Thursday, September 25, 2008

Statistically Speaking: Michigan State | by Pat

It Came From the Game Notes

• Notre Dame has not surrendered a single punt return longer than 10 yards or a kickoff return longer than 25 yards.

• Freshman WR Michael Floyd recorded seven receptions last week against Michigan State. It was not only the most catches for an Irish player in a single game since Jeff Samardzija in the 2007 Sugar Bowl against LSU, but also the most ever for a Notre Dame rookie.

• The 16 rushing yards against Michigan State were the fifth-fewest in the Weis era.

• Maurice Crum, Jr. has 260 career tackles, needing just 35 tackles to enter Notre Dame’s top-10 list for career tackles.

Battle for First Down

I'm still feeling my way through this metric. Perhaps it won't really prove very illustrative of anything on its own, or perhaps I just need to be more patient with it. This week's results were not what I expected.

Against the run, ND held 12 of MSU's 20 1st down run attempts (not counting the final kneel down) to 2 yards or fewer. That's good for a win rate of 60%, the highest of the season. That's not exactly what you'd expect given the fact Javon Ringer topped the 200 yard mark and seemingly barreled ahead for a decent gain on nearly every run. And yet there it is. Even more notable is that ND "won" the final 7 straight 1st downs.The fact that MSU had the lead and was running the ball on first down exclusively played into this, but it's still impressive given how much the defense had been on the field by that point.

Perhaps against the run there are only "very good" win rates (over 80%?), or very bad ones, and everything else is sort of meaningless. Last week I wrote that a Win Rate of over 50% against the run was probably very good. I'm now thinking that is way too low for the cutoff of "very good."

Against the pass, ND also notched its highest mark of the season with a Win Rate of 57%. It was a low sample size though as MSU only passed 7 times though on 1st down. The boom or bust effect was definitely seen here as MSU only completed 3 passes on 1st down out of the 7 attempts, but they went for 4, 16, and 28 yards.

Ultimately, I think these numbers might be best used not as a indicator themselves, but as a comparison to the performance of last year's defense or other previous defenses. As some class homework, why don't people take a game box score from last year and figure out the 1st Down Win Rates against the run and pass? It shouldn't take more than a few minutes for each game. Remember, a win is 2 yards or fewer, 3 yards is a draw, 4 or more is a loss. Here are the box scores with the play-by-play data. Just add up the numbers and post your results as a comment. (Hurray for crowd-sourcing!)

Here are the numbers to date this season.

Gimme M.O.E.

This is an interesting week because the results highlight the need to always keep stats and metrics in context. ND's M.O.E. result for the week was 12%. It might seem abnormally low given how poorly the Irish performed, but that's why it's important to remember that this particular metric deals with offensive mistakes rather than efficiency. A 50 yard run and a 1 yard run are the same thing to this particular metric.

The ND offense couldn't get the running game going at all, but for the third straight game kept the self-inflicted mistakes to a relative minimum.

Michigan State on the other hand played their run-heavy, conservative offense to a very low M.O.E. of 6%. Their biggest mistakes were dropped passes. Without those they would have had a nearly flawless offensive game. Most of the quantifiable mistakes used in his metric are tied to the passing game (sacks, interceptions, dropped passes) so it does make sense that teams that rely on the run have lower M.O.E. scores.

Here is the season to-date M.O.E. chart.

Season Long Running Averages.

It's still a bit early to put too much emphasis on these numbers, but we're getting close to the point where extremely low rankings will be hard to salvage for the season. Here are the numbers.