STANFORD, CA - OCTOBER 01: Andrew Luck #12 of the Stanford Cardinal stands on the sidelines during closing minute of their game against the UCLA Bruins at Stanford Stadium on October 1, 2011 in Stanford, California. (Photo by Ezra Shaw/Getty Images)
In this week's Morning Tailgate Mailbag, Bill takes a look at surprisingly strong defenses in Stillwater and Palo Alto, the difficulty in retaining amazing numbers against terrible teams and the impact of Andrew Luck. Plus, a numbers primer! Who doesn't like numbers primers??
Some fun, relevant questions this week. Keep them coming!
Two things quickly become evident watching Stanford:
1) Andrew Luck is really, really good.
2) That team is not just Andrew Luck.
Like Oklahoma State, Stanford is a team that has posted better defensive numbers than offensive (16th in Off. F/+, seventh in Def. F/+) even though the marquee players are on the offensive side of the ball. They are incredibly efficient, especially against the run; they are 10th in Adj. Line Yards Allowed and first in Adj. Sack Rate, meaning their defensive line is one of the best, if not the best, in college football. Fourteen players have taken part in at least one sack, 19 in at least one tackle for loss. At their current rates, senior linebacker Chase Thomas, junior end Ben Gardner and possibly junior linebacker Trent Murphy might all end up with double-digit tackles for loss, and in all, the defense's rates have not regressed without injured star linebacker Shayne Skov. They overpower you, put you on your heels, then assault you from every possible angle.
They are also, as currently constituted, not a team dominated by seniors. We all assume Andrew Luck is gone after this year (and he probably is), but even assuming that, seven current offensive starters and six current defensive starters are projected to return next year. This team should remain interesting after Luck's departure, even if the passing game (which should lose Luck, senior receivers Chris Owusu and Griff Whalen, and tight end Coby Fleener) quickly regresses. The defense should carry the team to satisfactory results even if the Cardinal are no longer national title contenders.
The question was about Luck, however, so let's do some quick math.
At this moment, the 120 FBS teams have run 53,501 plays, gained 302,665 yards and scored 22,106 points. That results in averages of 5.7 yards per play and 13.7 yards per point. Andrew Luck and the Stanford passing game have averaged at stellar 9.5 yards per pass attempt (not including sacks) thus far; the national average: 7.2.
For the purposes of this rough hypothetical, let's say that next year, with Brett Nottingham (or whoever) at quarterback for the Cardinal, Stanford regresses from a Top 5 passing offense to one that gains the national average. Meanwhile, they continue to attempt 31.8 passes per game. Both of these assumptions could be woefully inaccurate -- the passing game could completely fall apart, and Stanford could lean much more heavily on the run -- but let's go with it. That results in regressions of 2.3 yards per pass and 73.1 yards per game. If 13.7 yards are worth approximately one point, then 73.1 are worth 5.3 per game. Doesn't seem like much, does it? Realize this, however: in 2011, Stanford has won games by 54, 30, 27, 26, 41 and 30 points. So even if Luck had gone pro last year, and his departure resulted in a regression of 20 points per game, the Cardinal would almost certainly still be undefeated right now. Would they be good enough to beat Washington, USC (in Los Angeles), Oregon and Notre Dame? Probably not. But there is enough talent on this Stanford team that, even without Luck, they would still potentially go 8-4 or 9-3 with this schedule. This year almost certainly represents the peak year in Stanford's current upswing, but they should remain competitive, at least in the near future, following Luck's departure.
In your 10/7 Mailbag you discussed the wide margin of Bama's Def S/P+ score vs everyone else's. One week later, Boise State is #1. I know they held Fresno State well below their average offensive numbers on the season, but Fresno's really no juggernaut. So what caused the swing? Do you have any expectations for how these numbers will look by season's end?
-- Dave S.
The top of the Def. S&P+ list is a nice reminder of the small sample sizes we are dealing with at the moment, and a nice reminder not to get carried away with superlatives (a recommendation I am sure to ignore). Through five weeks, Alabama's offense had been otherworldly, but after two games in which they merely played well against two awful offenses (Vanderbilt and Ole Miss), they slipped from an easy No. 1 in Def. S&P+ to No. 4.
These past two weeks, Alabama's defense proved what Boise State is forced to prove all the time: it is difficult to maintain a truly dominant level of play when you are playing opponents of a lower caliber. Vanderbilt currently ranks 104th in Off. F/+, Ole Miss 111th. They have truly pitiful offenses, and for Alabama to maintain their numbers from a couple of weeks ago, they would have almost literally have needed to hold the Commodores and Rebels to under 100 yards and no points. They allowed just seven points and 331 yards, but that was merely awesome, not transcendent. Meanwhile, Boise State held Fresno State (a decent 62nd in Off. F/+) and Colorado State (94th) to a combined 501 yards and 20 points and made up ground.
It is still early enough in the season, and these defenses (along with those of LSU, Oklahoma and Penn State) have been so good, that the rankings could still oscillate quite a bit from week, depending on who merely dominates and who destroys wills to live. But I'll just say this: there is no defense in the country I feel more confident in than Alabama's, and I assume they will finish in the Top Two when all is said and done. (That said, they also have a good enough offense that they could slide to Top 5-10 and still win the national title.)
OkSt off is disappointing in FEI/S&P (#38/#9) yet classified as explosive, their def is mocked, yet #4/#9 in FIE/S&P. What gives?
With my Mizzou hat on, I tackled this topic yesterday at Rock M Nation. When you dive into the statistics, something becomes quite evident about this Oklahoma State team: they become more like Oregon every day. It's been an easy comparison to make for a while now; hell, I made the comparison three months ago, and I was far from original in that regard. Let's look at some of the similarities...
1) The Sugar Daddy. Both Oregon (Phil Knight) and Oklahoma State (T. Boone Pickens) find themselves in a situation where a single donor/booster has both donated an incredible amount of money to the program and developed a strong amount of sway. Oklahoma State now plays at T. Boone Pickens Stadium. You can perceive each relationship as dependent or unseemly if you want, but each has been incredibly important for the development of both the football and overall athletic programs.
2) The Recent Surge. Oregon had been a solid program for quite a while, but riding a dynamic offense, they took a considerable, sustained step forward about four or five years ago. Oklahoma State, meanwhile, is 35-10 since the start of the 2008 season. Oregon made the national title game last year, while Oklahoma State is 6-0 and should be favored in every game between now and what could be an incredible Bedlam battle with Oklahoma on December 3.
3) The Uniforms. My god, the uniforms.
4) The Statistical Oddity. This has been the most recent line of symmetry between the two, and the subject of this question. Last year, Oregon drew attention because of their explosive, hyper-speed offense, but from a per-play basis, their offense was only very good, not transcendent. Meanwhile, their defense was quite underrated, limiting big plays, making a ton of their own, and carrying the offense in the rare moments where it was a necessity. Without a speedy, opportunistic defense, the offense would have drawn the same acclaim.
Thus far, it is exactly the same story for Oklahoma State this year. Theirs is one of the most efficient offenses in the country, but they are succeeding as much because of pace (which, as we've discussed before, maximizes a per-play advantage, which they have obviously had in every game this season) and defense as Brandon Weeden and Justin Blackmon. The passing game has actually been somewhat limited in the big-play department -- of the team's top six targets, only converted tight end Tracy Moore averages more than 13.0 yards per catch, and onlyHubert Anyiam averages over 8.4 yards per target -- but it hasn't mattered because Weeden is more than capable of engineering nine-play drives without screwing things up, and because when the offense does stall, the defense does its job.
The Cowboys have the most underrated set of safeties in the country in Daytawion Lowe, Markelle Martin and Deion Imade. They are good but not great in the front seven, but with those safeties and some aggressive cornerbacks, they prevent big plays and take advantage of whatever mistakes you make. They are a surprise entry in the Def. F/+ Top 5, but thus far they have earned it. Bend-don't-break is hard to maintain from year to year, but it is absolutely working for OSU right now.
Is there a guide to F/+ somewhere? I googled it and got Ford stock quotes and other nonsense.
I fear sometimes that I have the tendency to start speeding down the road without everybody in the car. I mercilessly use the F/+ and S&P+ measures without going out of my way to make sure people know what the hell I am talking about. For starters, go to the primary F/+ page at Football Outsiders. It gives you the basic explanation, which is simply that F/+ is the combination of S&P+, my opponent-adjusted play-by-play measure, and FEI, Brian Fremeau's opponent-adjusted drive efficiency measure.
I love F/+ for one simple reason: it combines two solid, informative measures and creates one that is more predictive and evaluative. You cannot win thousands of gambling dollars using F/+ (because Vegas very, very much knows what it is doing), but you can get an infinitely better feel and grasp of quality using these measures than simply looking at yards, points and win-loss record.
The measures are presented in different formats. FEI is presented using thousandths of a point, with 0.000 being the national midpoint and anything over 0.300 signifying a really, really good team. S&P+'s midpoint is 200.0; this is because both Off. S&P+ and Def. S&P+ are presented with the idea of 100.0 representing something perfectly average (and S&P+ representing the combination of Off. S&P+ and Def. S&P+). Anything over 200.0 is better than average, anything below is worse. Why are FEI and S&P+ presented in different formats? Simple: Brian and I created these measures completely and totally independently of each other. They were hatched in their own incubators and combined well after their development.
To present F/+ as cleanly as possible, I simply display it in terms of a percentage. Right now, Arizona State, for instance, is 20th with a "+14.3%" F/+ rating. What does that mean? it means they are 14.3 percent better than the average team. Meanwhile, No. 100 Kentucky is 14.7 percent worse than the average team.
More information is presented in the Football Outsiders Almanac 2011, which is available in print or download at Football Outsiders (download the college portion for just $5!) and available in print at Amazon.com. It is a strong publication, I love that my name is associated with it, and I recommend you picking it up.
One more thing: beginning next week, I will be attempting to provide a weekly stat workbook of sorts at Football Study Hall. Hopefully giving individuals the open opportunity to play with these numbers will result in a more widespread understanding of them.