because they're (a) less affected by big crowds, (b) more prepared for the NFL, and (c) better team leaders. While all of these may be true in a general sense, the draft isn't about generalities. It's about identifying and acquiring specific talented players. I'm sorry, but when you limit your available talent pool to 6 or 7 college football conferences, you're going to miss the "diamonds in the rough" that teams need to sustain success. Also, McCloughan has made it seem like picking players from BCS conferences is only a Day 1 strategy. The stats say the opposite though: A higher percentage of McNolan's Day 2 picks have come from BCS conferences (90.5%) than have its Day 2 picks (88.6%). Oh, and did I mention that McNolan's first pick ever was a QB from a non-BCS conference (Smith), and that they traded up in 2007 to take an OL from a non-BCS conference (Joe Staley)? That's 2 of McNolan's 4 non-BCS picks being taken with 1^st-round picks. Again, is he BSing us here or is this some kind of smokescreen? It's pretty interesting how a general BCS guy like McCloughan can fall in love so much with specific non-BCS players.

And now, I present a second way in which McNolan's conference preferences are similar to Donahue's: They suck just as bad at picking BCS players despite being so BCS-reliant. Wait, let me correct that. They suck even worse. Here's the evidence:

McNolan
Starter?	BCS	Non-BCS	Total
Yes	10	2	12
No	21	2	23
Total	31	4	35

Donahue
Starter?	BCS	Non-BCS	Total
Yes	12	1	13
No	13	1	14
Total	25	2	27

You'll recall from Part 2c that Donahue's BCS starter rate (48.0%) was worse than Walsh II's (62.5%) even though Donahue was much more BCS-heavy with his picks. Well, even Donahue's sorry BCS starter rate makes him look like a draft maven when compared to McNolan's. That's because, as the table shows, only 32.3% of McNolan's BCS picks have become regular 49er starters! As I said in Part 2c, if you're going to be limiting your talent pool so drastically, you better be good at evaluating the talent you don't ignore. Clearly, McNolan has not been good in this regard, and has been even worse than the guy for which I invented the argument.

Now, I argued in Part 2c that Donahue's preference for BCS players and lack of success in actually picking good ones was borne out of his dispositional laziness and inattention to draft detail. With respect to McNolan, however, I'm sure this hasn't been the case. To its credit, the McNolan regime has been almost pathologically engaged in draft preparation. After all, they coached the Senior Bowl 3 straight seasons, Scot McLoughan was Director of College Scouting for the Seahawks, and Scot's brother, David, now serves in the same capacity for the 49ers. So I'm kind of perplexed as to why, given their affinity for evaluating college talent, McNolan has shied away from non-BCS players. One might argue that it's because of the Alex Smith experience. However, as has been noted, McLoughan remains a big Smith supporter. So what gives here? Please help.

Bottom line - Here's what I think are the main things to take away from this section:

Like Donahue, almost every player McNolan has drafted has been from a BCS conference.
Despite its BCS-lust, the McNolan regime has had an even worse BCS starter rate than Donahue.

GENERAL ANALYSIS AND CONCLUSION

As I've alluded to throughout this article, the McNolan drafts can be described by one word: Goldilocks. In other words, their porridge hasn't been too hot and it hasn't been too cold; just lukewarm. Here's the evidence:

McNolan's draft contexts have been a mixture of its 3 predecessors' contexts.
McNolan's been right in the middle in terms of a 49er regime's "draft vs. free agency" bias.
McNolan's stockpiled picks in rebuilding drafts ala Walsh II, but also traded away picks in "we feel good about ourselves right now" drafts ala the Policy regime.
McNolan's been right in the middle in terms of a 49er regime's Day1/Day 2 split.
McNolan's been right in the middle in terms of Day 1 starter and Pro Bowl rate.
McNolan has evenly split their picks between OFF and DEF overall.
McNolan's used Walsh's "bites at the apple" strategy, but only on Day 2.
McNolan's been on par with previous regimes in terms of picking OFF starters.
Though it's made Donahue-esque draft strategy errors, McNolan's either learned from those mistakes or has a more valid reason for why they've made them.

All in all, it sure seems like the McNolan regime has been a mixture of the good and bad of the 3 previous 49er regimes. They rebuilt like Walsh II in 2005 and 2006, but they sat on their laurels like Donahue and the Policy regime in 2007. They draft for need on Day 1 like Walsh II, but, like Donahue, only when they have to. They take a lot of bites at the apple like Walsh, but they wait until Day 2 to do it. You get the picture.

To me, understanding the Goldilocks nature of McNolan drafts helps explain why the 49ers have been stuck in "no better than 7-9" mode for the past 3 seasons. If you draft a propensity of mediocre players, or if you have a bunch of mediocre drafts, you're going to find yourself having a mediocre team. Here, another word comes to mind that has been used - with great comic effect - to describe McNolan: vanilla.

In evaluating the 4 Niner regimes since 1994, something has become very clear to me; something that I didn't fully realize until I got to McNolan. Bill Walsh was the only 49er GM in the past 15 years who went out and tried to be great in the draft. He's the only one who tried to bat 1.000 every time he stepped to the plate. The Policy regime was content to sustain success through free agency, Donahue was content to ride the coattails of Walsh II's drafts, and McNolan has been content to only be great on Day 1. Is it any wonder then that the Walsh II regime was the only one that had an upward win trajectory? Is it a coincidence that Walsh was the only GM to leave the 49ers better off than when he arrived? I mean, the 49ers regressed under the Policy regime and Donahue, and they've stagnated in 7-9 territory under McNolan. After reading my regime breakdowns, do you have any doubt that one important reason for this is because Walsh II's draft strategy differed so drastically from that of McNolan, Donahue, and the Policy regime? In re McNolan, do you think that the "hybrid" defense, not giving Hill the QB job outright, and not firing Nolan after 2007 can be described by any other word except "vanilla?"

Whether it's McSingle nee McNolan or someone else, until the person/people in charge of running the 49ers' drafts stop placing artificial constraints on themselves, and start trying to be great with every pick- by whatever means necessary - I fear the Niners are going to wallow in mediocrity.

Here's a perfect example of what I mean. The Niners need a pass rusher right now more than anything. Go out and draft the best damned pass rusher. If the consensus best pass rusher isn't the one you've identified as the best, then trade down for value and get the guy you want. If the pass rusher you think is going to be great played at (site decorum) Directional University in the Calfornia Penal League, take him anyway. Just don't sit back and wait to draft some OK pass rusher and hope he turns out to be good with a little coaching! Address the (site decorum) need already! In other words, try to be great!

Another example. Right now, you have an arguable draft bust, a good but ragged-armed journeyman, and a player KC didn't even want as your QBs. Go out and draft the best damned QB. Do whatever it takes. If the consensus best QB isn't the one you've identified as the best, then trade down for value and get the guy you want. If the QB you think is going to be great played at (site decorum) Directional University in the California Penal League, take him anyway. Just don't sit back and wait to draft some OK QB and hope he turns out to be good with a little coaching! Address the (site decorum) need already! In other words, try to be great!

You know what team drafts like this? The New England Patriots. But that's Part 3, so you'll have to wait until tomorrow to hear the details. For now, here's the idiot's guide to the McNolan regime's draft strategy:

Stockpile picks when your team sucks, but trade away picks when (you think) your team is good.
Draft for need on Day 1, especially when your team sucks.
Use Day 2 to take a lot of bites at the apple.
BCS or die!
We're Goldilocks (aka Vanilla McNolan).

So, based on this strategy, as well as some specific draft history trends that you can find in my Excel spreadsheet, here's what (and what not) to expect from the 49ers in the 2009 draft:

Don't expect them to take more than 1 player from a non-BCS conference.
Don't expect them to take a pass rusher, QB, or WR at #10.
Don't expect them to take a RB on Day 1.
Expect them to take a WR or two on Day 2.
Expect them to take an OL on Day 1.
Expect more DEF picks (esp. DLs and DBs) on Day 2 than Day 1.
Don't expect them to take a FB, K, or P.

Remember, this is what I expect them to do or not do based on their history. If they end up going against tendency, it's not because the tendencies were wrongly identified. It's because they did something that was unexpected given their history. And there's no harm in that. It's actually worked once or twice (See Willis, Patrick).

That's it for now. Tomorrow, I'll compare the overall 49er draft strategy since 1994, which I identified in Part 1, to that of the Patriots. TO BE CONTINUED...

¹ DVOA statistics used to produce this article were obtained from Football Outsiders.

Poll

How much blame would you assign McNolan's drafts for the 49ers continuously being unable to take that next step in recent seasons?

All of the blame

37 votes

Most of the blame

83 votes

Some of the blame

103 votes

None of the blame

9 votes

232 votes | Poll has closed

The Art (and Science) of Drafting: IIc. The Donahue Era (2002-2004)

Florida Danny — Tue, 14 Apr 2009 06:35:39 -0000

Yesterday, in Part 2b of my review of 49er draft history, I broke down the Walsh II regime's picks from 1999-2001 by round, position, and conference. Just to refresh everyone's memory, here was my description of Walsh II's draft strategy:

Stockpile picks when your team sucks.
Draft for need on Day 1, especially when your team sucks.
This is the pros, not college. There's no need to dominate your neighborhood.
Focus Day 2 on non-BCS players because even one diamond in the rough more than makes up for several lumps of coal.

And, in comparison, here was the Policy regime's draft strategy that I detailed in Part 2a:

Draft picks are commodities best used for moving up.
Once you've addressed positional needs in free agency, there's no need to focus on them in the draft.
Dominate the neighborhood (i.e., take Pac-10 players and those near Youngstown, OH).
There's plenty of talent outside the BCS conferences, but wait until Day 2 to acquire it.

So basically, the main differences between the two regimes were (a) Walsh II used the draft to rebuild the roster, whereas the Policy regime used the draft to tinker around the edges of the roster; (b) Walsh II drafted for need (esp. on Day 1), whereas the Policy regime didn't; and (c) Walsh II didn't have a geographical bias, whereas the Policy regime did. As I said in Part 2b, difference (a) is no doubt due, in some part, to differences in context. Nevertheless, it's still a difference worth noting, and it's overshadowed by differences (b) and (c), which have nothing to do with context.

I bring up both of these draft strategies because they can serve as opposite ends of a draft strategy continuum going forward. In other words, because they're such polar opposites, we can use them as draft strategy templates with which to describe the Donahue and McNolan strategies. Speaking of Donahue, his regime's drafts are the topic of this article.

After the jump, I'll put the Donahue drafts in context; break the results down by position, draft day, and conference; and provide the idiot's guide to Donahue draft strategy...

Here again is the link to my Excel spreadsheet of 49er draft picks in the Salary Cap Era:

49er Draft Picks 1994-2008

IN CONTEXT

Fans of a certain age might remember Dion's song, "The Wanderer." Well, from 2002-2004, Donahue apparently remade the song in his image and called it, "The Squanderer." Probably the easiest way to describe this is in a graph:

As shown, team wins peaked in the last year of Walsh II and began its decline in the first year of Donahue. That's a pretty stark display of how Donahue squandered Walsh II's success, and is much more vivid than anything I could say in writing.

The obvious questions here are, "Why the complete turnaround?" and, "What did Donahue do to precipitate it?" Well, knowing the specific contexts of each Donahue is probably a good place to start:

2002 Draft - Key Losses: OL Ray Brown (free agency) and DB Lance Schulters (free agency). Weak 2001 Stats: 20^th in ST DVOA.¹ Lingering Issues: None. Needs: OL, DB, ST.

2003 Draft - Key Losses: DL Dana Stubblefield (free agency) and DL Chike Okeafor (free agency). Weak 2002 Stats: Dropped to 18^th in DEF DVOA; 26^th in ST DVOA. Lingering Issues: Major injuries to OL Dave Fiore and DB Zack Bronson; half of the OFF at least 30 years old. Needs: DL, ST, OFF depth, DEF depth.

2004 Draft - Key Losses: QB Jeff Garcia (free agency), RB Garrison Hearst (free agency), WR Terrell Owens (trade), WR Tai Streets (free agency), OL Derrick Deese (free agency), OL Ron Stone (free agency), DL Travis Kirschke (free agency), and DB Zack Bronson (free agency). Weak 2003 Stats: 19^th in DEF DVOA; 28^th in ST DVOA. Lingering Issues: Still hadn't replaced Stubblefield; Starting OLs missed 18 total games with minor injuries. Needs: QB, RB, WR, OL, DL, DB, ST.

Looking only at 2002 and 2003, it sure seems like Walsh II set Donahue up pretty well for the future, seeing as how Donahue had to deal with 6 fewer key losses per offseason in preparation for his first two drafts. Basically, Donahue could do whatever his heart desired with his 2002 and 2003 draft picks. This certainly wasn't the case, however, with respect to the 2004 draft, after Donahue blew up the roster in a Walsh-esque salary cap purge.

So with respect to evaluating Donahue's drafts we have an interesting context to consider. He had even fewer needs in the first two drafts than the Policy regime did back in the halcyon days of 49er teams past. However, his final draft was Walsh-like in that he was tasked with a self-induced roster overhaul. Therefore, for the first two drafts, we need to see whether Donahue was sitting back on his laurels in a Policy-channeling-Nero sort of way; whereas for the 2004 draft, we need to see whether he (successfully) attacked the draft in a Walshian sense.

Before breaking things down, it's useful to detail how Donahue replaced all of those starters in 2004 so we can get a handle on his "draft vs. free agency" slant. The comparison with Walsh II is very straightforward here because both were handcuffed by the salary cap. In other words, they both had the same constraints when considering whether or not to address certain needs in free agency. So without further ado, here are Donahue's 2004 replacements along with the method via which they were acquired:

QB Tim Rattay (2000 Walsh II draft pick) for Garcia
RB Kevan Barlow (2001 Walsh II draft pick) for Hearst
WR Brandon Lloyd (2003 Donahue draft pick) for Owens
WR Cedrick Wilson (2001 Walsh II draft pick) for Streets
OL Kyle Kosier (2002 Donahue draft pick) and OL Kwame Harris (2003 Donahue draft pick) for Deese
OL Justin Smiley (2004 Donahue draft pick) and Kosier for Stone
DL Anthony Adams (2003 Donahue draft pick) for Kirschke
DB Ronnie Heard (2000 Walsh undrafted free agent signing) for Bronson

As you can see, none of the replacements were acquired via traditional free agency; very Walshian indeed. However, before we start applauding Donahue as a draft maven, notice that there was an even split between how many replacements were Walsh II leftovers and how many were of Donahue's making. Namely, Donahue's 50.0% "holdover help rate" is considerably higher than Walsh II's 16.7%. In other words, Donahue had a lot more help with replacing starters in 2004 than Walsh II did while in a similar predicament from 1999-2001. Of course, this says nothing about how Donahue handled the 4 starters he had to replace in 2002 and 2003 or about how many of his draft picks ever became 49er starters.

So there's the context. For his first two drafts, Donahue was in an even better position (thanks to Walsh II) than the 12-wins-per-season Policy regime from a need perspective. Essentially, he had none. In contrast, Donahue's 2004 draft found the team in a predicament similar to what Walsh II dealt with from 1999-2001. He had to turn over the roster through the draft due, in part, to the fact that he was constrained in free agency by salary cap woes. But here's a thought: Perhaps Donahue didn't "find" himself in the 2004 predicament. Maybe 2004 was the result of his own draft failures in 2002 and 2003. Just maybe, rather than being an innocent bystander vis-à-vis circumstances that lead to the 2004 fire sale, what if he was instead a catalyst of his own misfortune? In other words, what if these two different contexts (need-light heaven and need-heavy hell) are not mutually exclusive? For the answers to these questions, we need to break down Donahue's drafts.

PICKS, PICKS, AND MORE PICKS

From 2002-2004, the 49ers made 27 selections in the NFL draft, which works out to 9 picks per season. Once again, that was very Walshian of Donahue. Given how Walshian this amount of picks was, you might think that Donahue followed the "trade away picks when good, and acquire picks when bad" strategy of his two predecessors, such that he had more picks in the 2004 rebuilding draft than he did in the "happy days are here again" 2002 and 2003 drafts. Interestingly, this wasn't the case: Donahue had the same number of picks in 2002 (10) as he did in 2004, and he had the full repertoire of 7 picks in 2003. So, all in all, Donahue seemed to learn the lesson of Policy regime failure: Don't ignore the draft just because you're winning.

With respect to Donahue's starter and Pro Bowl rates, he was, again, very Walshian. He had the same exact starter rate (48.1%) and a similarly poor Pro Bowl rate (3.7% vs. 7.4%). Based on these stats, and given the similarity in winning percentages between the two regimes, we can come to the same conclusion about Donahue's general draft success that we did regarding Walsh II. Namely, he wasn't necessarily any better or worse than his predecessor(s).

Bottom line - Here's what I think are the main things to take away from this section:

Donahue was eerily Walshian with respect to the number of draft picks he had and how successful those picks were overall.
Donahue stockpiled picks; even when the 49ers were good.
After you account for differences in context, the Donahue, Walsh II, and Policy regimes were pretty much equals when it came to drafting 49er starters and Pro Bowler.

DAY ‘N' NIGHT

The first difference we see between Donahue and his predecessors is related to the percentage of draft picks they had on each draft day. Specifically, whereas Policy had a 52/48 split and Walsh had a 41/59 split between draft days (which I incorrectly reported as more like 50/50 yesterday), Donahue made only 33.3% of his selections on Day 1. Part of this had to do with draft day trades in 2002 and 2004, but it was also due to compensatory selections that were awarded by the NFL (which can't be traded), and having to forfeit a Day 1 pick in 2002 thanks to Policy Era salary cap shenanigans. Either way, Donahue's draft day disparity was nowhere more evident than in the 2002 draft, when only 2 of his 10 selections came on Day 1. Perhaps if he had a few more Day 1 picks that year, maybe he wouldn't have found himself in the sorry situation that was the 2004 offseason? Hmmm.

Looking at the relative success of their Day 1 and Day 2 picks, we find that Donahue was slightly worse than Walsh II at picking starters on Day 1 (55.6% vs. 64.3%), and slightly better than Walsh II at picking starters on Day 2 (44.4% vs. 30.8%). With respect to Pro Bowlers, Donahue failed to pick one on Day 1 (Walsh II = 9.1%) and was as (un)successful as Walsh II at picking them on Day 2 (5.3% vs. 6.3%). Given the similarities in team context between the Donahue and Walsh II eras, we can conclude that they weren't all that different in draft day success except for, perhaps, Donahue's Pro Bowl swing-and-miss on Day 1.

Bottom line - Here's what I think are the main things to take away from this section:

Donahue had far more picks on Day 2 than Day 1; unlike the Policy regime and Walsh II.
Donahue was similar to Walsh II (and the Policy regime) with respect to drafting 49er starters and Pro Bowlers on a given draft day.

ASSUME THE POSITION

I established earlier that Donahue didn't really have any glaring positional needs in 2002 and 2003, and that he had a veritable sh*t-ton of them in 2004. But for the sake of comparison, here's how Donahue's Day 1 and Day 2 picks break down by team unit:

Unit	Day 1	Day 2	Total
OFF	4	9	13
DEF	5	7	12
ST	0	2	2
Total	9	18	27

As the table shows, Donahue was more Policy-esque (i.e., evenly split) than Walshian in terms of the positional breakdown by draft day. Of course, this is to be expected given that his first two drafts were ones in which he didn't have to address any glaring positional needs. He basically had "draft freedom," and he divided that freedom equally between OFF and DEF.

Donahue's 2004 draft was a different story altogether, with replacing over half of their starting OFF becoming a distinct Niner priority. As I said earlier, Donahue luckily had 10 picks in the 2004 draft. So how did he use them? Well, let's just say he must not have read the team needs like you and I just did. Specifically, Donahue drafted only 4 OFF players in 2004: 2 WRs, 1 OL, and 1 QB. I'll give him credit for taking 3 of these 4 OFF players on Day 1, which I argued in Part 2b is a good idea. However, I do have one gigantic bone to pick in re addressing needs during the 2004 draft.

Donahue waited until the 6^th round to take his lone QB draft pick despite Garcia's release having left the team with two 7^th-round QBs and a 5th-round QB on their roster. Keep in mind, here, that 2004 is now considered one of the best QB draft classes of the past 30 years (thanks to the Big 3), and that Donahue traded down twice in the 1^st round. I can only assume that Donahue tried (and failed) to trade up for Ben Roethlisberger. But once he realized he wasn't getting Roethlisberger at Pick 16 (or earlier), how does he not take a QB before the 7^th round? I want to be clear here. This is not about results: Donahue was right to trade down after the top 3 QBs were off the board, and odds are that if he had taken a QB before the 7^th round, that QB wouldn't have panned out. It's about effort. Given how Walshian other aspects of Donahue's draft strategy were, I sure wish he would have employed Walsh II's "bites at the apple" strategy with respect to QBs.

One more thing I'll say about (lack of) QB bites at the apple is that Donahue took 3 QBs from 2002-2004: 1 (Brandon Doman) was selected in the 5^th round, and 2 (Ken Dorsey and Cody Pickett) were selected in the 7^th round. This is certainly not a serious attempt to solidify the QB position long-term. Perhaps if he would have been more serious in 2002 and 2003 about it, he wouldn't have found himself in the 2004 predicament.

Here are two final observations I'll make about Donahue's position picks. First, he took as many Ps in 2004 as he did QBs or OLs. Granted, he drafted ROBO-PUNTER (aka Andy Lee), but what is he doing drafting a P when he has half of a starting lineup to replace? Also, when you have Hearst playing on a bionic ankle at RB in 2002 and 2003, how do you take 0 RBs in 3 drafts? Sure, Barlow showed potential; but not even 1 RB in 3 drafts? Not even for depth? Pure madness. And I haven't even mentioned the fact that he actually took a K in the 4^th round of the 2002 draft.

Now, in terms of draft success, here's how Donahue compared to Walsh II at the positions I identified earlier as needs over the course of their respective tenures:

Donahue
Starter	QB	RB	WR	OL	DL	DB	Total
Y	0	0	2	4	2	3	11
N	3	0	2	0	3	2	10
Total	3	0	4	4	5	5	21

Walsh II
Starter?	RB	OL	DB	Total
Yes	3	3	2	8
No	3	2	4	9
Total	6	5	6	17

Once again, Donahue can be described as Walshian here, and he was actually super-Walshian with his OL picks.

So far I've seemed to suggest that Donahue had more in common with Walsh than Policy when it came to drafting. If that's really the case, we still haven't figured out why the team regressed so precipitously during Donahue's tenure, and why it seems he left the franchise in such a bad situation after 2004. Well, the first clue has entirely to do with the 2004 draft. Here it is:

Rd	Player	Pos	Starter?	Pro Bowl?
1	Rashaun Woods	WR	No	No
2	Justin Smiley	OL	Yes	No
2	Shawntae Spencer	DB	Yes	No
3	Derrick Hamilton	WR	No	No
4	Isaac Sopoaga	DL	Yes	No
4	Richard Seigler	LB	No	No
6	Andy Lee	P	Yes	Yes
6	Keith Lewis	DB	Yes	No
7	Cody Pickett	QB	No	No
7	Christian Ferrara	DL	No	No

You might say, "Well, he got 4 starters and his lone Pro Bowler out of that draft. That's not so bad!" Remember the context though. He had just lost 8 starters in a fire sale, 4 of which had been Pro Bowlers during their 49er careers. This was no time for an OK result. Donahue needed a draft like Walsh's 2000 effort: one that totally revitalizes the team, and sets it up for a division championship in the near future. Unfortunately for Niner fans, Donahue's 2004 draft was just OK; not an A+, not an F, just a C.1^st-round pick? A bust. 3^rd-round pick? A bust. 4^th-round picks? A miss and a DE/DT tweener who spent 4 seasons trying to find a position. 6^th-round? These guys have contributed a lot, except only on ST. 7^th-round? Nevermind. So basically, although he had the chance to pull a Walsh II with 10 picks, Donahue instead only drafted 2 players (both 2^nd-rounders) that could be considered to have had a perennial positive impact on the OFF or DEF starting lineups. A Walsh-circa-2000 this draft was not. It's no wonder then that SF has yet to recover from the 2003 fire sale.

Bottom line - Here's what I think are the main things to take away from this section:

Like Walsh II, Donahue focused his Day 1 picks on the most glaring positional needs (when he had to).
Donahue seemed to sit on his laurels when the team had few needs. He didn't draft any RBs despite having a bionic starter, he didn't draft a QB before the 5^th round despite having a geriatric starter and 7^th-round backups, and he actually chose a K in the 4^th round.
Donahue didn't use Walsh II's "bites at the apple" strategy.
Donahue had a just-OK draft in 2004 when he needed to have a great one.

CONFERENCE ROOM

By far, the biggest difference between Donahue and his predecessors was with respect to the conferences from which he drafted his players. Before elaborating further, here's how Donahue's 27 picks shake out by conference:6 Pac-10 picks, 6 Big East picks, 5 SEC picks, 4 Big 10 picks, 2 Big 12 picks, 1 ACC pick, 1 Division 1 Independent pick, 1 MWC pick, and 1 WAC pick.

The difference I'm talking about was not related to "dominating your neighborhood." On this count, Donahue revived the Policy regime's preference for Pac-10 players. Between his ties to UCLA, his permanent SoCal home, and his (infrequent) trips to Santa Clara, Donahue had to have watched a lot of Pac-10 football.

Rather, the difference I speak of is Donahue's near-perfect BCS rate. Specifically, 92.6% of his draft picks came from BCS conferences. This rate was well above the Policy regime's (79.7%), and dwarfed Walsh II's (59.3%). Now, I'm going to speculate here about what Donahue's clear preference for BCS players might mean, so feel free to provide opposing viewpoints in the comments section.

To me, Donahue's BCS rate was just another indicator of his lackadaisical work ethic, something that 49er fans and media have accused him of since 2002. Exhibit A: His BCS rate. It sure doesn't take much work to pick players from the conferences with mega-million-dollar broadcast contracts, i.e., the conferences whose games can be found from August to January simply by turning on a TV set. Exhibit B: The 2 non-BCS conferences represented in Donahue picks, the MWC and WAC. First, MWC and WAC teams are in close proximity to California and - I don't live out there so correct me if I'm wrong on this - they have games that are broadcast locally in parts of California. Second, if you look at Jeff Sagarin's conference ratings from 2002-2004 (here, here, and here), you find that the MWC was the top-rated non-BCS conference all 3 years, and that the WAC was the 4^th-, 4^th-, and 3^rd-highest rated non-BCS conference in 2002, 2003, and 2004, respectively. In other words, he wasn't exactly digging into the diamond mines with the 2 non-BCS picks he did make. Exhibit C: His preference for working from home in LA. This one's been well-documented, so there's no need for me to elaborate further.

Finally, Exhibit D: He wasn't any good at it. Here are two tables comparing Donahue's success on BCS picks with Walsh II's:

Donahue
Starter?	BCS	Non-BCS	Total
Yes	12	1	13
No	13	1	14
Total	25	2	27

Walsh II
Starter?	BCS	Non-BCS	Total
Yes	10	3	13
No	6	8	14
Total	16	11	27

Wait a second. Donahue, the BCS-phile, was actually worse than Walsh II, the non-BCS-phile, at picking starters from BCS conferences? Apparently so: Walsh II had a 62.5% starter rate for BCS picks, whereas Donahue's was only 48.0%. Look, if you're going to be lazy, and if you're going to limit your talent pool so drastically as to have a draft strategy that says, "BCS or die," you better be darn good at it. "Darn good at it" Donahue was not. Despite it being just as easy for a Donahue pick to start as it was for a Walsh II pick to start (given their similar 3-year winning percentages), Donahue was as good as the flip of a coin at picking BCS starters.

This begs the question, "Why did he continue to ignore non-BCS players even though he kept missing with his BCS picks?" Surely, if he noticed the trend, and wasn't lazy, he might have shifted strategies midstream. He didn't, and I think it's because he was, in fact, uninterested in those mundane college scouting details that Walsh was so fond of. In other words, as much as Donahue tried to be a Walsh draft clone, he didn't possess the underlying quality that made Walsh successful: attention to detail.

So there's my take on the second piece of the "What did Donahue have to do with the 49ers' steep decline?" puzzle. Answer: laziness. Although his drafts were Walshian in several respects, Donahue simply didn't work hard enough to find those "diamonds in the rough" that are vital for sustaining NFL success. What's worse, he failed to adjust after this lack of success became painfully evident. In the words of Einstein, "The definition of insanity is doing the same thing over and over again and expecting different results." As I said, that's just my educated guess based on some statistical (and journalistic) evidence. Take it for what it is, and feel free to offer an opposing viewpoint in the comments section.

Bottom line - Here's what I think are the main things to take away from this section:

Like the Policy regime, Donahue took a relatively large percentage of players from the Pac-10.
Almost every player Donahue drafted was from a BCS conference.
Even Donahue's rare non-BCS picks came from the best and most-watched mid-majors.
Despite Donahue's BCS-lust, he could have flipped a coin and been just as successful with his BCS picks.

GENERAL ANALYSIS AND CONCLUSION

I began this article by wondering if Donahue's 2002-2004 drafts had anything to do with the decline in 49er wins over the course of his tenure (and beyond). Initial signs pointed to "they didn't."

He stockpiled picks just like Walsh II, even though he could have just as easily repeated the Policy regime mistake of trading them away when the team is good.
Like Walsh II, he didn't over-utilize free agency to replace starters.
He had a similar overall starter and Pro Bowler rate to Walsh II's.
Like Walsh II, he focused his Day 1 picks on need positions when the team context dictated it.
He was as good as or better than Walsh II at picking starters at need positions.

Alas, however, the devil is in the details. Sometimes, even though it looks like a duck, walks like a duck, and talks like a duck, it's still not a duck. Similarly, even though Donahue was Walsh II's doppelganger in many respects, he made two fatal errors that were anything but Walshian. First, he had a C-grade 2004 draft when an A+ draft was necessary to rebuild the roster. Second, he limited his available talent pool - rather than expanding it like Walsh did - by relying so heavily on the overexposed (and easily scouted) BCS conferences.

The 49ers are still paying for it to this very day. And it's too bad given the tutelage Donahue enjoyed as Walsh's understudy. All I know is that, if he would have taken more notes in class, he migh have gotten that A+ draft in 2004. So, based on everything I've said in this article, here's the idiot's guide to the Donahue regime's draft strategy:

Stockpile picks, even when your team is good.
Draft for need on Day 1, especially when your team sucks.
Dominate the neighborhood.
BCS or die!

That's it for now. Later today, I'll break down the McNolan drafts (2005-2008), compare them to the 3 regimes I've discussed so far, and give a heads-up on the draft strategy you might see in a couple of weeks. TO BE CONTINUED...

¹ DVOA statistics used to produce this article were obtained from Football Outsiders.

Poll

How much blame would you assign Terry Donahue's draft failures for causing the 49ers' current streak of 6 straight losing seasons?

All of the blame

60 votes

Most of the blame

132 votes

Some of the blame

83 votes

None of the blame

6 votes

281 votes | Poll has closed

The Art (and Science) of Drafting: IIb. The Walsh II Era (1999-2001)

Florida Danny — Mon, 13 Apr 2009 06:19:31 -0000

AUTHOR'S NOTE: Happy belated Easter everyone. Sorry for the delay on getting this one posted, but there's just a ridiculous amount of info to sift through. Hopefully the article's worth the wait. Also, just wanted to let you know that I'll be posting my review of the Donahue Era drafts later today.

In Part 2a of my 49er draft history review, I broke down the Policy regime's picks from 1994-1998 by round, position, and conference. Just to refresh everyone's memory, here was my description of the Policy regime's draft strategy:

Draft picks are commodities best used for moving up in the draft.
Once you've addressed positional needs in free agency, there's no need to focus on them in the draft.
Dominate the neighborhood (i.e., take Pac-10 players and those near Youngstown, OH).
There's plenty of talent outside the BCS conferences, but wait until Day 2 to acquire it.

As I said at the end of Part 2a, the Policy regime's use of draft picks can be considered a form of negligence. Despite having glaring needs at RB, OL, and DB that were brought on by aging starters, retirements, and free agent defections, they chose to trade away a lot of picks and use the remaining ones on "wants" rather than "needs." To me, this was an insanity that didn't end until Bill Walsh returned to the organization in time to run the 1999 draft.

To say that draft strategy changed under Walsh would be an understatement. Today, I'll detail why.

After the jump, I'll put the Walsh drafts in context; break the results down by position, draft day, and conference; and provide the idiot's guide to Walsh draft strategy...

Here again is the link to my Excel spreadsheet of 49er draft picks in the Salary Cap Era:

49er Draft Picks 1994-2008

IN CONTEXT

On this Easter holiday, I think it's fitting to be discussing the Walsh II Era (1999-2001), which can accurately be called "The Resurrection." The dynasty died in 1998, and three days (ok, years) later it was born again thanks to the second coming of Bill Walsh: the Niners went from 4 wins in 1999 to 6 wins in 2000 to 12 wins in 2001. To fans, this seemed like a familiar tale, what with the similar progression that took place during Walsh's previous 49er messiah incarnation (2 wins in 1979; 6 wins in 1980; 13 wins in 1981). Put in that context, the 49er drafts from 1999-2001 have to be labeled a success, right? Well, stay tuned. First, here's more contextual detail:

1999 Draft - Key Losses: RB Garrison Hearst (career-threatening injury), FB Marc Edwards (trade), OL Kevin Gogan (trade), OL Kirk Scrafford (retirement), DL Roy Barker (trade), DL Chris Doleman (retirement), DB Merton Hanks (free agency). Weak 1998 Stats: Dropped from 1^st to 15^th in DEF DVOA; 26^th in ST DVOA.¹ Lingering Issues: Still hadn't found a worthy replacement for TE Brent Jones or CB Rod Woodson; major injury to DL Bryant Young; starting QB Steve Young, WR Jerry Rice, and OL Ray Brown all over 35 years old. Needs: RB, FB, TE, OL, DL, DB, ST, QB depth, WR depth, OL depth.

2000 Draft - Key Losses: Steve Young (retirement), OL Chris Dalman (retirement), DL Gabe Wilkins (released), LB Lee Woodall (free agency), DB Darnell Walker (free agency), DB Mark McMillan (free agency), DB Tim McDonald (retirement). Weak 1999Stats: 18^th in OFF DVOA; 31^st in DEF DVOA; 31^st in ST DVOA. Lingering Issues: Still hadn't found a worthy replacement for Jones, Barker, Doleman, Woodson or CB Marquez Pope; Rice and Brown were now 37 years old. Needs: TE, DL, LB, DB, ST, QB depth, WR depth, OL depth.

2001 Draft - Key Losses: RB Charlie Garner (free agency), Rice (free agency), TE Greg Clark (released), DL Anthony Pleasant (free agency), DL Brenston Buckner (free agency), DL Junior Bryant (career-threatening injury); LB Jeff Posey (free agency), LB Ken Norton (retirement), LB Winfred Tubbs (released), and K Wade Richey (free agency). Weak 2000 Stats: 28^th in DEF DVOA; 22^nd in ST DVOA. Lingering Issues: Still hadn't found a worthy replacement for Jones and Barker; Brown was now 38 years old. Needs: RB, TE, DL, LB, K, ST, WR depth, OL depth.

In terms of specific needs, DEF and ST were definitely the priority given the stats and player departures. However, taking a broader perspective on these draft needs, you notice that, as compared to the Policy Era, there were a lot more of them. Beginning with Policy's last draft and continuing through the Walsh regime, the 49ers needed players at every position. Indeed, in comparison with the 19 key offseason losses from 1994-1998 - or about 4 per offseason - there were 24 key offseason losses from 1999-2001 - or 8 per season. Essentially, Walsh was tasked with cleaning up after Policy's free agent extravaganza and draft neglect, and therefore spent most of his 2^nd stay in SF separating the wheat from the chaff on the 49ers' roster, whether through cutting players, letting them walk in free agency, "gently" forcing them into retirement, or trading them for draft picks.

Another thing you notice from 1999-2001 is the return of Walsh's signature move: unceremoniously showing old-timers the door at their first sign of performance deterioration. What Joe Montana, Ronnie Lott, and others experienced during Walsh's first go-round in SF, Steve Young and Rice experienced during his second. Remember, both Young and Rice wanted to continue playing even though everyone (except them) knew that they were a shadow of their former self (in Rice's case) or couldn't stay healthy anymore (in Young's case).

OK, so we know that Walsh II (correctly) got rid of a lot of players. But how did he acquire their replacements? Well, of the departed players I listed above (and there were a lot of them), here's how they were replaced via traditional free agency:

Garner replaced Hearst
Wilkins replaced Doleman
OL Scott Gragg replaced Dalman (after moving OL Jeremy Newberry permanently to C)
DL Dana Stubblefield replaced Buckner
LB Derek Smith replaced Norton
K Jose Cortez replaced Richey

Now, before you assume that the other 18 were all replaced with Walsh II draft picks, here are 4 other starters who were replaced during Walsh II's tenure with free agents originally signed by Policy:

OL Dave Fiore (undrafted free agent) replaced Scrafford
Posey (undrafted free agent) replaced Barker and Woodall (after being moved to LB)
DB Zack Bronson (undrafted free agent) replaced Hanks
Buckner replaced Bryant

And finally, here is 1 replacement that Walsh II made by signing an undrafted free agent:

QB Jeff Garcia replaced Young

Let's digest that for a moment. Out of 24 starters that left the team during the offseason of each draft, Walsh II only replaced 6 of them with free agents from other teams. How's that for doing a 180 from the Policy regime's ...um...policy?

So there's the context. In preparation for his drafts, Walsh II had to deal with a ton of holes on both sides of the ball - but especially on DEF and ST - not a lot of help from Policy Era free agent holdovers, and no room under the salary cap to sign his own free agents. Solution? The draft of course.

PICKS, PICKS, AND MORE PICKS

From 1999-2001, the 49ers made 27 selections in the NFL draft, which works out to 9 picks per season. In comparison to the previous regime, Walsh II had nearly as many picks in 3 drafts as the Policy regime did in 5. Given that each team is allotted 1 pick for each of the draft's 7 rounds, this difference means that Walsh II had 6 more picks than he was supposed to, whereas Policy had 6 less. Why? The aforementioned context.

With respect to Walsh II's and the Policy regime's starter and Pro Bowl rate, conclusions again require context. Specifically, Walsh's starter rate was 48.1%, whereas the Policy regime's was 41.4%. Taking into account the fact that it was just a tad bit tougher for a player to start on those stellar teams from 1994-1998 than it has been since 1999, it's more accurate of a comparison if we count only those Policy regime picks that initially became starters during the Policy regime (i.e., when the team was good). Making this adjustment, the Policy regime's starter rate drops to 31.0%. So the question becomes, "Is the difference between 48.1% (Walsh II's starter rate) and 31.0% (the Policy regime's adjusted starter rate) bigger than or smaller than the difference in how good the teams were?" In other words, "How much of that 17.1% difference in starter rates is due to Walsh being a better drafter, and how much is due simply to the fact that it was easier to start after 1998?"

Well, one unbelievable crude way to answer this is by comparing the Niners' winning percentage from 1994-1998 to that of the 5 seasons afterwards as a measure of the difference between how good the teams were during those eras. From 1994-1998, the 49ers had a 76.3% winning percentage, whereas they've had a 40.0% winning percentage since 1998. Therefore, using this crude measurement, the Niners were 36.3% better from 1994-1998. Applying this to our question (and again, this is unbelievably unsophisticated statistically speaking) it was something like 36.3% easier for a draft pick to start after 1998 than it was from 1994-1998. Based on this difference, you might conclude that the 17.1% difference in starter rates for Walsh II and the Policy regime is more than made up for by the fact that it was harder for players to start during the Policy years. Nevertheless, as I don't by any means think we're using top-notch statistical methodology here, I'd just conservatively say that Walsh II wasn't necessarily better at drafting starters even though his starter rate was higher.

The same analysis can be done with Pro Bowl rate. 20.7% of the Policy regime's picks became 49er Pro Bowlers, and that adjusts down to 10.4% when you consider only those draft picks that made the Pro Bowl from 1994-1998. Walsh's Pro Bowl rate, in contrast, was 7.4%. Given that better teams get more players chosen to the Pro Bowl, we can crudely applying the winning percentage stats above as measure of "how easy it was to make the Pro Bowl." Doing so, you might conclude that the 3.0% or so difference in Pro Bowl rate was more than made up for by the fact that it was harder to make the Pro Bowl after 1998. Again, though, I'd just conclude that Policy wasn't necessarily better at drafting Pro Bowlers even though his Pro Bowl rate was higher.

Bottom line - Here's what I think are the main things to take away from this section:

Walsh had 27 picks in 3 drafts, an amount nearly equal to what the Policy regime had in 5 drafts.
After you account for differences in context, Walsh II and the Policy regime were pretty much equals when it came to drafting 49er starters and Pro Bowlers.

DAY ‘N' NIGHT

Based on what I said in the last section, one might be tempted to guess that Policy subtracted picks by trading up in the draft, whereas Walsh II added picks by trading down. Surprisingly, though, this wasn't actually the case. Of the Policy regime's 29 picks, 51.9% of them were on Day 1, whereas a nearly identical 51.7% of Walsh II's picks were on Day 1.

Looking at the relative success of their Day 1 and Day 2 picks, we find that Walsh II (64.3%) and the Policy regime (60.0%) were pretty much equally effective at finding starters on Day 1, whereas Walsh II (30.8%) was slightly better than the Policy regime (21.4%) on Day 2. With respect to Pro Bowlers, Walsh II (Day 1 = 9.1%; Day 2 = 6.3%) was considerably worse than the Policy regime (Day 1 = 20.0%; Day 2 = 21.4%). However, once again, we run into that pesky context problem here: Walsh's acumen for finding starters on Day 2 might have been because it was easier for a Day 2 pick to crack the starting lineup after 1998, and the Policy regime's acumen for finding Pro Bowlers on both draft days might have been because it was easier for any 49er in general to make the Pro Bowl before 1999.

However, all is not lost. One way we can avoid the above-applied statistical gymnastics is by examining differences within each regime in starter and Pro Bowl draft picks. In other words, instead of comparing, for instance, Day 1 success of Walsh drafts to Day 1 success of Policy regime drafts, we can compare, for instance, Day 1 success of Walsh drafts to Day 2 success of Walsh drafts, because, if one regime was better than the other, that superiority is likely to cut across draft days regardless of the "team goodness" context in which the picks were made.

Doing this within regime comparison, the stats say, once and for all, that the two regimes didn't differ that much: the Policy regime picked 75.0% of their drafted starters on Day 1 and 25.0% on Day 2, whereas Walsh II picked 69.2% on Day 1 and 30.8% on Day 2. In terms of Pro Bowlers, the stats say the same: both regimes had a 50/50 split of Pro Bowlers between draft days. In other words, keeping the pick context constant (i.e., looking at success within regimes), both regimes picked 3 times as many starters, and an equal percentage of Pro Bowlers, on Day 1 as they did on Day 2. If either regime were objectively better at drafting, whether overall or by round, the difference would have shown up in this within-regime comparison (e.g., if Walsh had a 5-to-1 Day 1/Day 2 ratio for starters, whereas the Policy regime only had a 2-to-1 Day 1/Day 2 starter ratio, then Walsh would be an objectively better Day 1 drafter). So, any way we look at it, we can't really conclude that one draft regime was more successful than the other with respect to starters and Pro Bowlers.

Bottom line - Here's what I think are the main things to take away from this section:

Although Walsh II made a lot of trades to acquire draft picks, he still ended up with an even spread between Day 1 and Day 2.
When you control for context, Walsh II and the Policy regime drafted similar percentages of their total number of starters and Pro Bowlers on a given draft day, aka they're equals.

ASSUME THE POSITION

One conclusion we can make about differences between the two draft regimes has to do with the positions on which they focused their Day 1 and Day 2 picks. On Day 1, Walsh II used 81.8% of his picks on DEF, whereas the Policy regime used 60.0% of their picks on OFF. In raw terms, Walsh II used only 2 of his Day 1 picks on OFF. Can you name them? If you said QB Giovanni Carmazzi and RB Kevan Barlow, you win the no-prize. The point of this, though, is not to give a pop quiz or remind you of an epic draft fail like Gio. Rather, it's to show that Walsh II actually made picks according to need. The Niners' needs were predominately on DEF, and he spent practically all of Day 1 for 3 consecutive drafts picking DEF players. Contrast this with the Policy regime, who basically gave the Italian salute to their positional needs.

Yeah, about those positional needs. Here's what Walsh II did with his 27 picks from 1999-2001: he used 17 on DEF, 10 on OFF, and 0 on ST. By position, he took 6 DLs, 6 DBs, 5 LBs, 3 RBs, 2 QBs, 2 WRs, and 1 OL. Relating these numbers back to team needs, TE Eric Johnson replaced Clark, DL John Engelberger replaced Doleman, DL Andre Carter replaced Pleasant, LB Julian Peterson replaced Posey, LB Jeff Ulbrich replaced Tubbs, DB Ahmed Plummer replaced Walker, and DB Jason Webster replaced McMillan. The obvious thing to notice here is that Walsh II drafted over half of the starting defense that went 12-4 in 2001, but only 1 starter on offense. Again, context is everything here: if you need to rebuild a defense, then actually rebuild it; don't instead go taking players at "want" positions ala the Policy regime.

Speaking of success, here's how Walsh II and the Policy regime compared with respect to drafting starters at "need" positions:

Policy
Starter?	RB	OL	DB	Total
Yes	0	1	2	3
No	1	3	1	5
Total	1	4	3	8

Walsh II
Starter?	DL	LB	DB	Total
Yes	3	3	2	8
No	3	2	4	9
Total	6	5	6	17

On the surface, these tables again suggest that Walsh II (47.1%) and the Policy regime (37.5%) were equals when it came to drafting starters. However, what's important here are not the percentages, but the totals. Walsh drafted over twice as many players at need positions, and, low and behold, drafted over twice as many starters at those positions. This is another example of that "bites of the apple" argument I've made previously.

And just to drive the point home, here's a look at Day 1 of Walsh II's 2000 draft:

Rd	Player	Pos	Starter?	Pro Bowl?
1	Julian Peterson	LB	Yes	Yes
1	Ahmed Plummer	DB	Yes	No
2	John Engelberger	DL	Yes	No
2	Jason Webster	DB	Yes	No
3	Giovanni Carmazzi	QB	No	No
3	Jeff Ulbrich	LB	Yes	No

How's that for addressing positional needs on Day 1? Out of 5 DEF picks, Walsh II batted 1.000, picking 5 starters: 2 at LB, 2 at DB, and 1 at DL. That's practically half of the starting DEF on the 12-4 2001 team right there. Sure, he screwed up his lone OFF Day 1 pick, but he had Garcia entrenched as the QB starter at that time. Therefore, Gio was more of a "want" pick (i.e., for depth) in the grand scheme of things. Besides, when you hit on 5 players at need positions, who cares that you missed on 1 player at a "want" position? After all, no one (not even Walsh) is perfect.

Herein lies the main lesson when comparing Walsh II's drafts with those of the Policy regime: focus your Day 1 picks on need positions. Taking more bites at the apple softens the blow when you miss. I mean, who can look at that Day 1 in 2000 and say, "Taking Gio set the franchise back for years?" In contrast, focusing your Day 1 on players at "want" positions (ala the Policy regime) sets the franchise back doubly when you miss. Not only did you pick a bust; you failed to fill needs in the process. That's why people look back at the Policy regime's 1995 and 1997 drafts and say, "With so few picks to work with, taking JJ Stokes and Jim Druckenmiller set the franchise back for years!"

Bottom line - Here's what I think are the main things to take away from this section:

Unlike the Policy regime, Walsh II focused his Day 1 picks on the most glaring positional needs.
Walsh drafted twice as many starters, Pro Bowlers, and players overall at need positions as the Policy regime.

CONFERENCE ROOM

Another clear way that Walsh II differed from the Policy regime was related to the conferences from which they drafted their players. Here's how Walsh II's 27 picks shake out by conference: 5 SEC picks, 4 Big 10 picks, 3 Big East picks, 3 WAC picks, 2 Pac-10 picks, 2 Division I Independent picks, 1 Big 12 pick, 1 Atlantic 10 pick, 1 Big Sky pick, 1 Conference USA pick, 1 Ivy League pick, 1 OVC pick, 1 SAC pick, and 1 Southern pick.

A couple of things jump out from this list. First, unlike those of the Policy regime (and the 49ers overall), Walsh II's picks were not, shall we say, homegrown. The Pac-10 was represented by less than 10% of Walsh II's picks, as opposed to the 17% or so of Policy's (and the 49ers' overall) picks. Also, 2 of the top 3 conferences represented in Walsh II picks did not play a majority of their games a hop, skip, and a jump from DeBartolo country (i.e., Youngstown, OH). The second thing that jumps out is actually nothing, literally: the ACC is absent. That's right, Walsh II took 0 players from ACC schools.

Now, where things get really interesting is in the BCS breakdown. As compared to the 75.9% BCS rate of the Policy regime and the 79.7% BCS rate of the 49ers overall, Walsh II used only 59.3% of his picks on players from BCS conferences. In other words, he didn't shy away from taking players from smaller schools. On the contrary, he embraced it.

Walsh II's fondness for taking non-BCS players might suggest that he was particularly adept at it from a results perspective. Interestingly enough, though, the opposite was actually the case. Of his 11 non-BCS picks, only 27.3% became Niner starters, and none made the Pro Bowl. In contrast, 62.5% of his BCS picks became Niner starters, and 12.5% made the Pro Bowl. So Walsh II was actually better at picking BCS players than non-BCS players. Furthermore, when you compare his starter and Pro Bowl rates for BCS and non-BCS players, you find that Walsh II was more successful than the BCS-heavy Policy regime when it came to picking BCS players, but less successful than the non-BCS-phobic Policy regime in terms of picking non-BCS players. So it turns out that, whereas the Policy regime probably should have drafted more non-BCS players, Walsh II probably should have drafted more BCS players.

Walsh II didn't follow the 80% rule of overall 49er draft strategy when it came to positions, which isn't surprising given his overall BCS rate of 59.3%. His preference for non-BCS players was especially evident in his QB and OL picks, positions at which he didn't select a single BCS player. The lone position at which he reached an 80% BCS rate was LB. When combining positional BCS rates by team unit, we find that Walsh II preferred BCS players on DEF (64.7%) more than he did on OFF (50.0%). Of course, this is to be expected given that he took more DEF players overall.

The next question to address is, "When did Walsh II make his non-BCS picks?" Here's a table showing the number of BCS and non-BCS players that he selected on each day of the draft:

BCS?	Day 1	Day 2	Total
Yes	9	7	16
No	2	9	11
Total	11	16	27

Converting these numbers into percentages, 18.2% of Walsh II's Day 1 picks came from non-BCS conferences, whereas 56.3% of his Day 2 picks came from non-BCS conferences. In one sense, this is expected given that a tripling of the non-BCS rate from Day 1 to Day 2 is standard for the 49ers overall in the Salary Cap Era. However, what's most interesting here is that, on Day 2, Walsh actually picked more non-BCS players than BCS players.

So given that Walsh II wasn't that good at picking starters from non-BCS conferences, and that he picked an inordinate amount of non-BCS players on Day 2, you'd expect that his starter rate for Day 2 non-BCS picks was atrocious. The tables below provide the evidence:

BCS Picks
Starter?	Day 1	Day 2	Total
Yes	8	2	10
No	1	5	6
Total	9	7	16

Non-BCS Picks
Starter?	Day 1	Day 2	Total
Yes	1	2	3
No	1	7	8
Total	2	9	11

Based on the second table, we'd conclude that, yes, Walsh II was horrible at taking Day 2 non-BCS players. But, then again, he was almost as bad taking Day 2 BCS players. What's important to recognize, though, is the discrepancy between these Day 2 non-BCS results and his Day 1 BCS results. Namely, he was nearly perfect at picking players from BCS conferences on Day 1, and nearly imperfect (or whatever the polar opposite of perfect is) at picking players from non-BCS conferences on Day 2. Going back to my earlier discussion, the moral here is that we as 49er fans don't care (or even remember) that Walsh II sucked so bad on small-school fliers because (a) he waited until Day 2 to take them, and (b) he hit on so many Day 1 BCS picks.

Bottom line - Here's what I think are the main things to take away from this section:

Walsh II did not have a preference for players from conferences that were close to home.
Walsh II was less inclined than the Policy regime (and the 49ers overall) towards players from BCS conferences.
Walsh II waited until Day 2 to take non-BCS players, and he actually took more non-BCS players than BCS players on Day 2.
Despite having a higher non-BCS rate than the Policy regime, Walsh II was actually less successful with his non-BCS picks.
Despite having a lower BCS rate than the Policy regime, Walsh II was actually more successful with his BCS picks.
Walsh was highly successful on Day 1 BCS picks, and highly unsuccessful with Day 2 non-BCS picks. But, then again, no one cares or even remembers.

GENERAL ANALYSIS AND CONCLUSION

In this article, I've identified a few stark contrasts between the philosophies of Walsh II and the Policy regime as they relate to the draft. First, Walsh II was far more inclined than the Policy regime to take advantage of the draft. It's obviously up for debate whether this was borne out of free will (i.e., because he preferred the draft to free agency) or determinism (i.e., because he was handcuffed by the salary cap). Whatever the reason, it's clear that the context of the situation (i.e., needing to rebuild the roster from scratch) and Walsh's previous experience (i.e., rebuilding the roster through the draft during Walsh I) made the draft more likely than free agency as an organizational focus.

Second, Walsh II's approach to the draft was not bound by geographical proximity or bias toward "big time" college football. Perhaps this was due to previous experiences like the one that resulted in an unheralded, yet highly successful, G.O.A.T. from Mississippi Valley State University. Or perhaps Walsh's approach was based on his desire to always be one step ahead of the competition, even with respect to the mundane details of college scouting. In short, Walsh thought outside the box. When he arrived in 1979, he thought outside the box and revolutionized offensive football. In his second incarnation as 49er savior, he focused on the draft while most other GMs - including his predecessor - were smitten with the advent of free agency.

If I had to come up with one criticism of Walsh II's drafts, it's that - like me in my fantasy football drafts -he was almost too much in love with finding "diamonds in the rough." Even though it might be contrary to the previous paragraph, I think - like my affinity for young, backup RBs behind an old NFL starter (ala Chris Perry ca. 2005) - he might have gone overboard with his affinity for non-BCS players. Given his success with BCS picks, he might have had an even better draft record from 1999-2001 had he chosen a few more of them on Day 2. But, even here, the negatives of a poor success rate with Day 2 non-BCS picks are mitigated by the fact that they were taken on Day 2, because, generally speaking, fans remember the Druckenmillers and Stokes of 49er draft history (i.e., Day 1 carriages that turned into pumpkins) more than we remember the Tyrone Hopsons and Menson Holloways (i.e., Day 2 lumps of coal that never became diamonds).

The final distinction between Walsh II and the Policy regime is also related to draft days, and also plays a major role in how fans look back at their respective drafts. Namely, Walsh II used Day 1 on glaring positional needs (of which there were many), whereas the Policy regime did not. There's a longstanding debate about draft strategy that is basically framed as "need vs. best player available." Some have even entered a third dog into the fight named "best player available at a position of need." Really, though, if you're drafting for need, aren't you going to take the best player available at that position? It kind of renders your draft board useless if not. Anyway, based on the long-term draft results of Walsh II and the Policy regime, I'm thinking that the evidence supports drafting for need. I'm going to get into this more with respect to the McNolan regime vis-à-vis DL Kentwan Balmer, but suffice it to say that Walsh's success on Day 1 of his drafts provides a pretty strong argument in favor of drafting for need, especially in the context of a talent-starved roster.

So, based on everything I've said in this article, here's the idiot's guide to Walsh II regime's draft strategy:

Stockpile picks when your team sucks.
Draft for need on Day 1, especially when your team sucks.
This is the pros, not college. There's no need to dominate your neighborhood.
Focus Day 2 on non-BCS players because even one diamond in the rough more than makes up for several lumps of coal.

That's it for now. Later today, I'll break down the vomit-inducing Donahue drafts (2002-2004). TO BE CONTINUED...

¹ DVOA statistics used to produce this article were obtained from Football Outsiders.

Poll

Which of these do you consider the better strategy on Day 1 of the draft?

Drafting the best player available at a need position

146 votes

Drafting the best player available regardless of position

125 votes

271 votes | Poll has closed

The Art (and Science) of Drafting: IIa. The Policy Regime (1994-1998)

Florida Danny — Fri, 10 Apr 2009 20:48:51 -0000

AUTHOR'S NOTE: Apologize in advance, but the team and player page tags aren't working for me right now for some reason. I'll attempt to link them shortly.

On Wednesday, in Part 1 of my review of 49er draft history, I broke down their picks since 1994 by round, position, and conference. Just to refresh everyone's memory, here's the overall Niner draft strategy I came up with based on my breakdown:

When we're good, let's trade away picks. When we're bad, let's acquire more picks.
Trade up into the 1^st round using 2^nd-round picks. Trade down into the 6^th and 7^th rounds using 5^th-round picks.
Take players from BCS conferences.
Dominate the neighborhood (i.e., take Pac-10 players).
Wait until Day 2 to draft QBs, RBs, and WRs.
Grab TEs, LBs, and DBs in the late rounds for depth and special teams.
Don't pull an Al Davis (i.e., Ks and Ps are not draft-worthy).

As I mentioned at the end of Part 1, this overall strategy actually represents a conglomeration of strategies for 4 different 49er GMs (and coach-GM duos):

Carmen Policy, who ran the draft from 1994-1998
Bill Walsh, who ran the draft from 1999-2001
Terry Donahue, who ran the draft from 2002-2004
Scot McLoughan and Mike Nolan (aka McNolan), who ran the draft from 2005-2008

Of particular interest, above and beyond the Niners' overall draft strategy, are the individual strategies (and results) for each regime. There are a few reasons why knowing the regime info is important. First, it allows us to compare the 4 regimes both in terms of the strategies themselves, as well as their effectiveness at drafting starters and Pro Bowlers. I'm sure the conventional wisdom out there among 49er fans, for example, is that Walsh and Policy were more successful than Donahue and McNolan. So 2 potential questions that arise out of a knowledge of regime results are, "Is this conventional wisdom true?" and, "If so, are differences in draft strategy an explanation for why it's true?"

The second reason why it's important to know the regime strategies is related to the first. Namely, we can compare 49er regimes with those of other teams. I'll be doing just that in Part 4 of this piece. Finally, it's important to know the regime strategy for McNolan in particular because it gives us a rough blueprint for what to expect in this year's draft.

I know I said in Part 1 that I was going to tackle all 4 regimes in today's post, but I soon realized that such a post would be the longest in the history of blogging. Therefore, I've decided to change up the schedule a little bit. Rather than detailing all 4 regimes in one post, I'm going to break Part 2 into 4 subsections (2a-2d), one for each regime. So here's the revised schedule:

4/10/09 - Policy regime drafts (1994-1998)
4/11/09 - Walsh regime drafts (1999-2001)
4/13/09 - Donahue regime drafts (2002-2004)
4/14/09 - McNolan regime drafts (2005-2008) and comparisons
4/15/09 - Comparison of SF and NE team draft strategies
4/17/09 - Comparison of SF and NE regime draft strategies

Before I begin with the regime breakdowns, a couple of "buyer beware" caveats are necessary that will apply to each of the 4 posts. First, because we're talking about 118 picks spread out over 4 regimes, the sample sizes are pretty small for each specific regime. So I'll be trying as hard as possible to focus only on strategies and strategy differences that are extremely glaring. Second, because we're talking about a 15-year time span here, I'm going to be vigilant about keeping each regime's picks in the context of where the Niners were as a team when the picks were made. For instance, the team had increasingly deteriorating DEF stats at the end of the 1990's, so it's no wonder that Walsh took DEF player after DEF player from 1999-2001. The strategy means nothing without the context, and, in fact, this phenomenon illustrates an additional aspect of draft strategy in and of itself, i.e., whether or not each team (or regime) drafted according to needs.

After the jump, I'll put the Policy drafts in context; break the results down by position, draft day, and conference; and provide the idiot's guide to Policy draft strategy...

Here again is the link to my Excel spreadsheet of 49er draft picks in the Salary Cap Era, which can serve as both a permanent reference on Niners Nation and a resource for your own investigations into 49er draft picks:

49er Draft Picks 1994-2008

If you haven't taken a look at it yet, here's what's in it for those who are interested:

List of 49er picks and pick details (year, round, pick, position, height, weight, school, conference, BCS?, starter?, and Pro Bowler?) from 1994-2008
List of 49er picks and pick details from 1994-1998 (Policy drafts)
List of 49er picks and pick details from 1999-2001 (Walsh drafts)
ist of 49er picks and pick details from 2002-2004 (Donahue drafts)
List of 49er picks and pick details from 2005-2008 (McNolan drafts)
Number of picks in each overall and regime total broken down by round, position, and conference
Percentage of picks in each overall and regime row total (with breakdowns)
Percentage of picks in each overall and regime column total (with breakdowns)
Percentage of picks in each overall and regime total (with breakdowns)

OK, on with the show...

IN CONTEXT

First, the context...and what a context it was! From 1994-1998, the 49ers won just over 12 games per season, made the playoffs all 5 years, and won Super Bowl XXIX. Obviously, that success was built over the course of several drafts that preceded 1994, but Carmen Policy was nevertheless the primary architect beginning in 1991 (with the help of Vinny Cerrato).

So what possible needs could there have been for the 49ers during these halcyon days? More than you think:

1994 Draft - Key Losses: FB Tom Rathman (free agency), DL Ted Washington (trade), DL Kevin Fagan (retirement), and LB Mike Walter (retirement). Weak 1993 Stats: 23^rd in rushing TDs allowed, K had 61.5% accuracy rate. Lingering Issues: Still hadn't found a worthy replacement for LB Charles Haley; still couldn't stop the Cowboys' running game. Needs: FB, DL, LB, K.

1995 Draft - Key Losses: RB Ricky Watters (free agency) and DB Dieon Sanders (free agency). Weak 1994Stats: None. Lingering Issues: Except for William Floyd and Derrick Deese, entire starting OFF was 30 or older. Needs: RB, DB, OFF depth.

1996 Draft - Key Losses: WR John Taylor (retirement), OL Bart Oates (retirement), DL Rickey Jackson (retirement), and DB Eric Davis (free agency). Weak 1995 Stats: 23^rd in ST DVOA¹. Lingering Issues: Still hadn't found a worthy replacement for Watters; had major injuries to Floyd and Deese; old-timers QB Steve Young, OL Steve Wallace, and OL Harris Barton missed 13 total games due to minor injuries. Needs: RB, WR, OL, DL, DB, ST, FB depth.

1997 Draft - Key Losses: Wallace (retirement), Barton (retirement), and K Jeff Wilkins (free agency). Weak 1996 Stats: None. Lingering Issues: Still hadn't found a worthy replacement for Watters or Sanders; old-timers Young, TE Brent Jones, and LB Gary Plummer missed 15 total games due to minor injuries. Needs: RB, OL, DB, K, OFF depth, LB depth.

1998 Draft - Key Losses: Floyd (free agency), Jones (retirement), OL Jesse Sapolu (retirement, DL Dana Stubblefield (free agency), Plummer (retirement), and DB Rod Woodson (free agency). Weak 1997 Stats: OFF DVOA dropped to 13^th. Lingering Issues: had major injuries to DB Marquez Pope and old-timer WR Jerry Rice. Needs: FB, TE, LB, DB, OL, DL, LB, DB, WR depth.

For the most part, a vast majority of the needs were on offense, and overall, their needs had to do with youth and depth. However, rather than through the draft, the 49ers addressed most of these needs via free agency. Specifically, Jackson was signed to replace Fagan, RB Garrison Hearst was signed to replace Watters, Pope (feebly) and Woodson (fleetingly) were signed to replace Sanders, DL Chris Doleman was signed to replace Jackson, DB Darnell Walker was signed to replace Davis, K Gary Anderson was signed to replace Wilkins, OL Kevin Gogan was signed to replace Sapolu (after permanently moving OL Chris Dalman to C), and LB Winfred Tubbs was signed to replace Plummer. The point here is that the Niners' issues from 1994-1998 had more to do with age than performance, and yet they chose to address those issues by signing high-performing free agents. In NFL Dynasties 101, the Policy regime must have skipped over Chapter 10 of the required textbook, which is entitled, "Free Agency: A Faustian Bargain."

SECOND FIDDLE

From 1994-1998, the 49ers made 29 selections in the NFL draft, which works out to about 6 picks per season. Given that their overall rate has been about 8 picks per season over the past 15 years, you can already tell that the Policy regime was to blame when it came to trading away picks. As I just said, part of the reason for this had to do with the fact that they chose to address most of their needs during free agency. Another likely reason for the "fold ‘em" draft strategy - and this is obviously just speculation - is that the Policy regime saw the gaudy stats and continued playoff success during his tenure, and figured the best strategy in that context was to draft for immediate help - rather than depth or potential - by trading up for the blue chippers. Indeed, the Niners only had 4 picks in 1995, and no 1^st-round pick in 1996 because they sold the farm to take WR J.J. Stokes. Similarly, they only had 3 picks in 1997 because they traded their 5^th-, 6^th-, and 7^th-round picks to take - wait for it - FB Mark Edwards and TE Greg Clark.

So reframing the argument I made in Part 1, if we define "drafting for now" as getting younger and deeper at key positions, then the Niners did not "draft for now" when they were good during the Policy regime years; they also didn't "draft for the future" given the plethora of picks they traded away. Rather, it's more that they simply decided to basically not draft at all. That is, the draft played second fiddle to free agency.

ASSUME THE POSITION

So what did the 49ers do with the 29 picks they actually held onto from 1994-1998? Overall, the Policy regime took 16 OFF players, 12 DEF players, and 1 ST player. In terms of positions, they took 5 WRs, 5 LBs, 4 OLs, 4 DLs, 3 FBs, 3 DBs, 2 TEs, 1 QB, 1 RB, and 1 K. Relating these numbers back to team needs, they drafted Floyd to replace Rathman, DT Bryant Young to replace Washington, LB Lee Woodall to replace Haley, K Doug Brien to replace K Mike Cofer, WR Terrell Owens to replace Taylor, and DB R.W. McQuarters to replace Woodson.

Although this certainly indicates that they did address some positional needs in the draft, it's more interesting to notice (a) what positions they didn't address, and (b) the timing of their positional need picks. Specifically, Policy's 49ers drafted only 1 RB despite the obvious negative implications of losing Watters. Some people (like me) think that Watters' departure was the beginning of the end of the on-the-field dynasty, so its' mindboggling that the Policy regime chose to muddle through 1995-1997 with RBs Derek Loville and Terry Kirby. And the RB that the Policy regime did pick? That would be Stephen Pitts in the 6^th round; definitely a household name. Similarly, despite losing 4 Pro Bowl OLs during his tenure (not to mention the increase in Steve Young's injury proneness), the Policy regime only took 4 OLs in 5 drafts. He also took only 3 DBs despite losing both starting CBs and much of the secondary depth beginning in 1995. Indeed, some analysts were shocked (shocked!) that the Niners chose to trade away their 1997 picks for a QB, FB, and TE; rather than addressing their glaring, perennial needs at OL and DB.

Now, the second interesting piece of the positional need puzzle: timing. During the same time that the Policy regime basically ignored RB, OL, and DB - the Niners' 3 biggest needs - he sure did choose an inordinate number of WRs and LBs. Think about that for a second. The 49ers had the G.O.A.T., and they chose WR as the position on which to focus most of their draft picks. Granted, Taylor was getting old, so drafting to replace their #2 WR was reasonable. However, trading most of your draft to move up in the 1^st round is something you do to replace your #1 WR, not your #2. Now think about LB. They had - arguably - the best LB corps in the NFL from 1994-1997, yet they chose to focus most of their draft picks on that position.

You might be thinking to yourself, "but you said before that they were supposed to be using the draft for depth (and youth)!" You're right, but only if you ignore the context. My point here is that, when your major positional needs are RB, OL, and DB for 5 years running, and you have only 29 picks to work with, you don't go focusing your youth and depth solutions at WR and LB. Sure, hindsight is 20/20; and I know it's easy to look back and say Stokes, WR Corey Fleming, WR Iheanyi Uwaezouke, WR Ryan Thelwell, LB Kevin Mitchell, LB Anthony Peterson, LB Antonio Armstrong, and LB Sam Manuel were wasted picks given their performance (or lack thereof). But, to me, their (poor) performance is only relevant in the context of draft strategy: they shouldn't have been taking so many WRs and LBs in the first place, so missing on most of them only served to exacerbate their lack of attention to RB, OL, and DB. I mean, if after losing Watters, Sanders, Davis, and all those OLs, they would have drafted repeatedly for these positions, would we have b*tched as much if they didn't hit? I don't think so because at least they were giving it the ol' college try (pun intended). Alternatively, might they have hit more at these need positions if they took more bites at the apple? Probably, given their relative success overall (which I detail in the next paragraph). Instead, as fans, we were faced with a double dose of dynasty destruction: drafting busts at the "want" positions and ignoring the "need" positions. The Policy regime's drafts leave a sour taste because, in essence, he was fiddling while Rome burned.

Looking back, this lack of attention to need positions in the draft really is a shame because it's not like Policy and company were bad talent evaluators. To the contrary, 41.4% of their 29 picks became regular starters for the Niners, and 20.7% made the Pro Bowl as a Niner. Breaking these down by position, they drafted 3 starters at FB, 2 at WR and DB, and 1 each at TE, OL, DL, LB, and K. In terms of Pro Bowlers, they drafted 1 each at FB, WR, OL, DL, LB, and DB. Given the overall stats I presented in Part 1 (42.4% and 9.3%, respectively), Policy's regime was by far the best of the 4 when it came to drafting Pro Bowlers; they drafted more than the other 3 regimes combined (5). Of course, as I'm sure you realize, Pro Bowlers tend to come from winning teams; so maybe that's why the Policy regime "found" more of them. You know, it's a kind of chicken-and-egg thing. And just in case you didn't notice, that was 3 starters (30.0%) and 2 Pro Bowlers (20.0%) among the 10 total picks the Policy regime made at RB, OL, and DB. Although the lone DB Pro Bowler actually played S (Lance Schulters), that's still a good Pro Bowl rate; even more evidence in favor of my "bites at the apple" argument that more picks at these positions might have led to more need-based draft success.

Of their 5 drafts, the Policy regime's most impressive was their 1994 effort, which produced 4 players (Young, Floyd, Brien, and Woodall) who started as rookies for the Super Bowl XXIX squad, and 2 49er Pro Bowlers (Young and Woodall). Not far behind was the 1998 draft, which produced 3 starters (OL Jeremy Newberry, Shulters, and FB Fred Beasley), all of whom made the Pro Bowl as a Niner.

One last thing I'll say about positional needs, and how they were addressed. In case you were wondering, here are the Policy regime's positional picks by draft day:

Pos	Day 1	Day 2	Total
QB	1	0	1
RB	0	1	1
FB	2	1	3
WR	3	2	5
TE	1	1	2
OL	2	2	4
DL	2	2	4
LB	1	4	5
DB	2	1	3
K	1	0	1
P	0	0	0
Total	15	14	29

As you can see, they actually ended up with nearly an even split of picks between Day 1 and Day 2 despite trading away so many. That's because trading away Day 2 of the 1997 draft balanced out the effect of trading away Day 1 in 1995. In addition, the only position in which there seemed to be any draft day preference whatsoever was at LB (Day 2), which I suppose softens the blow of picking too many LBs. However, lack of attention to positional needs on Day 1 (0 RBs, 2 OLs, and 2 DBs) re-hardens the blow (that's what she said; sorry, had to).

Bottom line - Here's what I think are the main things to take away from this section:

The Policy regime mostly took WRs and LBs despite having more pressing needs at RB, OL, and DB.
The Policy regime traded away a lot of picks to move up in the draft for positions they didn't need.
The Policy regime had a Pro Bowl rate that was over twice the 49er average from 1994-2008, and the rate at need positions suggests more attention should have been focused on them.
Except for taking more LBs on Day 2, the Policy regime spread their positional picks out evenly between draft days. In other words, there was no Day 1 focus on glaring positional needs.

CONFERENCE ROOM

As was the case with the overall total from 1994-2008, the highest percentage of Policy regime draft picks came from the Pac-10 (6). Among the other conferences represented, there were 4 Big 12 picks, 3 Big 10 picks, 3 Big East picks, 3 Notre Dame picks, 2 Big West picks, 2 SEC picks, 1 ACC pick, 1 Atlantic 10 pick, 1 MSFA pick, 1 PSAC pick, 1 Southern pick, and 1 WAC pick. What stands out here is their affinity for picking players from schools that are close to home. Over half of the picks (65.5%) came from conferences that mostly play in or west of the Central time zone, which isn't so coincidental given the Niners' location, and that of Eddie Debartolo's home (Youngstown, OH) and alma mater (Notre Dame). Also, it didn't hurt that Policy himself was born, and got his undergraduate degree, in OH.

If we break it down according to the BCS, we find that 75.9% of the Policy regime's picks came from BCS schools. Going back to the 49ers' overall BCS rate from 1994-2008 (79.7%), it seems that the Policy regime was par for the course when it came to drafting big-school talent. What's interesting about this, though, is how good the Policy regime was at finding small-school diamonds in the rough: Woodall (PSAC; Division II), Owens (Southern; Division IAA), and Schulters (Atlantic 10; Division IAA) all made the Pro Bowl as a 49er after being drafted from non-Division-IA - let alone non-BCS - schools. Indeed, although their BCS and non-BCS starter rates (40.9% and 42.9%, respectively) were as expected from Part 1, the Policy regime's Pro Bowl rate for non-BCS picks was 42.9%, which dwarfs both their Pro Bowl rate for BCS picks (13.6%) and the Niners' overall non-BCS Pro Bowl rate in the Salary Cap Era (12.5%).

When it comes to positions, the Policy regime pretty much followed the 80% rule of overall 49er draft strategy. The only exceptions were at DB (66.7%), LB (60.0%), and TE (50.0%). However, when combining positions into team units, the conclusion is different. In line with expectation, 81.3% of the OFF picks were from BCS schools. However, only 66.7% of the DEF picks came from BCS schools, which is well below the overall 49er rate of 79.3%.

Here's a table showing the number of BCS and non-BCS players that the 49ers have selected on each day of the draft:

BCS?	Day 1	Day 2	Total
Yes	14	8	22
No	1	6	7
Total	15	14	29

Not surprisingly, the Policy regime took 93.3% of their Day 1 picks from BCS schools, which is in line with the 89.8% overall BCS rate for the Niners. However, on Day 2, rather than the non-BCS rate tripling as it did in the overall numbers, it instead increased 6-fold to 42.9%! Clearly, the Policy regime absolutely adored those diamonds in the rough, most likely because he was actually good at finding them.

So I think I've established that the Policy regime was good at picking players from non-BCS schools, and that they picked more of these non-BCS players on Day 2. To further illustrate this point, the tables below show starters by draft day broken up into BCS and non-BCS picks:

BCS Picks
Starter?	Day 1	Day 2	Total
Yes	8	1	9
No	6	7	13
Total	14	8	22

Non-BCS Picks
Starter?	Day 1	Day 2	Total
Yes	1	2	4
No	0	4	3
Total	1	6	7

Among players taken from BCS schools, the 49ers' starter rate dropped from 57.1% on Day 1 to 12.5% on Day 2. Likewise, the non-BCS starter rate dropped from 100.0% on Day 1 to 33.3% on Day 2. However, what's interesting here is that the Policy regime's starter rate was better for non-BCS picks on both draft days. Not to beat a dead horse, but the Policy regime hit on Day 2 non-BCS players at an inordinate rate when compared to the overall 49er stats from 1994-2008. Again, this begs the question, "Why, then, did they trade away Day 2 in 1997?" For Edwards and Clark no less! Your guess is as good as mine.

Bottom line - Here's what I think are the main things to take away from this section:

The Policy regime took players from conferences that were close to home.
The Policy regime had a higher BCS rate for OFF picks than DEF picks.
The Policy regime took more non-BCS players on Day 2 of the draft than Day 1.
On both draft days, the Policy regime had better starter and Pro Bowl rates for picks from non-BCS schools.

GENERAL ANALYSIS AND CONCLUSION

During the course of doing my research and stats for this article, it became apparent to me that the period in 49er history from 1994-1998 should be named, "Auto Pilot." I think both the Policy regime and 49er fans entered into a mindset that the dynasty would never end; that there was some supernatural force bestowing Hall of Famer after Hall of Famer on SF; that the mere presence of a West Coast offense, a Walshian approach to football, and a benevolent owner was enough to bring wins in perpetuity. The fact of the matter is that what kept the dynasty going for so long was a constant replenishment of superior talent through the draft. It was saying, "We're great, but we want to be perfect, and the way to be perfect is to get younger and deeper." If that meant letting an old-timer go or "gently" forcing him into retirement so that his younger understudy could get a chance to shine, that's what was done (See Young, Steve vis-à-vis Montana, Joe; Romanowski, Bill vis-à-vis Turner, Keena, etc.). Rather than by nibbling away at the edges of the envelope, the dynasty sustained itself by replacing the old envelope with a new one throughout its lifespan. To put it as poignantly as possible, Walsh saw a dynasty and tried to perfect it through the draft, whereas Policy and company saw a dynasty and wanted to put it on auto pilot.

I realize that there was no free agency before1993, but focusing on it after 1993 -at the expense of the draft - was the epic fail of the Policy regime. Rather than getting rid of the old guys and replacing them through the draft, they sat on their laurels, watching the old guys get older, and tradied away the valuable draft picks that could be used to replace them. What's worse is that, when it eventually came time to replace the free agent losses and retirees, the Policy regime chose free agency - rather than the draft - as its primary means.

In case you didn't know, I'm a Yankee fan. Don't fret, though. I'm not one of those obnoxious "the Yanks can do no wrong" or "Hey, there's no salary cap, so we're just playing by the rules when we spend $200M a year on payroll" types. On the contrary, I'm hypercritical of what the Yankees (esp. Brian Cashman) have done for the past 8 years; to the point now that I have a hard time watching their games or getting emotionally involved in their seasons. The reason I'm hypercritical is because it's a déjà vu experience; the same thing that I saw with the Niners in 1995.

The Yanks' success from 1996-2000 was achieved on the backs of homegrown talent (Jeter, Posada, Rivera, Pettitte, Bernie, etc.). When they used free agency, it was to sign the second-tier guys like Paul O'Neill and Tino Martinez. Since Cashman became GM, however, they watched the old guys get older, traded away their prospects, and signed high-priced replacements via free agency. What have they gotten out of it? Over the past 8 years, they've gotten a lot of deceptively good seasons, but no championships.

And you know what? That's exactly how I look at the Policy years in SF. Rather than going with what got them there (i.e., the draft), the Niners chose the high-priced free agent route. They watched their stars get old, and did almost nothing. And what did it get them from 1995-1998? A lot of deceptively good seasons, but no championships. As is the case with the Yankees today in MLB, the Policy regime's strategy was no way to sustain an NFL dynasty in the mid-90s. Furthermore, as is the case with the Yankees today, what was needed in 1999 was for someone to blow up the Niners' franchise and go back to its drafting origins. Indeed, it wasn't until Bill Walsh himself retook the reins of the organization that the Policy regime's free agency attachment and draft neglect were stopped once and for all.

But that's the subject of Part 2b, which I'll discuss in detail tomorrow. So, with my diatribe complete, here's how I would describe the Policy regime's draft strategy, a strategy to which we'll be comparing subsequent regimes:

Draft picks are commodities best used for moving up in the draft.
Once you've addressed positional needs in free agency, there's no need to focus on them in the draft.
Dominate the neighborhood (i.e., take Pac-10 players and those near Youngstown, OH).
There's plenty of talent outside the BCS conferences, but wait until Day 2 to acquire it.

That's it for now. On Saturday, I'll break down the Walsh drafts (1999-2001). TO BE CONTINUED...

¹ DVOA statistics used to produce this article were obtained from Football Outsiders.

Poll

How much blame would you assign the Policy regime's draft neglect in causing an end to he 49ers' dynasty?

All of the blame

45 votes

Most of the blame

135 votes

Some of the blame

147 votes

None of the blame

36 votes

363 votes | Poll has closed