The Forgotten Man: Are the Spurs Better Off?
I previously made some comments within the recap thread on the loss to Denver pointing out that the Spurs seem to win at an extraordinarily high rate with DeJuan Blair in the starting line-up, and that perhaps he should have started against Denver. I noted that since 2011-12, with Blair starting (most persistently with TP & TD), the Spurs have gone a gaudy 103-29 (comprising 75% of all Spurs games up to the present), for a win percentage of .780. And during the 44 games in which DeJuan did not start during this stretch, the Spurs have gone a solid but less than stellar 28-16, for a win percentage of .636. Other commentators on that thread suggested that it's more than likely that DJB was just a coincident indicator in this success because his on-court/off-court numbers (among other factors) show that the team actually performs better overall when he is off the court.
So this brings us to today's post, which attempts to broach the same subject matter, but from a slightly different angle. Until couple of games late last season (against the Jazz, and Lakers), DeJuan Blair, as the 37th pick in the 2009-10 draft, had played in virtually every Spurs game since he donned the Silver & Black at the outset of his rookie season. Prior to that day at home against Utah (4/8/12), I'd be hard-pressed to think of one game in the previous 214 where he received DNP-CD. And yet now, during the 2012-13 campaign, his fourth with the team, Blair has received 6 DNP-CD's in the first 28 games of the season.
So I thought that it would make sense to explore the following question: Does DeJuan Blair help the Spurs win games when he plays an active, meaningful role, or would the Spurs be better off using him sparingly?
I'll just focus on a few general areas, like DJB's basic production (points & rebounds) during games in which he played zero to 16 minutes, and those in which he played 22 minutes or more, and what the Spurs win-loss record has been during each of those series of games.
I'll start off with providing some aggregate data that covers the period from the beginning of 2010-11 (when DJB began to start more regularly), and extends up until the Spurs most recent win against the New Orleans Hornets.
Since 2010-11, DeJuan Blair has played in 166 Spurs games, and has started in 132 of them (80% of the time he played, he started). DeJuan's minutes logged in these games has run the gamut from a high of 41:35 (in a win against NYK in 2010-11), to low of 3:36 (in a loss against DEN in 2011-12).
I'd now like to focus a bit on the 27% of his games (45) where he was used more sparingly in bits and pieces (defined as playing 16 minutes or less), as well as the 42% of games (70) in which he played more of a significant role (defined as playing 22 minutes or more).
DJB games ≤ 16 minutes 2010-11 2011-12 2012-13 Total
Points/Rebounds per game 4.1/3.6 5.3/3.0 3.7/3.4 4.4/3.4
Points/Rebounds per 36 11.8/10.4 15.3/8.6 10.7/9.8 13.2/10.2
Spurs W/L (win %) 16-5 (.705) 11-4 (.733) 5-4 (.556) 32-13 (.711)
DJB games ≥ 22 minutes 2010-11 2011-12 2012-13 Total
Points/Rebounds per game 11.9/9.4 13.3/7.4 14.7/8.3 12.6/8.5
Points/Rebounds per 36 15.9/12.5 18.1/10.0 21.2/12.0 17.3/11.7
Spurs W/L (win %) 29-8 (.784) 23-7 (.767) 3-0 (1.00) 55-15 (.786)
In each season above the Spurs have won at a significantly higher clip with DJB playing a significant role, versus him playing a relatively insignificant one. In aggregate, the Spurs were 32-13 when DeJuan played (whether starting or not) 16 minutes or less, and 55-15 when he played for 22 minutes or more.
Let's take a moment to look at DJB's basic production, in particular comparing his per-36 numbers in the "small minute games" to his per-36 numbers in the "larger minute games". This simple amount of data allows us to make some tentative observations and educated guesses, and also ask some pertinent questions. First of all, DJB at least seems to help the team win at a high rate (.786) in the games in which he plays 22+ minutes. Certainly averaging 12.6 and 8.5 (17.3/11.7 per-36) in 26 mpg is some significant production from a 4/5 guy playing next to Tim Duncan, even if he had some drawbacks, like a higher than optimum turnover-to-steal rate, a higher than average foul rate, or below average defensive numbers, as three possible examples.
Looking at the 16 minute and below numbers (12 minute avg.), which often involved 2-3 short stints, his production per-36 was a solid step below his twenty-two-minute and above, per-36 numbers. However, the biggest difference was in scoring, which is not really the main reason DJB is out there. It's more about his rebounding, which while lower (10.2 vs. 11.7), was still very solid overall.
Without examining all the data in excruciating detail, I would guess that DJB's overall game was weaker in his 16 minute and below games, particularly in his efficiency numbers (e.g., FG% ...). That said, the team managed to lose at a much higher rate with DeJuan receiving modest to insignificant minutes, when compared to those games in which he played a more significant role. As a result, these basic facts raise a host of interesting questions.
Is DeJuan Blair just an inconsistent player, and will he always be, regardless of how he is developed? Were his numbers and play so bad during his low-minute games that it was prudent to keep his play to a minimum in those games? Or, was DJB playing well enough in many of those low-minute games that he should have been allowed to play through the rough patches? Is it inconceivable to suspect that the Spurs may have actually ended up with a higher win percentage in those low-minute games if DJB was allowed to play through some of those "slow-start" games? And is the hook on DJB generally too quick, and could the Spurs be missing out on improved play later in the game?
This brings me to the final set of data. This data will involve a considerably smaller sample of games (28), but I'd like to revisit the following fact presented near the outset of this post:
... during the 2012-13 campaign, his fourth with the team, Blair has received 6 DNP-CD's in the first 28 games of the season.
Which begs the question: is it a good idea to play Blair zero to 16 minutes in over 56% of Spurs games for which he was available (i.e., not injured) this season? (DJB has had 9 games with 16 minutes and under in playing time this season).
Keep in mind some of the pertinent background: the Spurs win at a considerably higher rate since the beginning of the 2010-11 season when Blair is a starter than when Blair is not a starter (.780 vs. .636). The Spurs also win at a much higher clip when DJB plays 22 minutes or more than when he plays 16 minutes or less (.786 vs. .711). As well, keep in mind that the Spurs are 12-3 this season with Blair as a starter (.800), and 8-4 when he does not start (.667). Okay, so let's look at the 15 games in which Blair did not play at all when available (i.e., not injured), and logged 16 minutes or less.
During those 15 games, the Spurs have managed to post a record of 10-5 (.667). In their losses, their opponents combined W/L record is 66-37 (.641). In their wins, their opponents combined W/L record is 65-63 (.508). In the games in which he played but was held to 16 minutes or less the Spurs were 5-4 (.556). In the games for which he was available, but did not play at all, the Spurs were 5-1 (.833). However, these 5 wins came against opponents with a combined record of 49-80 (.380). The lone loss was against a team with a record of 14-13 (.519). All-told, 6 of the 10 wins out of those 15 games (those where DJB played anywhere from zero to 16 minutes) were against opponents with a combined W/L record of 54-90 (.375), the other 4 wins were against teams with a combined W/L record of 60-43 (.583), with the average margin of Spurs victory just 3.5 points. So, at least on the surface, it appears that the Spurs when going completely without Blair win at a high rate (.833) with solid margins against the poorer teams (.380), and generally speaking when marginalizing Blair (playing him from zero to 16 minutes) exhibit a similar pattern by achieving solid wins against poorer teams, and low-margin wins against the better competition.
Now again, this most recent portion of the analysis is done on an insufficient sample size. However, since the data from this season only tends to basically coincide with the larger sample of data covering all games since 2010-11, the potential implications of this data cannot be dismissed out of hand. In addition, the fact that Kawhi Leonard and Stephen Jackson does not seem like a sufficient excuse to me. For example, when KL became a starter during DJB's last 23 regular season games last year, the Spurs went 20-3. Thus, it does not appear that a return of KL would reduce the Spurs fortunes if DJB was in a 16+ mpg.-role. Furthermore, the Spurs were 11-2 with Blair starting after Leonard went down, and 9-2 after Jackson went down.
The foregoing analysis is simply an inductive form of reasoning whereby if the premise is true, the conclusion may also be true. I'm quite aware that this exercise does not prove anything, but in my view it reasonably raises some important questions about the diminished playing time of DeJuan Blair, and how this could impact the Spurs going forward. The margin for winning in this league is super tight, and this team simply must optimize every angle possible. Also, this analysis is not about whether DJB's on-court/off-court numbers are better than Diaw's, Splitter's, or Bonner's (it's not that those factors are irrelevant either). It simply raises the question, are the Spurs more likely to win at a higher rate with DJB occupying a 16+ mpg.-role on this team? Is it possible that putting DJB in a particular role (such as starting the halves, for example) somehow allows the rest of the team as a whole to be put in the best position to come out with a "W"? How does one adequately explain away the win-percentage-effect in favor of Blair starting over the last 176 games? Is Splitter really that much better now, particularly against the playoff bound teams, to fully offset the extreme reduction in DJB's playing time? Is Diaw just coasting through the first 28 games, and so he and Splitter are perfectly capable of sucking up DJB's minutes when it matters the most? Do we plan to go small-ball more now that Jackson's here for the whole year, thus negating the need for a 4th big?
As far as I can see, there are a lot of questions, and not enough really solid answers that can be compellingly and/or conclusively documented. Nevertheless, it would be interesting to hear some meaningful debate on this subject. The decisions made by PATFO regarding the questions raised in this post (and other questions, of course) could very well decide whether the Spurs make a Finals appearance this season for the first time in 6 years.