“WINNING AND LOSING WITH YOUR DEFENSE”

I thought it might be interesting to do a companion entry to the diary which I posted earlier in the week, and then expand on that analysis a little bit. Using the same 120 Division I football teams as in the previous diary, I started to do similar sorts of the data and did find some intriguing relationships as well as some intriguing quirks in the relationships.

As in the other diary, I first broke down the team averages on some basic dynamics by win total. Below are the results for basic defensive statistics:

For a point of comparison, the division means for each:

The relationship here is roughly linear, but not precisely linear.

 Looking at averages for total yards, for example, there are two small dips in the trend – one is at 7 and 8 wins, and the other occurs between 3 and 5 wins. One thing that is intriguing about the 7-8 win range is that there is a preponderance of ACC, Big Ten, Big 12 and Pac-12 teams that sit in this range, most notably Michigan, Wisconsin, Michigan St. (bowl stats included here), as well as Arizona, Oklahoma St., Baylor, Clemson and many others. Indeed, three of the top ten defenses and seven of the top twenty defenses by national ranking (yards / game) sit in this area of the overall chart.

The second anomaly – between 3 and 5 wins  - is largely inhabited by teams from lesser discussed conferences, such as the Mountain West, Conference USA, and the bottom tier of teams from the MAC conference. My guess on this one is that it is a quirk of their 2012 schedule and the results of these games, at least when it comes to total yards. In other words, many of these teams seem to be the bottom rung of teams in conferences that don’t have many high-powered offenses. I could be wrong on this, and at some point, I may very break this up into passing vs. rushing to elaborate on this effect.

Next, let’s take a look at top decile and bottom decile performers in the areas of total yards, yards per play, TDs surrendered and average yards per game:

Actually, you can see this in terms of average yards per play right here:

One of the things that I found here that actually did surprise me a little bit was just how significantly less sensitive the relative performances can be to defensive play sometimes. There are indeed teams which have the firepower on offense to outgun their opponents, and when you look at some of the teams that show up as the worst performers on defense, you see many teams that also walked away with decent records for the season as well.

Overall, you still see how much it is worth in terms of wins and losses to have a good defense. The whole idea here was to reach an approximation of that essentially using the otherwise obvious football axioms about having a good defense and what it can get you. It seems that, similar to the diary on offense, we are looking at anywhere from 1-3 wins above the mean, depending on the measure used.

REALLY TL;DR PHOTO: