As a follow up to my last diary entry(Statistical Analysis of the Defense During B10 Play), I decided to do a comparitive analysis to a team considered an average defense generally speaking. For this particular comparison, I decided to use Penn State's defense as a baseline, based on the fact that they play in the same conference(hence some direct comparison), currently rank #52 nationally in total defense, #6 in B10 in total defense, and #38 nationally in points allowed.

Data - I used the 6 metrics used in the diary mentioned above, and compared UM & PSU on a game by game basis during conference play:

a.) Yards/Drive

b.) Stop Percentage - defined as a drive by the opposition that doesn't yield points

c.) Punt Percentage -

d.) 3 and out Percentage

e.) Points/Drive

f.) Didn't put this on the actual chart, but analysis will follow

Format - The following chart lists the the Iowa & Illinois games first based on the fact that both UM & PSU have played both teams(a direct comparison), followed by the game vs. each other. After that, I decided to do a side by side comparision of the UM & PSU defenses according to quality of competition. Ex - comparison of UM's defensive performance against MSU vs. that of PSU's defense against OSU. It would be ill advised to do a side by side comparison of UM's defensive performance against Purdue vs. that of PSU's defensive performance of OSU, etc. Although this certainly doesn't control for variation in competition quality in an absolute manner, it does somewhat cut down the inflation due to inferior opposition. 

Results - I think many will be very surprised to see the chart below, and certainly the averages for B10 game will shock some in comparison to PSU. It may suggest that the defense really isn't bad at all.

 Looking at the above statistics begs the question - how is it that Michigan's defense allows less yards per drive, forces punts at a higher rate, forces 3-and-outs at a higher rate, and yet still they have a lower stop percentage and allow more points per drive?

Now for the all important metric - FIELD POSITION.

Game Average - I accumulated total yards in terms of field position for each game, and divided that by the total number of drives for each game.

B10 average - Simply adding the averages for each game and dividing by number of games  would be flawed since each game is weighted differently because the number of possessions in each game varies. I totaled the field position for each individual drive(ex starting at your own 20 would yield an input of 20 yards, while starting at the opposition's 20 would yield an input of 80 yards) of each individual game, and divided by the total number of drives for the duration of B10 games.

Here's the Chart:

 UM

 PSU

FACTORS:

Drives per Game - Looking at the stats above, you can see that Michigan's opponents have started 72 drives in 6 games, which means that Michigan's opponents get an avearge of 12 drives per game. On the other hand, PSU's opponents have only had 63 drives for 6 games, an average of 10.5 drives per game. This essentially means that our offense simply scores too quick, and total yards is not a metric for which our defense should be judged by. Since our defense yields an average of 36.83 yards/drive, you would simply multiply 36.83 yards by 1.5, which comes out to 55.24 yards per game that we give up simply because our offense scores quickly. If we used the PSU average of 10.5 drives per game, the defense would allow 378.6 yards per game, which would be good for #64 nationally. For comparison purposes, if PSU opponents averaged 12 drives per game, they would give up 469 yards per game based on their average of 39.11 yards/drive, which would be #120  nationally.

Field Position -  special teams and TO's are the things that are most detrimental to our defense right now. The opponents average starting field position against our defense is basically the 35 yard line. We allow on average 36.83 yards per drive. Using the averages, opponents will get to the Michigan 28 yard line on each possession, and that is FG range for most kickers in the B10. Compare that with PSU's defense that allows 39.11 yards per drive, but their opponents starting field position is around the 25 yard line on average. Using the PSU averages, the opponents average drive would end around the PSU 36 yard line, which is not with FG range for most B10 kickers(53yd FG). This explains why our points/drive and stop%  metrics are lower than PSU's despite the fact that we win almost every other category.

Conclusion:

Simply put, the defense isn't as bad as we currently think. The defense is put in a position to fail nearly every time they take the field. Whether it is the poor field position that we allow, or TO's, you simply can not allow the opposition to have a 65 yard field and expect good results.

Our offense is so explosive that it puts our defense at a disadvantage because it forces the defense to be on the field for more drives than an average team. Second, with the average field position we allow opponents due to Special Teams and TO's, the stats above indicate that we will yield 3 at least 3 points on almost every drive by the opposition.

What Does This Mean?

Our defense certainly isn't good, but it's not as bad as some think. To be honest, I was shocked at how favorably we stacked up against PSU's defensive numbers. Our defense is in fact approaching average, if the appropriate metrics are used. To use an analogy - basketball defensive statistics are often inflated by teams such as Boston simply because they play a slow game, which means the opposition doesn't have as many FG attempts, and hence a lower point total. A more realistic portrayal of a particular defense is FG% and points per possession.

The same can be said for Michigan's defense, we need to use metrics such as yards per drive, and take field position into account as well in order to get a true guage of their performance.