February 12, 2010

Pair Essentials

Statman Jon Nichols has retired from his excellent blog, presumably for the greener pastures of some franchise or another. He emitted a pretty fun last burst of data for NBA fans to remember him by: how effective teams are with various pairs of their players on the floor. We shall sift through the Warriors' data in our usual fashion, and so's we don't drown, we'll restrict things to the eight guys that matter the most: Biedrins, Morrow, Randolph, CJ, Maggette, Monta, Ronny and Curry. (Nichols posted these charts on January 25th, so these numbers represent each pair's results through the Dubs' first 42 games.)
















First thing to note: Biedrins's results are pretty wacky across the board, and given the small number of minutes he'd played with some of the guys, it's probably not worth taking his numbers seriously. The Warriors are not actually doomed to offensive failure when Biedrins takes the court with Randolph or CJ; they performed well enough when he played with those guys last year.

The most interesting results here involve the divergent patterns in the backcourt. The pairings of Curry with the Warriors' best weapons -- Morrow, Maggette, Randolph, Biedrins --lead to good things for the Warriors' offense. The pairings of Monta with those guys do not... of the six other guys, only Turiaf shows equal results with Monta as he does with Curry. And worrisomely, Monta is Curry's least effective offensive pairing.
















Once again, Biedrins brings the wacky... it is not likely, for instance, that the pairing of Biedrins and Turiaf actually leads to the worst defense in NBA history. (Indeed, Biedrins's results have already normalized a bit since Nichols posted these numbers.) In a similar vein, while it's nice to think that the pairing of Randolph and Turiaf leads to shutdown D, the sample size there is too small to tell us much of value. More than anything, this chart serves as a useful snapshot of each player's overall effect on the defense. Turiaf and CJ have been defensive assets; Curry and Biedrins have been defensive detriments; the other four guys rate somewhere in between.













Of the 28 pairings of these players, only six have seen the Warriors play net-effective basketball on their watch: Randolph-Turiaf (+7), CJ-Turiaf (+3), Curry-Turiaf (+2), Monta-Turiaf (+1), Biedrins-Curry (+1) and Randolph-CJ (+1). Conversely, four of the five worst pairings involve Biedrins. Turiaf posted better results than Biedrins last year as well, and it seems safe to say that Ronny fares better in the Warriors' free-wheeling than Andris, but the true gap between the two is surely not as extreme as this chart suggests.

Once again, the striking results here involve Monta and Curry... Curry goes better with every other guy except for CJ, the guy who's least likely to be around next year. Monta rates as one of the least effective pairings for Randolph, Morrow, Maggette and Curry, and as an above-average pairing only for Andris, whose numbers are all over the map.

These numbers don't suggest that the Warriors have to trade Monta Ellis. But they do represent a stark reminder that what he's doing isn't working: as things stand, he does not marry well with any of the team's other young building blocks. The Warriors need to find a way to make him an effective team player again, and if Don Nelson isn't up to the task, they need to find a coach who is.

No comments: