I think both Dudley and Gortat's play was hurt by playing with Beasley… I don’t have the stats, so I might be mistaken, but it would be nice to see eg. a +/- breakdown for both Dudley and Gortat with and without Beasley on the floor. (ilintar)
It certainly seems that no one plays well when Beasley is on the floor. (Seth Pollack)
This post was inspired by the two comments shown above posted under "Suns wing Jared Dudley thriving in a starting role," a recent story by Jacob Padilla. Coincidentally, I had just completed a relevant analysis (covering the first 25 games). I myself had long suspected that Beasley was hmm … incompatible with the rest of the team; as a scientist I needed a solid proof though.
The +/- stats are a useful, but not very accurate measure of player's performance; other advanced statistics paint a more complete picture of a player, but even those render an image of a player in isolation. As anyone working on experimental data modeling or, say, process optimization knows, analysis of correlations frequently reveals hidden dependencies and occasionally alters the big picture. While correlating advanced statistics would perhaps be more insightful, I believe that looking at correlated +/- values sampled over 25 games is adequate to understand the Beasley effect on other players.
Using +/- data from nba.com I assembled a table for the regular Suns rotation showing for each player +/- per 36 minutes for the total minutes as well as for minutes played with and without Beasley, Dudley, or Gortat. The remaining four players (Garret, Marshall, Zeller, and Johnson) played insufficient minutes for the correlated results to be statistically meaningful.
The Beasley, Dudley, and Gortat effect is also presented below graphically, where +/- per 36 minutes with the player of interest on/off court is shown by red/blue circles, respectively, and the weighted averages (weighted by minutes played with or without the player of interest) are shown by dotted lines.
Every single regular rotation player had a negative +/- value over the first 25 games for minutes played with Beasley on court, while they all had positive +/- when Beasley sat on the bench. The weighted +/- average for the team with Beasley on and off the court (weighted by each player minutes) was -9.9 and 7.0, respectively. Therefore, the Beasley effect, or perhaps more appropriately, the Beasley hole, was on average nearly -17 (per 36 minutes). Clearly, all rotation players (not only Dudley and Gortat) were hurt by Beasley’s on-court presence with O’Neal, Scola, and Telfair suffering most.
Given that the Suns had lost several games by 3-4 points (or in OT), it seems to me that sitting Beasley (or, even better, not signing him in the first place) would had perhaps made the Suns a winning team. So, what now, should Beasley be traded, moved to the end of the bench or even send to the D-League? Well, his game so far look certainly appealing to any tanking team, but other than that it seems unlikely that other teams would trade for him. Since quite a few people still believe in Beasley’s high ceiling with Gentry being, in my opinion, one of them (given that Gentry’s job is on the line, Beasley’s eventual emergence might be Gentry’s best hope), playing Beasley limited minutes, exploring match-ups, and developing him is what will likely happen.
Probably the most surprising result of this analysis is the huge positive effect of Dudley’s on-court presence on +/- of any other Suns player. The average +/- over the first 25 games for minutes played with Dudley on the floor was higher by 17 as compared with that with Dudley on the bench. Dudley seems to make any other player on the Suns roster to perform, or at least to look, better. He may not be the franchise player, he may not be a superstar, but no one should doubt that he is a legitimate starter on this team. I believe that Dudley and Gortat are between the most unappreciated NBA players, but this could be perhaps a subject of another posting.
For Gortat, the on/off court effect was relatively small on average and possibly even statistically insignificant, although it is just my guess not supported by any rigorous error analysis. I should mention that for majority of the rotation players (except for Beasley and Dudley, of course), the average on/off court effect was small. Gortat clearly made Beasley a better player (or, again, a better looking player). In fact, when playing with Gortat, Beasley’s performance (and by performance I mean his +/-) was not much below the team average. The same was also true for Beasley playing with Dragic. Maybe this is why Gentry played Beasley with Dragic (64% time) and Gortat (63% time) longer than with any other player. It is also apparent that Scola, Dragic and Dudley played better (their +/- was better, which may not necessarily be the same) without Gortat. Scola and Gortat’s playing space somewhat overlap, so it may be the reason. As for Dudley, it would be interesting to see if his performance with Gortat improved after he was moved to 3.
The case of Dragic and Gortat pair is an interesting one. For most other pairs there is a clear beneficiary of a pairing. However, both Gortat and Dragic seemed to play better with the other player on the bench. While they were -6.0 when playing together, Gortat was 8.9 without Dragic and Dragic was 13.6 without Gortat. If things do not change, this may pose in the future a big problem for the Suns, maybe even more serious than the Beasley effect (hmm…, nah), as Gortat and Dragic play most of their time together (nearly 84% of their on-court time over the first 25 games). The Dragic/Gortat pairing should certainly be explored in more detail. Well, maybe the next time. Now, I just hint that the Dragic/Gortat pair played roughly 50% of their time with Beasley and his mighty suckiness might have been a factor. But then again, eyeballing their game together (e.g. PnR) tells me there might be something else there.
Worth of noticing is also synergy of Telfair/Gortat pairing. They were +8.6 playing together and both were in negative +/- with the other player on the bench. On a side note, with a huge positive effect of Dragic on O’Neal’s play (+/- on/off court difference of +20), maybe Dragic, Gortat, Telfair, and O’Neal’s pairing time should be somewhat tweaked. It should not be disregarded though that the Dragic/Gortat pair played a significant fraction of their time against opposing team starters, while the Telfair/Gortat pair against the opposing team bench.
The correlated +/- analysis is a useful tool, but one should keep in mind that it has limitations. Beasley and Dudley are clear-cut cases. Statistically speaking, Beasley has been a serious on-court liability for the Suns, although, obviously, not the only Suns problem. Also statistically speaking, Dudley has been a legitimate Suns starter and perhaps MVP of this team so far. As for Gortat and the rest of the roster, things are not so clear and more complex analysis is necessary to uncover the whole truth.