January 30, 2008

Why UCLA is very lucky to have Kevin Love, part 34,785,900

If any of the following numbers confuse you, it means you probably ought to read this. It's my previous post on the subject. Long story short, I’m using a metric devised by Professor David Berri called Win Score to analyze the LA collegiate basketball teams. If you want to check out my previous post, on USC’s basketball team, go here, please. Thank you. The following post will cover “Your… UCLA BRUINS!” –cue Van Halen’s Right Now-. Anyway. Please, again, keep in mind that all conclusions reached are working within the constraints of the metric used, and should not be considered blanket statements in and of themselves. As to such conclusions, given the design of the statistic in question, nothing was particularly unusual.

Again, let’s just check our metric. 40 Win Score equal approximately one team victory. UCLA have accumulated 805.5 total Win Score to date. 40 from 805.5 is about 20.1, and the Bruins have 18 wins at the time. A little high - in other words, the stat suggests UCLA should actually be doing a little bit better than it actually has (umm... that UCLA should be undefeated PLUS .1 of a game, heh). Interesting. Also interesting - if one used Win Score to set a starting line-up, it would go Collison at point, Shipp at 2-guard, Luc at the 3, Mata at the 4, and Love in the middle. O.J. Mayo would not be in the top 6 (well... don't blame me, according to Win Score he wouldn't, heh). My guess is this is because I feel like Win Score rewards rebounds a bit too highly, and the Bruins as a team rebound extremely well, with Kevin Love leading the way. Speaking of Kevin...

Love is UCLA’s best player by a humongous margin. My apologies to all the other Bruin favorites, but it's true. Much like in Ken Pomeroy's Offensive Rating stat, Kevin Love is one of the 5 best players in the country in Win Score. I was able to check this thanks to the excellent work of Erich Doerr, who showed me some of his numbers he has compiled for the season to date. Thanks a bunch, Erich - that was really helpful. Love is 3rd nationally in Win Score, behind Kenny George of UNC-Asheville, and Michael Beasley of K-State. But... Why?

Well, to start, he scores a large number of points without using very many possessions. He shoots a very high percentage on his 2s, but that's not the only thing he does well. He has brought up his numbers, so that he's a solid 3-point shooter. And he gets to the free throw line with regularity, where he knocks down freebies at an excellent rate for a big man. On top of all this scoring efficiency, he’s one of the best rebounders in the Pac-10, and the country to boot. His other numbers are fairly pedestrian, but he is so excellent around the rim – be it with the ball, or going for a board – that his moderate other values are of less importance. He adds so much value in the things he does well, while not being terrible in the negative categories, that his WS/40 is more than double any of his teammates.

Probably the biggest surprises in this group of stats are the high rankings of Mata-Real and The Prince. Zo and Luc (despite his 3-point follies) are high-percentage shooters that rebound the ball extremely well. If it sounds familiar, it should - nearly all high-value players in Win Score do so. (ed note - There's actually a statistical correction by position to account for the fact that big men usually rebound much better, but I forgot to do so with the USC stats, so we'll be moving forward without it at the moment; if I come back to it later and find that it materially changes the analysis I'll make sure to note it then) Because he's a smart, well-coached player who picks his spots to shoot while going after every rebound Mata-Real scores well in this stat. Furthermore, he's actually one of the better shot-blockers in the Pac-10. Let's correct for playing time, as a theoretical exercise. Assume a linear increase to Zo's stats, relative to his minutes played. Now change the number of minutes Mata-Real plays to the number of minutes Kevin Love has played so far. In that case, Zo would average around 2 blocks a game. That’s around the same level as the much more touted shot-blocking of Taj Gibson.

In comparison to Lorenzo's focus on boards and blocks, Luc's high score shows the value of a well-rounded game. He doesn’t necessarily excel at any one aspect, and in fact has regressed statistically each year he's been at UCLA (more on this in a later post). But he still provides plenty of value to the team in nearly every category. His turnovers are a bit high for a 3/4 on a team with plenty of ball-handlers, but that’s about his only real weakness. Much like Lorenzo's offense, Luc plays defense very intelligently - he avoids fouls, grabs a handful of steals and blocks when he can, and protects the defensive glass.

