## 2009: A Look Back at Offensive and Defensive Efficacy

Last week, I offered up a statistical teaser based on the 2009 Big Ten schedule, for what I'll be putting together for all of college football in 2010. It looked at 9 team strength ratings that were based on a variety of calculations. The basis for all of them is tempo and competition adjustments. In other words, how effective was your team's offense or defense per possession (how many yards or points did they score per possession, factoring out end of half knees) and how did they do in each particular game against that opponent's average offensive and defensive ratings per possession.

Here are the 2009 offensive and defensive performance ratings for Big Ten teams, only in the 8 games of Big Ten play (as I mentioned last week, these formulae require closed systems for maximum accuracy and the 2009 Big Ten season was the alpha-test for all D-1 games this season):

The Will Rating is the combination of the two statistics, Offensive Performance and Defensive Performance. Essentially, it is showing the delta of the two as a measure of a team's strength where 1.00 equals an average football team. It is essentially offensive production minus defensive production where the delta should give some estimate as to the team's strength. To look at what we are actually subtracting, we have to look at the individual offensive and defensive data...more after the jump:

Above is the 2009 Big Ten Offensive Efficacy rating. It is calculated by determining the yards per possession the offense gained, i.e. the ability to move the ball and then matched against the opponents' defensive average yards given up per possession, yielding a percentage of how much better one offense was able to move the ball against each defense than the rest of the Big Ten. Each game performance is then averaged over the course of the entire Big Ten schedule to determine the final rating.

For example, let's take a look at the Penn State-Minnesota game, where all statistics would show that Penn State dominated all facets of the game, but was unable to get the points to match, in a 20-0 win. PSU averaged 51.556 yards per offensive possession, which equaled a factor of 1.51 or 151% of what every other Big Ten team averaged per possession vs. Minnesota's defense. This was then averaged against every other offensive output for all Big Ten teams, which is what appears above.

By this metric, it shows that Wisconsin (1.17) and Penn State (1.13) were the two strongest teams in terms of the ability to sustain long drives. Iowa (.87) and Michigan (.90) were the worst offenses in the Big Ten.

For a second, you might ask, "well, why did Iowa win so much and Michigan lose so much?" Here is one of your answers. The other is found in the STIC rating I mentioned last week, which measured ability to manufacture points through special teams, defense and short fields. Here you see that Illinois (no surprise) and Indiana (equally of no surprise) were the two worst defenses in Big Ten games in terms of yardage given up per drive as compared to the opponent's ability to move the ball. Iowa (.77) and Ohio State (.75) were astoundingly good here. In other words, expect your typical drives to be shortened by 25% vs. those defenses.

Where this gets interesting is in some of the statistics in the table that don't show up in these graphs but help to explain the more difficult questions. For example, "what's up with Wisconsin?! They look like the strongest team according to this measure." This requires us to 1) look back at the graph from last week, which showed they had the greatest range in terms of just how strong they were:

This tells me that Purdue and Wisconsin were the two most schizophrenic teams in the Big Ten. Like a box of chocolates, you never knew what to expect playing them. Wisconsin also rated fairly poorly in terms of the STIC rating. I believe this stemmed from the amount of points they gave up in relation to yards their defense gave up. They gave up a lot of free or at least easily earned points.

On the other hand, the defensive measure above, shows Penn State not to be nearly as dominant as we thought when looking at yearly defensive measures such as yards per game. The simple answer is that we (along with Minnesota) averaged the least amount of possessions per game. The opponents didn't have many chances to score against PSU, but that doesn't mean they couldn't move the ball against Penn State. They did, but we also had the second lowest points given up per possession (behind Ohio State, which dominated every defensive statistic in my table). That is bend but don't break defense to a T.

TheJuggerNittAugust 17th, 2010 - 23:14

Your yards per drive statistics…perhaps another one that might be a bit more telling (at least for overall team performance) could be something along the lines of “yards from the endzone when the drive stalled”. Say you are averaging 50 yards per drive, but every drive is starting on your own 10. That would look better than only averaging 30 yards per drive, but you were starting at midfield (or even on your opponents 30).

adminAugust 18th, 2010 - 20:25

The challenge with that, is that a 30-yard drive starting from your opp 30, resulting in a TD is far better than a 30-yard drive from your own 10. If starting from own 10, the thinking is that by this metric, a good offense would average 40+ yards (or whatever is towards the higher end) and that would bear out in this particular measure. Now on the other hand, like you said a 50-yard drive that stalls and gets 0 points means nothing, which is why we balance it with the STIC (special teams, intangibles, coaching) which measures ability to generate points through short fields, special teams and defensive touchdowns. One might at first think those two statistics would counteract each other, but OSU ended up being very highly rated in both. Wisconsin was good at moving the ball and stopping the other team (to some extent) but not a top tier team when it came to the STIC measure or generating points. Iowa wasn’t great at moving the ball, but was very good at stealing free points (high STIC). PSU was fairly solid in all of the above except in two games. You can guess the two.