March 1, 2011

Golden Stats, 3/01/11

The Warriors have opened the season's second "half" with three straight losses, the last to a Minnesota Timberwolves team that is comprised of, well, Minnesota Timberwolves. The trade deadline has come and gone, leaving Oakland one young 'un lighter and one blooper heavier. Hope, in its many-splendored forms, has been extinguished from the hearts of all but the most crazed Dubs fans, and there's nothing left to do but take a peek at what's left in this grim cupboard. Line up, Warriors: you're getting inspected, numbers-style.

Last season, we Worriers experimented with a statistical amalgam we called SCORE. We've put that bad boy on ice, partly because BP's WARP, a tentpole metric, has not yet been published for this season, and partly because statistical amalgams are for fucking babies. This year, we're busting out a fat chart. This chart will feature

- John Hollinger's famously flawed and nevertheless useful PER;
- Justin Kubatko's readily available and plentifully plausible Win Shares per 48 minutes;

Without further ado: said chart! Above-average numbers are in green, below-average in red.

An alphabetical approach leads us to the recently re-signed Jeff Adrien! The beefy rook fares pretty well here, but caveats are warranted: the Dubs' improved play in his limited minutes is due to extremely efficient team offense, something that probably doesn't have much to do with Adrien (9.8 points per 36, .470 TS%). Lou Amundson, while almost surely the better player, has fared worse than Adrien thus far, with an even more frightful TS% and free throw shooting that is actually sub-Biedrins. Charlie Bell's good plus-minus showings are due to sample-size noise and his healthier Bucks work of yesteryear, respectively; the production-based stats tell the real story here, and the story is that the dude's done.

Production-based stats do not tell the whole story of Andris Biedrins, who looks useful (albeit less so than he used to) by PER and WP48. The Dubs have been significantly worse on both offense and defense with Goose in the game. We have a soft spot for him, but at this point even we can't deny it: there's something rotten in Latvia. The story is similar for the other longest-tenured Warrior; Monta Ellis has produced laudable individual numbers, but greatly hurt the team's chances in the process. (The glass-half-full take on 'Ta is that he's not hurting the offense, as he did a season ago. He's just hurting the defense... very, very badly.)

We may notice that we skipped someone, and it's because Ol' Babyface has earned his very own paragraph! Stephen Curry, we warmed to you slower than most, but it's official: you are the best player on the Warriors, and one of the roster's two unalloyed positives. Your passing numbers, though still poor, don't stop the offense from soaring when you take the floor, and your defense, though still really poor, isn't bad enough to stop you from being a big asset, 21st-in-the-league-in-APM-as-of-this-writing big. We're still not sure that you can lead a team to the playoffs, but we'd rather see you try it than anyone else in Oakland. Good job, Junior.

Acie Law gets scapegoated for losses more often than he should, but he certainly hasn't done much to demonstrate he's a worthy NBA rotation guy. David Lee's been an offensive asset, as his passing has made up for his weaker scoring... he's been every bit the defensive nightmare we warned about, though, and does not rate as a winnin' dude. Worth eighty million, he ain't. Jeremy Lin is Lee's opposite, both financially and basketball-wise: he's brought it on D (4.7 steals and 1.7 blocks per 36) and looked lost offensively (a .389 TS% that rates among the league's five lowest). His minutes have been too scant for his plus-minus to mean anything, and he's probably in over his head. At this point, though, there'd be no real cost in pleasing the Oracle faithful and giving him Law's backup PG job to close out the year.

Had we been more on the ball in the first half of the season, we'd have devoted a column to the quiet resurgence of Vladimir Radmanovic. The Hairy One has played smart, active basketball, and surely the fact that his contract is expiring is a coincidence. Ekpe Udoh has some of the weirdest plus-minus results in the league this season: his production has been nothing short of unacceptable (7.7 points and 6.4 boards per 36, .485 TS%), and yet both the offense and defense have thrived when he's played. His energy has surely had some effect, but this disparity can't last forever... if he can't boost his numbers, he can't boost the team long-term. Reggie Williams, who looked quite pedestrian when last we checked, has upped his production up to the fairly lofty levels he reached last year. He's done yeoman's work with the second unit, and while his defense still stinks, he's a useful piece. That negative adjusted plus-minus is rising by the week.

You know what's interesting about Dorell Wright? Compare his numbers to those of the average starting NBA small forward.

Creepy, right? We're not saying Dorell is average, though... he scores a tiny bit more often and a tiny bit more efficiently than your average three, and grabs a few more rebounds and a few more dimes and a few more steals. These little things add up to make the guy a solidly above-average small forward.

Adjusted plus-minus is a lot more effusive than that: Dorell currently has the tenth-highest APM in all of basketball. We're frankly not sure where that metric's getting that from. Neither this year's Warriors nor last year's Heat profited all that much from Dorell Wright's presence, and he only played 73 minutes two seasons ago; to the regression-ignorant eye, Dorell doesn't look like any sort of game-changer. We'll need to see more until we believe the guy is that good. He's good, though. Like Curry, he helps the cause.

These Warriors of ours have never seemed wise to the utility of statistical analysis... we can't expect them to get numbers-savvy in a heartbeat. They could help themselves a lot by learning the importance of two simple numbers.

#1 and #30 are good. And if you want to start winning, you should start with them.

No comments: