David’s Stat Corner: EM in the Big 12

This year I’m going to be doing weekly Big 12 rankings based entirely off of efficiency margin. Efficiency margin, for those who aren’t nerds, is a team’s points per possession scored minus the teams points per possession allowed. In essence, a measure of how much better or worse a team has been compared to their competition on a per possession basis.

It’s easy (and fairly effective) to look at teams efficiency margin in conference play, and simply rank them that way. However, I wanted to go a step further, so I’ll be comparing their actual efficiency margin in each game versus their opponent’s efficiency margin, as a means of comparing how they’ve performed based on what we’ve come to expect out of the competition.

To illustrate, let’s look at KU’s schedule so far. They’ve played TCU, K-State and Texas Tech in conference play. TCU’s EM in conference play so far is sitting at -7.1. In that game, KU’s efficiency margin was +8. So KU’s “score” in that came for the purposes of these rankings was .9, and TCU’s was -.9. K-State’s Big 12 EM is +5.9, and in their game against KU it was -3, so KU “scores” 8.9 points in that one. Texas Tech’s EM sits at -11.6, and against Kansas Saturday it was -26, giving the Jayhawks a 14.4. You add these together, and KU’s rating through three games is a 24.2.

I’m going to call this measure Marginal Performance Rating (MPR) for now, because I’m not good at naming things. If you have a more creative name for this, I’m definitely all ears. Obviously, with a 3 game sample size, this is going to be flawed. But, since the teams all play each other twice throughout conference play, it should smooth itself out as the season goes on. The idea is to simply measure how a team has performed vs how a generic team would be expected to perform based on the opponent’s results to date. I’ve also thrown in the team’s offensive efficiency vs the opponent’s defensive efficiency and vice versa, as a means of measuring teams on each side of the ball. Here’s where the Big 12 sits through three games (all stats used are derived from KenPom)

Again, three games is too small a sample size to draw anything too definitive, but you can see that this is a pretty telling method of quantifying the teams’ results so far. West Virginia’s performance against a Tech team that isn’t looking too great is clearly pulling them down, while Iowa State’s defensive performances are really propping them up. Oklahoma is really bad, Kansas State has been just ok, and wow, KU’s defense...that’s another topic unto itself, as is their offense.

As always, leave any thoughts or ideas in the comments!