Looking at Collison, his stats are still depressed somewhat by that gimpy knee of his, in my opinion. I feel that, playing with the brace - and probably still slightly injured to boot - affected his shooting, passing, and defense very significantly (this, too, will come up again in the next post). I actually decided to go back and get the updated stats from after the Oregon State game, which is why the Sunday game stats are included for the Bruins, but not for the SC post below. I did so because Collison's huge game had a monster impact on his score - pushing him much higher than he had been previously. The benefits of getting healthy, I suppose. Even with his struggles until recently, he’s been excellent at avoiding turnovers, averaging one about every 18 minutes. That’s quite, quite good. He's distributing the ball efficiently, is UCLA's best player at getting a steal, and almost never fouls. As such, he rates highly in Win Score despite not being a great rebounder - and I believe his score will continue to climb. I expect his level of play going forward to be closer to that of this past weekend, knock on wood.

Let's continue on down the list. We already covered Mbah a Moute. Shipp is actually a pretty poor rebounder for somebody with his size and athleticism – he’s outrebounded by every major rotational player except Collison, once you correct for playing time. And Collison’s been playing with a gimpy knee. And Shipp still just BARELY outrebounds Darren. Yeah, the point is, he’s probably the worst rebounder on the team. This hurts his individual value. I stress 'individual,' because the team as a whole doesn’t suffer. The other Bruins range from very good to best-in-the-country in this category, so Shipp doesn't really need to focus on getting boards. But here, we’re just looking at his individual contribution to the team at the moment, and as such must then accept his rebounding ‘deficiencies’ as the major reason for his fairly low Win Score. Fortunately, he doesn’t turn the ball over much, and he’s UCLA’s most prolific scorer after Love, so he pulls his weight in other disciplines.

Westbrook has had turnover issues at times, and gets surprisingly high number of fouls for a perimeter player. My guess is his athleticism is both a blessing and a curse, because he sometimes still tries to do too much, which drags him down a bit. Speaking of foul problems, Aboya gets KILLED by his fouls. He averages less than 7 minutes per foul. In case you couldn't guess, that is really, really, REALLY bad. Plus, his blocks are much lower than I expected - I think almost every time he gets a "block" he gets called for a foul as well. Keefe is even worse, at 6 minutes per foul. Luc is the only major rotational player even close to that amount, at over 10 minutes per foul - all other Bruins are over 15.

Chase and Nikola haven't had the opportunity to add much given their limited playing time, but hopefully later in the season they'll get the chance to show what they can do. And this table certainly won't help those who have defended Mike Roll in the past - in 93 minutes, he's amassed a Win Score of zero. Yikes. At least it's not negative!

Next time, I say some things that are really controversial while comparing this season's UCLA team to last year's Final Four group.


Ranjit said...

As a former UCLA Math/Stats major (and enormous sports fan), I eat this stuff up. Thanks for publishing all of this.

Benjamin said...

this is great! Keep up the good work. I appreciate your recognition of the fact that good rebounders have an edge with win score but I also think that guards are at a disadvantage. Do you know the thinking behind why assists are halved? Also, what about defensive/hustle statistics that programs like UCLA keep and that include deflections, knocking the ball away from a dribbler, charges taken, etc. It seems that although the WS appears to be right on in many cases, we saw how valuable Luc's role within the team defensive scheme is when he wasn't able to play in at the end of the SC game and in Oregon. Even when he isn't recording steals or blocks the guy is a tremendous asset. Also, remember how Texas exploited Love's inability to rotate quicly enough within the defensive schemes? Can this be accounted for? What about the way DC pressures the opposing team's point guard disrupting the team's offensive sets?

Brett said...

Nice post, I'm a big fan of Berri's work and enjoy seeing it applied to some college hoops. In response to the earlier comments, read Dave Berri's The Wages of Wins for a thorough explanation of the metric. If you liked this post you'd probably enjoy the book. And Benjamin, it would be nice to take into account every single thing that happens on the court, but it would also be pretty cumbersome to do so. Win Score is meant to be a good way to value players that is easy to compute.

Walton's Wisdom said...

This is a great post. I guess I was bound to like it given my profession as a CPA and my love for the game of basketball. Oh, I also hail from the state of Oregon and am one of the few people in this state that aren't bitter about Love flocking south (who could blame him?!).

I am not surprised in the least bit by the findings of your analysis, and was shocked/apalled by UCLA's turning away from Love for good portions of its final two tourney games. Even Billy Packer, possibly the most dense announcer in the history of sports, wouldn't stop commenting on this during the games.

See my post on Love as an NBA prospect, although it involves more feeling/opinion than it does numbers: http://waltonswisdom.blogspot.com/2008/03/kevin-love-as-nba-prospect.html