If you don’t run, you can’t hide

The world of smart basketball got a huge boost entering last season when Ken Pomeroy quit his job as a meteorologist. This allowed him to focus full time on his blog, and in the past year he's added several new features. He's added coaching records, box scores, conference-only stats, and others. And now – thankfully – he's added the stat that I've been begging for.

Basketball is all about tempo. While Len Elmore and the rest of the old-school dinosaurs keep spinning rebound-margin and points per game as key indicators, they conveniently ignore the fact that different teams play at different paces. BYU scores a ton of points (77.1 per game to be exact)? Shocking, considering they run the break on virtually every possession. Miami doesn't score a lot (69.7)? They must win with their grinding defense. Never mind that on a possession by possession basis, Miami scores more than BYU.

To follow tempo, you have to follow Pomeroy's blog. Because he doesn't just track the number of possessions in each game, he corrects those data based on what a team's opponent is trying to do. A BYU vs Miami game might not feature as many possessions as BYU wants, but with some simple statistical tricks, Pomeroy is able to compare BYU vs every other Miami opponent and adjust accordingly.

The problem is that tempo has been lumped into one single number. Until now.

Pomeroy's newest addition is 'average possession length.' This is critical, because offensive tempo and defensive tempo are treated independently. One of the examples he uses in his latest post is Syracuse. Overall, Syracuse had the 245th ranked tempo in the nation, meaning that 244 of 347 D1 teams average more possessions per game (ie, are faster) than Syracuse. So Syracuse must be a slow team, right?

Wrong.

Evaluating their offense and defense separately, it turns out that Syracuse forces opponents to use a ton of time when Syracuse is on defense, and sitting in their famous zone. They're hard to get shots up against. They keep teams from running the fast break. All told, out of 347 D1 teams, Syracuse plays defense more than all but two teams. When 'Cuse gets the ball, they're actually pretty fast (41st fastest tempo in the nation).

This is important because all teams like to sell recruits on how much they like to get out and run the floor on offense. But now that data is public. Talk is cheap.

Using last year's ACC as an example, here are the offesive tempo rankings for the past four years (as far back as the data currently goes) for each of the ACC coaches. Some coaches overlap with previous jobs, and where possible, that data is included.

  2010 2011 2012 2013 avg
Roy Williams 6 9 3 14 8
Mark Gottfried     60 33 47
Leonard Hamilton 38 31 41 128 60
James Johnson       87 87
Jeff Bzdelik 121 37 191 80 107
Mike Krzyzewski 216 38 90 133 119
Brian Gregory 86 51 259 96 123
Mark Turgeon 207 274 157 85 181
Brad Brownell 262 167 131 262 206
Jim Larranaga 215 187 284 305 248
Steve Donahue 205 258 293 265 255
Tony Bennett 306 335 282 314 309

It's no surprise that Roy Williams is at the top or Tony Bennett is at the bottom. UNC wants to run on every possession, and Virginia longs for the days of no shot clock. But there are some surprises in the middle. Coach K has only run one offense faster than 90th in the past four years. while Leonard Hamilton has had three that were between 31st and 41st. For FSU, those years coincided with their elite defensive years, and so when you tuned into a game you likely heard from someone like Len Elmore how much FSU liked to grind it out. Meanwhile, Leonard Hamilton and staff were talking about running. Turns out that Ham was right, and it wasn't just coach speak. For Coach K, there's a ton of fluctuation in his numbers, which is also consistent with what he preaches. He might be the best coach in basketball at bending his system to fit the players. And the proof is right there.

Further refinements in the statistic could help smooth out some of the outliers, like teams who offensive rebounding is on the outer margins of the bell curve. But for now, even in its raw form, this is a very powerful stat.

It's also one that is going to be tossed around on the recruiting trail. Sellers beware.

Quantcast