SB Nation User Blog: PoliSam

Sergio Wants Out "I don't know what my future (is) going to be," Rodriguez said. "I think...

PoliSam — Sun, 03 May 2009 17:40:18 -0000

Sergio Wants Out

"I don't know what my future (is) going to be," Rodriguez said. "I think something is going to happen, either a trade, or something. But I think for me, I feel like I need a change. I need to play a different way. I really enjoy playing three years here. But I think I need something (different). I feel after three years, I need a change."

http://blog.oregonlive.com/behindblazersbeat/2009/05/pritchard_the_puzzle_master.html

Most disappointing/worrisome player performance? (poll)

PoliSam — Sat, 02 May 2009 00:10:06 -0000

I know we are starting to move on from the Houston series, but before we do, I wanted to settle an argument that I have been having with myself.: the Blazers lost several player match-ups in the Houston series, but I have not been able to figure out which is the most disappointing or worrisome development. I figure the best way to do it was to rely on The Wisdom of Crowds. In addition, I have never done a poll before and thought this would be an approrpiate topic. It's a somewhat negative topic and I generally dislike dwelling on negative outcomes, but I could not resist.

In choosing which is the most disappointing, use whatever criteria you think is appropriate, though I personally had in mind the long term implications of the series (what it revealed about the "true" underlying value of the player in question) AND how surprising it was to me.

Comments are welcome.

Poll

Which of the following was the most disappointing/worrisome aspect of the series with Houston? (note: if you feel that any of the statements are false, simply do not pick that answer)

Luis Scola outplaying LaMarcus Aldridge

41 votes

Aaron Brooks outplaying Steve Blake

38 votes

Van Wafer outplaying Rudy Fernandez

25 votes

Chuck Hayes outplaying Greg Oden

29 votes

Travis Outlaw not outplaying anyone

231 votes

Roy not playing as well as LeBron or Jordan (circa 1993) would have

7 votes

Other - Comment

9 votes

380 votes | Poll has closed

Oden off the Bench? An Analysis

PoliSam — Tue, 17 Mar 2009 14:29:12 -0000

Reports suggest that Oden will soon start playing again for the Blazers. There has been a lot of discussion about whether he should start or come off the bench when he returns. My guess is that he will come of the bench, simply because Nate tends to be risk adverse when making rotation decisions. Nonetheless, I wanted to see what an analysis of past games would say about the subject. My analysis suggests that the team has played quite a bit better with Oden coming off the bench, about 6.3 points better than it has played on average. By constrast, when Oden started the Blazers have played at their season average.

Measuring the Oden off Bench and Oden Starting "Effects"

Building off the methodology I used in a previous post I estimated the "effect" of starting Oden vs. bringing him off the bench. In particular, I compared the margin of victory in games that Oden came off the Bench to those in which Oden started, controlling for the following factors:

Opponent's Average Margin of Victory (so Cleveland is Good, Sacaremento is Bad)

Opponent's Average Home/Away Advantage/Disadvantage (which varies by team, see previous post).

Either Team playing on a Back to Back, which hurts teams, on average, about 1 point per game.

Whether or not either Team played 5 games or more in 7 days, which hurts teams, on average, about 1.7 points per game.

Whether or Not Roy Played. The Blazers have played about 7 points better per game when Roy has played than when he was out with injury.

Whether or Not Rudy Played. The Blazers have played about 10 points better when Rudy has played than when he is out.

Whether or Not Blake Played. The Blazers have played about half a point better when Blake has played than when he was injured (counting the Philly and Clippers games as injured games).

By my count, Oden came off the bench in 7 games and started 39 games. Controlling for the factors above, the Blazers have played about 6.3 points better with Oden comeing off the bench than in games where he started or was injured. By contrast, when Oden started, the Blazers have played no better or worse than average. Note that if the Blazers had played better than average with Oden as a starter, then one would have to compare the Oden Bench advantage to the Oden Starting advantage. However, because the Blazers played no better when Oden started than when he was injured, the "benefit" of playing Oden off the bench is simply 6.3 points per game.

Since Oden only came of the Bench in 7 games these results should certainly be seen as tenative. Indeed, the "Oden Off the Bench Effect" is not statistically significant, if one were to use a standard hypothesis test. That being said, it is interesting to look at the seven games he came of the bench to investigate the matter further. The following are Oden's Bench games:

Team1	Points1	Team2	Points2	Unexpected Margin	H/A
Portland	104	Miami	96	10.23562	A
Portland	82	NewOrleans	87	-0.8711184	A
Portland	88	Minnesota	83	-0.9047602	A
Portland	106	GoldenState	111	-2.983838	A
Portland	116	Chicago	74	32.24477	H
Portland	117	Sacramento	96	14.08401	A
Portland	92	Phoenix	102	-5.479679	A

The Unexpected Margin, is the margin of victory over what was expected, controlling for all of the factors listed above. The Chicago game is certainly an outlier and helps Oden's Bench average, but its important to note that there are no terrible games in that seven game stretch, which is significant in itself. Indeed, the worst game was a loss to the Suns in Phoenix before Amare got hurt and when the Suns were playing relatively well. My guess is that its hard to get pummeled when your back-up center collects 5 offensive rebounds.

Caveats

I put effect in quotes because this is certainly not definitive evidence that the Blazers should bring Oden off the bench rather than start him. There are a variety of alternative explanations for the team playing better with Oden off the bench that I cannot rule out. Among other things, it may have taken teams a few games to figure out how to defend Oden; injuries to Deng in the Chicago game and Martin in the Kings game probably overstate the results; and as mentioned above, the sample size is small and margins of victory are quite noisy-the result could be a fluke--the equivalent of winning a hand of Texas Hold'em starting with a pair of sevens (which is an okay, but not great starting hand).

Discussion

That being said, this analysis does suggest that there is a real chance that the Blazers will get a nice little "Oden Bump" if he, once again, comes off the bench when he returns. In fact, the model used to do the analysis above predicts that the Blazers would finish the season 51-29 31 if Oden does not return at all. If he returns and comes of the bench Wednesday, the model predicts that the Blazers will win 54 games. That difference is enough to give the Blazers home court advantage in the first round. He could easily make the difference in games against at Indiana and Milwaukee and would give the Blazers a better chance to beat LA, Houston, Denver, and San Antonio.

Will Nate continue to start Pryzbilla when Oden returns? Will Nate do it for the rest of the season? Should he? Does the analysis above pick up something real or was it a fluke? If its real, why does Oden fit better with the second unit? Would Oden still benefit from playing in the second unit if Bayless remains the back-up point guard? This one will be intersting to watch.

Home Court Advantage in the NBA, 2008-2009

PoliSam — Mon, 16 Mar 2009 17:27:52 -0000

With the Blazers continued road woes and the importance of scheduling in the Western Conference playoff race, I thought I'd take do a little analysis of home court advantage in the NBA this season.

Measuring Home Court Advantage

The simplest way to measure home court advantage is to compare a team's home court and away records. Because schedules in the NBA are unbalanced, however, it is possible that some teams have played more difficult schedules at home than away. In addition, as John Hollinger and others point out, margin of victory is a better predictor of future record than wins and losses. Since most of us are interested in projecting how the rest of the season and playoffs will unfold, it makes sense to measured home court advantage by margin of victory rather than wins and losses.

To create a measure of home court advantage for each team, I started by downloading a log of all of the 2009 regular season games from basketball-reference.com (through Saturday March 14th). Then, if the game went into overtime, I set the margin of victory to zero. Then I used standard statistical/econometric methods to control for strength of schedule and create an estimate of each team's home court advantage. (For those interested in the technical details, I created the measure by regressing margin of victory on a set of dummy variables for each team, a dummy variable indicating which team was home and set of team*home court interaction terms. The team dummy variable is the team "Fixed Effect", it accounts for each teams average margin of victory. Each team's home court advantage is the coefficient on the home court variable plus the team*home court interaction. The team*home court interaction is described as the "Extra" Advantage). The results are below.

Results

Home Advantage is the difference between a team's average home margin of victory and average road margin of victory, controlling for their opponents' average margin of victory and their opponents' home v. away advantage. Note that, by construction, a team's away disadvantage is just their home advantage multiplied by -1. The average home-court advantage in the NBA this season is 7.87 (if overtime games count as ties).

