blog | rpi | pomeroy ratings | stats

Monday, January 08, 2007

Game Plan

[Doing some housekeeping from yesterday’s post - it was a major oversight to mention great freshmen point guards and not think of Mike Conley first. Duh.]

As promised, there’s a new feature on the site today. It’s a fine mix of poorly formatted tables and preformatted text!

Just like just about everything on this site, this feature is built on the ideas of Dean Oliver. Please buy his book.

I’m calling it Game Plan, although that’s a bit of a misnomer. The first part is a sortable list of the four factors for each game played so far. Remember that I have to do some estimates on team rebounds, so these numbers are close approximations (except free throw rate, which is exact) of the calculations you’d get by using the official box score where team rebounds are broken down into offensive and defensive. Also, I’m calculating free throw rate (FTR) using FTA/FGA for both offense and defense.

The second part of the page is more powerful, but definitely not for everyone. Correlations are calculated for each parameter against offensive and defensive efficiency. Relationships significant to the 95% level are in bold and to 99% with a red star. For just about every team, you’ll see that shooting is significant to offensive efficiency, and shooting defense is significant to defensive efficiency. That’s not exactly front page news. But for other teams you’ll see different things.

Let’s use Texas Tech as an example of how I think this feature should be used.

Texas Tech’s defensive efficiency has a fairly strong connection to the turnovers they force. For example, sort their games by defensive TO%, and you see that their worst defensive game was against Air Force and that was the game where they forced the fewest turnovers, also. Air Force just doesn’t turn the ball over regardless of who they play.

Looking ahead, using the four factors offense page and sorting by conference and by turnover percentage, you can see that in the Big XII, Texas is ranked 5th nationally in turnover percentage. Might we see Texas have an unusually good offensive night against Texas Tech when they meet in Lubbock on January 31st? Perhaps. A couple reasons why this might not work out - Texas doesn’t shoot the ball well, and their Game Plan page says turnovers aren’t important to their offense. Ideally, I’d like to see mirrored connections with the two teams.

The idea is that you can figure out what the keys to an upcoming game should be. When somebody says “x” quality is really important to a team, now we can check if that’s actually been true. If you see a situation that looks better than this example, let me know, and I’ll post it on the blog. Maybe this has a valid application, and maybe not. I’ll have more to say about it on the usual Friday posting after I play around with it more and get some feedback from you all.

Posted on 01/08 at 11:00 AM
Permalink • E-mail me

Powered by ExpressionEngine