Monday, February 8, 2010

laying to rest a misconception

Dearly beloved, we are gathered here today to put to rest a common misconception about Virginia basketball. It is the halfway point of the ACC season and so it is only fit that we do this today.

As I need not explain to you, Tony Bennett came in as coach this year and made as his #1 priority the installation of the pack-line defense, a strict system that, when properly executed, is godawful painful to try and score upon. Just ask North Carolina. This, plus the demand that his team play a measured, deliberate pace, is the hallmark of the Tony Bennett system. The results on the defensive end have been impressive enough to earn a mention in the NCAA's official blog. Mention, hell: they wrote a whole article detailing the defensive improvement. The spectacular drop from 72.5 PPG allowed to just 60.9 is cited as the reason for the improvement on the standings sheet. They're not the only ones - it's basically conventional wisdom. Improve on defense, improve the record, it sounds simple. Hooray for Tony Bennett and his pack-line.


Wait, what? Yup. The truth is, the team hasn't actually improved that damn much on defense at all. Somewhat, yes. But the main reason we've gone from doormat to an outside shot at the NCAA tournament is because of the huge improvement on offense, not defense. I shall explain.

For those unfamiliar with the KenPom ratings, the fundamental point is this: What's the point of defense? Answer: to stop the other team from scoring on the current possession. Therefore the possession, not the game, should be the base unit of measurement. KenPom's offensive and defensive ratings are based on points per 100 possessions. Very very roughly, the average team gets about a point per possession, so above 100 on defense is bad and below is good. Roughly.

KenPom further adjusts his stats by applying some complex logarithmic exponential differential whatever formula to the raw stats. This is to take away the effects of strength of schedule and try to simulate how the team would do against a perfectly average opponent on a neutral court. We'll be using adjusted stats here.

This year we're averaging about 93.7 points allowed per 100 defensive possessions; last year we averaged 95 points allowed per 100 possessions. There's hardly any difference. Tony Bennett is known for his slow pace, and that is most definitely borne out in KenPom's stats. In 2009, a UVA game averaged 68.2 (adjusted) possessions, 92nd in the country; in 2010, that number is down to about 63 (adjusted) - 322nd in the country. (That's per team per game.) They weren't kidding when they said Bennett slows the game down.

Multiply 93.7 by 0.63 and you get about 59. What do you know - that is just about exactly how many points per game we actually give up. Only off by a point or so. The stats work. Do the same for 2009 - multiply 95 by .682 - and you get 65. That's actually quite a bit fewer than our actual PPG average, but the discrepancy is very easily explained the fact that we average four fewer turnovers than last year. That's four fewer possessions for the opposition, and possessions off turnovers - those caused by steals anyway - tend to be fastbreak opportunities, meaning more points. The four-turnover difference is actually more like an eight-possession swing in our favor: four fewer possessions for the bad guys, and four more chances for us that don't turn into an automatic zero.

But our offense? Improved? Obviously. Again, if you look only at points per game, you'll miss the point entirely. Last year, 70 points per game. This year, 70 points per game. But remember: that's on six fewer possessions per game. Whether you look at KenPom's raw or adjusted numbers, we are scoring ten more points per 100 possessions. We're 43rd in the country in offensive efficiency at 111 points per 100, up from 165th in the country last year.

The numbers don't lie. Our defensive efficiency is about the same as it was. (If you use real-world numbers the improvement is a lot greater, but that doesn't account for the fact that we have yet to play the hardest part of the schedule.) Offensively, we're much better. So what's the real difference?

One, we can shoot. Last year I repeatedly lamented our inability to hit a simple jump shot. This lousy shooting caused a lot of long rebounds and hurt our defensive numbers because teams were scoring easy transition buckets. We're shooting so much better it's unbelievable. Sammy has found a shot, JJ has found a shot, Sylven has found a shot. Such a difference when you can make your three-pointers.

Two, did I mention Sylven Landesberg? He's no longer a freshman and there are two major differences in his game: Turnovers are down and three-point shooting is up. This is big. He's not a deadeye shot from downtown, but he's good enough that people now have to respect it. Good enough to get a little pub for POTY, in fact.

Three, the ACC has been playing out exactly as I told you it would. We are better because we didn't lose any major pieces from last year's team, and the teams we've beaten, with the exception of Georgia Tech, are worse. NC State has taken a step back, Miami has taken about twenty steps back, and UNC has crashed into the goddamned mountain.

If all this sounds like it's taking credit away from Tony Bennett, it's not meant to. He did, as I've harped on, walk into a good situation in that his team is basically the same as last year only a year older, a year wiser, and a year better. And the ACC is, in general, worse. But the turnover numbers are all Bennett. He emphasizes "take care of the ball" more than Leitao did, and his slow pace makes it much easier to do so. Leitao's offense was almost literally indescribable, because nobody was quite sure what it was. (OK, that's not quite fair to Leitao, but the offense was still a mess last year.) Bennett has a philosophy and it results in better shots. Bennett is calmer; Leitao's style was a bit rough on a young team. And Bennett's slow pace keeps the game closer, a plus for a team facing a talent deficiency the way we have been against teams like GT and Wake.

And finally, there's this: Leitao played a pretty traditional man-to-man defense. The guys have been doing that for a while. It's not hard to learn and the players had experience with it. Bennett brought in a whole new thing, which demanded the players scrap their old ways of thinking and do things, in some cases, entirely the opposite of what they'd been taught: do not switch on screens, and such-like. Yes, the defense is only just as effective as it was last year, but to be as effective in a totally brand-new system is an accomplishment in itself - while at the same time improving rather dramatically on offense. Just you wait til this pack-line defense becomes second nature.

1 comment:

Anonymous said...

Excellent post. Best thing I've seen you write! Would love to see the national news pick up the idea.