"Extra" Advantage is the difference between each team's home court advantage and average home court advantage in the NBA. As with the raw numbers of home court advantage, a high number can be taken to indicate that a team plays well at home or poorly on the road. Similarly, low numbers can indicate that a team plays well on the road or poorly at home; it's impossible to distinguish between those characterizations of what the numbers mean.

Team	Home Advantage	"Extra" Advantage
GSW	15.39	7.52
UTA	15.30	7.43
POR	11.39	3.52
CLE	11.23	3.36
HOU	10.34	2.47
MIL	10.26	2.39
IND	9.84	1.97
MIA	9.54	1.67
CHI	9.45	1.58
PHI	9.37	1.50
ATL	9.27	1.40
PHO	8.97	1.10
DEN	8.48	0.61
OKC	8.44	0.57
DAL	8.43	0.56
CHA	7.70	-0.17
NOH	7.64	-0.23
NYK	7.46	-0.41
ORL	7.36	-0.51
TOR	7.12	-0.75
SAC	6.96	-0.91
LAL	6.38	-1.49
BOS	6.19	-1.68
WAS	5.42	-2.45
MEM	5.38	-2.49
SAS	5.28	-2.59
DET	2.93	-4.94
LAC	2.73	-5.14
NJN	1.86	-6.01
MIN	-0.09	-7.96

Discussion

The results conform, more or less to my expectations, with a few surprises. Portland has a "larger" home court advantage than the average team, but Utah and Golden State have even larger home court advantages (or road disadvantages). It's interesting that thee top four teams in terms of "extra" advantage are also relatively young, while three of the teams with the smallest home advantages are also young. It is amazing to see that Minnesota has actually played worse at home. I think the depressing faces that Phil Jackson referred to must be in Minneapolis-St. Paul, rather Portland.

Obviously teams like Utah and Portland would like to be able to play as well on the road as they do at home. However, given that the advantage of playing at home seems to be connected to how loud fans are... I'm not sure it's possible to just flip a switch and make the team play as if 20,000 people are enthusiastically supporting their every move. Perhaps some improvement can come with experience, but I would not expect any big changes this season. Thinking about this a bit reaffirms to me the importance of the Blazers getting home court advantage in the playoffs.

A few limitations of the method that I used: it assumes that each team's "strength" and home court advantage do not vary systematically throughout the season. Each team plays poorly some games and well in other games, but any variance other than home court advantage and opponent strength is (random) noise. One way to improve the analysis above would be to include "month" indicators for each team to account for streaks that occur during the season. There is no reason that "hot streaks" would care about the calendar, however, so this would be a pretty dirty fix. Another approach is to include lags and leads of margin of victory. I played around with this a bit, but saw little evidence of auto correlation. A more promising approach would be to include variables that indicate when key players have been out due to injury (or even playing while injured). If anyone knows of a nice data set on player injuries, I'd be happy to give that a shot. Indeed, I've already looked into this for the Blazers and the results suggest that having key players out matters quite a bit, in some cases. More on this later.

More Jefferson rumors from Stein at ESPN

PoliSam — Mon, 16 Feb 2009 15:36:19 -0000

More Jefferson rumors from Stein at ESPN

-Not really new, but additional confirmation that Jefferson + Ridnour is being discussed.

-A claim that Washington is not trading Butler.

-A few nuggets about Vince.

The No-Stats All-Star

PoliSam — Sat, 14 Feb 2009 17:35:04 -0000

There is a great article in the New York Times Magazine this week by the author of Money Ball, Michael Lewsi. It is about Shane Battier, Houston's GM Daryl Morey, and the "new" stats in basketball: The No-Stats All-Star. It's a must read.

A few highlights, comments:

-Some great discussion of exactly how and why Shane Battier is a good defender.

-One of the interesting claims made by Lewis and Morey is that in basketball, unlike Football or Baseball, the traditional statistics give players an incentive to be selfish. I always knew that to be true in basketball, but never realized that it was less of an issue in other sports.

