Talking tempo free

Mr. Rexrode was kind enough to mention us on his blog today. The blurb:

Before going any further, I'd like to appease the TPS Police before they come after me (no, I'm not talking TPS Reports, I'm talking tempo-free stats). I enjoy them, believe in them, still reluctant to use them in stories and spend several grafs explaining them. Here's a glance at them. And as I've said to the folks at TOC and elsewhere, when Izzo stops citing dinosaur stats such as rebounding margin and plain old FG percentage, I'll consider doing the same.

Rexrode is apparently a closeted tempo-free fan. I agree with him that most tempo-free numbers are more complex than what the median college sports fan is looking for. Explaining points per possession or effective field goal percentage in a newspaper column probably doesn't work all that well.

But I do think there's some middle ground for using tempo-free concepts to enhance the experience of the great unwashed masses without unduly confusing them. The two areas that rank highest on my wish list:

• Report two-point shooting separately. Pointing out that Michigan State made 22 of 36 attempts inside the arc on Sunday does a much better job of communicating how well they were able to create good shots close to the rim than saying they shot 52.2% from the field but only 20.0% on three-pointers does. And the former is actually a lot easier to wrap your head around numerically.
  
• Report offensive and defensive rebounding separately. Sunday's game isn't a good example here, since MSU dominated the boards on both ends, but often times a team is very good on the glass on one end of the court but not the other in a given game (offensive and defensive rebounding are distinct skills). Giving specific numbers on one end (e.g., MSU grabbed 28 of 31 defensive rebounding chances) paints a much clearer picture than rebounding margin does.

Rexrode obviously doesn't need any journalism advice from me, a purportedly-retired blogger. (His good deed is not going unpunished, as I pick on him regarding this general topic for the second time in two months.) But I spend a lot of time thinking about the most effective way to present data to people. And I think there's room for improved reporting of basketball statistics that falls well short of dumping PORPAG figures on people. (Another sign of progress and/or shark-jumping: Dick Vitale plugged John Gasaway during the Kentucky-Florida game tonight.)

As for Izzo, my take is that this equation generally holds true:

Tom Izzo's intuition + clunky stats > accurate stats

The rest of us need the best numbers we can get.

And, as Mr. Pomeroy pointed out, Izzo's intuition is fed by an extremely thorough video scouting and analysis process. The kinds of data he gets behind the scenes are probably much more sophisticated than basic four factor numbers.

Plus, in the case of rebounding margin at least, it's not that the number doesn't tell Izzo if his team is playing well. It's just that it doesn't accurately tell him exactly how they're playing well--defensive rebounding vs. field goal defense, etc. Needless to say, he's pretty good at figuring that out himself.

Since we're talking tempo-free numbers. I was struck by how similar most of the four factor numbers were between the two Michigan-Michigan State games:

For MSU, the low three-point percentage on Sunday offset the big increase in two-point shooting. Spartan three-point attempts rimmed out throughout the game, preventing the team from ever pushing lead up to 20. A larger number of free throw attempts certainly helped, although 10 of MSU's 20 attempts from the line came in the final two minutes.

For Michigan, the only significant change was in two-point shooting. Beyond that, the basic shape of the stats was the same in both games: MSU lost the turnover battle but dominated on the boards.

This all goes to highlight the dividends paid from defensive adjustment against Trey Burke. In Ann Arbor, Burke was 5 for 5 on two-point attempts and 3 for 5 on three-point attempts. Sunday, he was only 2 for 5 inside the arc and 2 for 6 outside it--with one of those thee-point makes being of the indefensible variety. Izzo implemented a scheme that forced other guys to beat MSU. Tim Hardaway, Jr. (3 for 9 from the field in Ann Arbor, 1 for 10 on Sunday) was again a non-factor. And Michigan just doesn't have any other players right now who can beat you offensively on a high number of possessions.

Again, Tom Izzo has a pretty good sense of what's driving the final scoreline.