-Some hints about the types of statistical analysis that teams keep "in-house." Stuff that is not publically available that is useful to teams when scouting.

-There was a suggestion that Houston believes that it's very important not to foul. That fouling is the worst type of defensive play. Greg Oden, I'm looking at you.

-Frankly, I would have liked to see a bit more discussion of the details of the "new" statistics and a bit broader coverage (discuss other teams that are interested in this trend).

I'm sorry for being immature, but this picture deserves some comment. Please submit your best...

PoliSam — Thu, 12 Feb 2009 22:02:48 -0000

I'm sorry for being immature, but this picture deserves some comment.

Please submit your best mock captions below. I know there are a lot of comedic geniuses here at Blazersedge. Make us laugh.

2nd Update: NBA from 1951 to 2009, POR, LAC, LAL

PoliSam — Fri, 12 Dec 2008 16:49:08 -0000

In response to the questions and comments on my previous post on ages and success in the NBA, I did a little additional work. I could not do everything that people asked about, but I tried to improve the previous analysis in four ways:

1. More Data!

It is relatively easy to download the entire statistical history of the NBA at databaseBasketball.com. Minutes played data "only " go back to the 1950-51, so that is the earliest season included below. It also does not include the current season, but I manually added the 2008-2009 Blazers. The coolest thing about having older data is that it gives us a target for a (somewhat goofy) record: the best record for a team whose average age is under 25 in the (modern) history of the NBA.

2. Better Measurement of Age

The databaseBasketball.com data has players' date of birth, which makes it easy to get calculate their age to the day (ie. 24.53 years rather than 24). Each player's age for each season is determined by their age on December 31st of the given season. This may not seem like a big deal, but it actually made this year's Blazers about half a year older.

3. Years of Experience

Several people were interested in seeing the results if the measurement of maturity was years of experience, rather than age. I measured experience in two ways. The first, which I refer to as "Years of Experience" is just the number of seasons since their first seasons in the league, counting their rookie season as their first season. So, if a player's first season in the league was 1990, they would have one year of experience in the 1990 seasons and six years of experience in the 1995 season. The weakness of this measure is that it counts a year of experience whether or not a player played any games. So, I created a second measure, referred to as "Years of Game Experience" that counts a season of experience as playing 77 games. Why 77 and not 82? It's actually fairly rare for anyone to play all 82 games of a season. Why 77 rather than 72? In short, I just picked a number, but since everyone's years are scaled in the same way it does not matter too much. In addition, the two measures ("years of experience" and "years of game experience") tracked fairly closely for star players like Kareem or Bird when I used 77 games.

4. Tracking Team Trajectories

A lot of people were interested in seeing the trajectories of particular teams over time. As you will see, I tracked a few team trajectories in only the most simplistic way: a did three seperate graphs that highlighted the team-age relationships for three franchises: the Blazers, Lakers, and Clippers. For those that really want to try to make projections about the future, this is probably not enough. It is, however, fun to see. Moreover, I did not see an obvious way to use the data to make projections that would be any more definitive than what we would all do in our heads.

Win % by Average Age, 1951 to 2009

So, what are the results? First, I plotted win percentage by the average age of the team on the floor. I also drew a line for the average win% pct given average age, just for reference. For an explanation of how I calculated average age and drew the line, go here... The two best under 25 teams are in red. This first graph is almost too awesome to believe. Keep in mind that this is after the Blazers lost two games in a row... and comes with the Blazers having played a brutal schedule so far (the graph looked even crazier before last night's game):

via i36.tinypic.com

Yes, this year's Blazers top competition for best record by a team whose average age was under 25 years is the 1976-1977 Portland Trailblazers. Here is a link to see the ages and minutes played for that legendary Portland team. 1977 was the year I was born, so I did not know that that team had a relatively pedestrian regular season record. Obviously, that is also the youngest team to ever win an NBA championship. That's one achievement that I do not think this year's team can match.

Win % by Years in the League, 1951 to 2009

How do the 2008-2009 Blazers compare when maturity is measured by years experience in the NBA? In some way this helps this year's team because of one of it's rookies, Rudy Fernandez, is an older rookie. On the other hand, Travis Outlaw and Aldridge have a lot of NBA experience for their age. As in the previous graph, in the graph below, I plotted win% by the average years of experience for the team. I highlighted in purple the teams with the best record and an average of less than 4 years of experience:

via i38.tinypic.com

First, note that when maturity is measured by years of experience rather than age, there is still a fairly strong maturity-success relationship. Second, there are actually a few teams more exceptional than the Blazers, including the 1991-1992 Golden State Warriors with Mullin, Hardaway, and Marciulionis and the 1970-1971 Milwaukie Bucks with Oscar Robertson and Kareem in his second season. That being said, the 2008-2009 Blazers are doing remarkably well by this measure.

Win % by Years of NBA Game Experience, 1951 to 2009

As I said above, I thought that these results might be misleading if a player hardly played his first few seasons or missed a season or two with an injury. While some experience can be gained from just being around the team and practicing, most people think game experience is what players need and what this year's team needs. In addition, this year's Blazers are doing well early in the season even with three rookies seeing significant minutes. The graph below follows the same format as the previous two, but with the best teams with less than an average of 3 years game experience in green:

via i38.tinypic.com

My guess is that this measurement of experience probably exaggerates the Blazer youth to some extent, but still reinforces that the success of the team is pretty extraordinary.

Team Trajectories: Blazers, LA-Lakers, and LA-Clippers

Lastly, I thought it would be fun to look at a few franchises in isolation/comparison. Tominhawaii joked that I should color code the teams and use their logos, which gave me an idea for how to display team trajectories and patterns. Since the graphs can can cluttered really quickly, I just did three seperate graphs for three fairly familiar franchises: the Blazers, LA Lakers, and LA Clippers (the Minneopolis Lakers and San Diego Clippers are considered separate teams in the databaseBasketball.com data and I did not see any reason to change things). Obviously, I could have done more teams or restricted any team to a particular era, but looking at just these three teams says enough. In all three graphs, I plotted the average age against win%, as in the first graph in this post. It's best to look at the three graphs together.

To lift your spirits, I put the Clippers after the Blazers:

via i35.tinypic.com

via i33.tinypic.com

via i34.tinypic.com

I found myself tracing the history of the Blazers from year to year by connecting the dots in my head. Try it, I think you'll find it enjoyable.

Analytically, these three graphs strongly suggest that if one were to account for "franchise effects" or "GM effects" the age-success relationship would be even stronger. I'll leave the rest of the commentary to you.

Ages of Success in the NBA: How Unusual are the Blazers? (Updated)

PoliSam — Tue, 09 Dec 2008 17:42:52 -0000

One of the major themes of this seasons has been that the Blazers have been unusually successful for their age. I wanted to know, more precisely, just how young the Blazers are and just how successful they have been, given their age. So, I did a little research and found out. The answer? In short, they are one of the younger teams--though far from the youngest--and on pace to be by far the most successful team for its age in the past five years.

Average Age

Most media sources call the Blazers the second youngest team in the NBA, with the Golden State Warriors being the youngest. This description comes from calculating the the average age of each roster in the NBA. As many people have pointed out, however, this is a less than ideal measure of age because some rosters have veterans or rookies that seldom or never play.

A better measure of a team's age, in my view, is the average age of the players on the floor. (To be more precise, define the average age of players on the floor as the expected value of the average age of the team on the floor from a randomly selected moment in the season). Believe it or not, this is relatively easy to calculate. It is just each player's Age * Minutes played divided by the total minutes played by the team.

Here are the average ages of teams calculated in that manner, with the teams currently projected to go to the playoffs by Hollinger in bold:

Team Average Age
MEM    22.8978
POR    24.1397
CHI    24.7834
CHA    24.8325
MIN    25.0516
OKC    25.2086
UTA    25.2347
GSW    25.3229
MIA    25.3542
NJN    25.6025
ATL    25.8435
NYK    25.9501
MIL    26.0521
PHI    26.3214
SAC    26.5633
IND    26.7588
CLE    26.8154
LAL    26.9129
TOR    27.0772
WAS    27.2451
ORL    27.5021
LAC    27.5592
DEN    27.7794
NOH    27.9039
BOS    27.9651
DET    28.2222
HOU    28.3585
DAL    29.2533
PHO    29.5305
SAS    30.6093

As you can see, when calculating the average age of teams in this manner, the Blazers are indeed the second youngest team, but the youngest team is the Memphis Grizzlies (now, whenever you hear an announcer say the Warriors are the youngest team, you can snicker at their ignorance). It's also pretty damn clear that older teams, in general, are much more likely to make it to the playoffs.

Average Age Part 2: An Alternative Measure of Maturity

Now, some might think that even the average age of players on the floor is a less than ideal measure of what we really care about when we talk about the youth of a team--it's maturity as a basketball team. What if, for example, a team's stars are veterans and its younger players are role players? Certainly that is a more mature team than a team whose stars are young and whose veterans are role players, right? Perhaps. The only difficulty is that it's a little tricky to objectively determine who is a role player and who is a star. Nonethelss, in order to investigate if measuring the maturity team in some way dramatically changes the picture, I calculated teams' average age weighted by the number of field goals attempted (each players age*fga/total field goals attempted by the team). Think of this as the average age of the team's field goal attempts:

Team FGA weighted Age:
MEM    22.5476
POR    23.9251
OKC    24.4223
MIN    24.5981
CHI    24.7143
CHA    24.7957
UTA    25.0517
MIA    25.2547
GSW    25.5418
NYK    25.8768
NJN    25.9572
MIL    26.025
PHI    26.096
SAC    26.2507
CLE    26.2931
ATL    26.351
IND    26.6009
TOR    26.8093
WAS    27.1953
LAL    27.2098
DEN    27.2108
ORL    27.4351
LAC    27.6127
NOH    27.9389
HOU    28.1777
DET    28.5663
BOS    28.8435
DAL    28.9353
PHO    29.5283
SAS    29.8134

As you can see, this doesn't change the story too much. The Blazers are a little younger, but still the second youngest. This is primarily because Pryzbilla does not attempt a lot of field goals per minute. One of the teams whose ranking changes the most when calculating team age in this manner is, interestingly, Boston.

Win % by Average Age and FG weighted Age

The next thing I wanted to do was to get a better sense for where the Blazers stood in terms of the success given their age. To do this, I plotted team's win percentage this season against their average age and drew a (non-linear) regression line that shows (roughly) the average win percentage of a team, given their age:

via i37.tinypic.com

The results are interesting. There is a pretty strong trend toward success (defined as win percentage) increasing with age up until a team's age reaches 28 or so. The top three teams in the league (Boston, LAL, and Cleveland) have an average age between 26 and 28, though so do some of the worst teams (Cllppers, and Wizards). The oldest teams, Dallas, Phoenix, and San Antonio, are all winners, but they are not the league's elite.

The Blazers are clearly the best 24 and under team and appear, to my eyes, to be on a trajectory towards joining the league's elite in the next couple of years. Of course, no one can be certain what the future will hold, and making projections into the future is dangerous, but it is easy to see why so many of us are optimistic about the future.

We see roughly the same picture if plot win percentage against teams' age weighted by field goal attempts. The only difference is, perhaps, a shortening of Boston's window for elite play:

via i37.tinypic.com

Power Ranking by Average Age and FGA weighted Age

Since the Blazers have played such a difficult schedule, I also plotted the Sagarin power ratings against their age and FGA weighted age:

via i35.tinypic.com

via i33.tinypic.com

The same general conclusions seem to hold. The Blazers are ahead of the curve and on pace to join the league's elite. These graphs suggests, even more strongly, that there are diminishing returns to age (teams peak at 28 or so). In addition, I was amused to see that the teams that have fired their coaches are all significantly below the age-power ranking curve.

Age and Win % in the Past Five Seasons

Finally, I was curious to see if the age-success relationship that we see this season holds for previous season. And, I wanted to know if there had been any other teams as young as the Blazers that had been as successful. So, I put the data together for the past five season (2004-2005 to 2008-2009.... I could do more, and I might).

In the graph below, I plotted win percentage against team average age. I labeled Portland this season and last, as well as the nearest competitor for the title of "best young team in the last five years":

via i37.tinypic.com

The only team in the past five years that could possible argue to be a better "young" team than the Blazers is the 04-05 Phoenix Suns. The average age of that Suns team was 25.1, a full year older than the Blazers. They were, however, led by the veteran Steve Nash, who was 30 years old by the end of the season. So, I give the title of best young team in the last five years to this years' Blazers.

UPDATE: Age and Win % in the Past 10 Seasons

I had a little more time this afternoon and added five additional seasons. And, because some were interested, I marked which teams eventually won the NBA championship in each season (they are the black diamonds). With five years of additional data, the 2008-2009 Blazers still stick out:

via i33.tinypic.com

Lastly, a question to ponder and discuss: the age success relationship appears to only get stronger as I add more data; why aren't there more older teams in the NBA?

Oden and Portland's Team Defense

PoliSam — Sun, 22 Jun 2008 01:01:52 -0000

If Portland is going to contend for NBA championships some day, it will need to improve its team defense. Although the season is going quite well so far (in my opinion), Porltand is only 20th in the leage in defensive efficiency. It is highly unlikely that the Blazers can win a NBA championship playing that level of defense. Fortunately, there are some very encouraging signs that Oden, even early in his rookie season, is having a positive effect on Portland's team defense. I have commented on this before, but thought I would provdie a little visual evidence to support that claim.

The majority of the discussion of Oden's effect on the team has focussed on the plus-minus statistic or comparison of statistics with Oden in the game and with Oden out of the game. Although there some valuable insights can be gained from looking at those kinds of statistics, they have a couple of drawbacks. One, it is possible that during Oden's minutes he has faced, on average, stronger players or played with weaker teammates. Since there are no readily available statistics to quantify the strength of his opponents or teammmates when he is on the floor, this is a significant problem. Two, having Oden in the line-up can affect how the team plays when he is out in the game as well as when he is on the bench. As is becoming increasingly obvious, Pryzbilla can play much more aggressively when he plays limited minutes because he does not have to worry as much about foul trouble and has more energey. In addition, when Oden plays, Frye is more likely to defend PFs rather than Cs, which is a big improvement.

For those reasons, I think a better way to assess the effect of Oden on Portland's defense is to compare Portland's defensive efficiency in games that Oden played to it's deffensive efficiency in games that he did not play. Since "good" offensive teams will, by definition, score more than "bad" offensive teams, we want to know if teams Porltand has played better defense without or without Oden, "controlling" for the offensive strength of their opponent. If Portland's defensive effiency was better with Oden, but they played weaker teams when Oden played, we wouldn't know much. A rough measure of offensive strength is simply their average offensive efficiency in games this season. So, a simple way to see if Oden might have an effect on team defense is to plot opponent's offensive efficiency when they played Portland against their average offensive efficiency and compare the games that Oden has played to the games that Oden did not play:

via photos-g.ak.fbcdn.net

This graph shows several things. One, as expected, teams that have better average offensive efficiencies tend to score more effieciently against Portland, on average. Two, the relationship between average scoring and scoring against Portland has roughly the same slope whether or not Oden plays. This is shown in the green and orange lines (the "Fitted" lines). The fitted lines show, in essence, Porltand's average defensive efficiency, given an opponent's offensive strength with and without Oden. Three, given an opponents offensive strength, Porltand has played, on average, signicantly better defense with Oden.... In particular, teams have scored between .1 and .13 fewer points per possession against Portland when Oden has played than when he was out with an injury. Though it is not shown in this graph, the effect is so large that passes standard tests of a statistical signficance (in a regression based t-test, the probability that the oden effect is zero is less than 5%).

Now, is this definitive, 100% ironclad evidence that having Oden in the line-up has caused Portland to play better defense. No. Certainly there are other deterimants of an opponents offensive efficiency than simply their average offensive strength. It's possible that, home court advantage, injuries, a "lucky streak" of good defense , or something else can explain this pattern. It is worth noting, however, that controlling for home court advantag actually increases the size of the "Oden effect" and controlling for opponent injuries does not eliminate eliminate it.

That being said, this is really just an early development that will be interesting to track the rest of the season.