Michigan Offense and Defense Against the Averages

Submitted by Bambi on

With our recent success, there has been a lot of skepticism from opposing fan bases about how good our team actually is. While they admit we have been impressive, they say that we haven't played any good teams yet and cite our opponent's offensive and defensive stats as proof. I wanted to look into this further, so that's what this diary is.

To start, I wanted to put together a table showing the national ranks of the opponents we've played in some key overall stats.

The total defense/offense stats are in YPG and the scoring stats are in PPG

Team Total Defense National Rank Scoring Defense National Rank Total Offense National Rank Scoring Offense National Rank
Utah 392 66 19.8 33 401.8 70 37 25
Oregon State 393.6 67 29.8 89 320.4 119 19.8 116
UNLV 421.5 96 26.8 73 382.5 86 28.3 74
BYU 400.7 78 26.3 71 408.2 63 27.7 79
Maryland 457.7 110 34.7 103 341.3 109 24.3 98
Northwestern 269.5 9 12.2 8 353.8 98 21.2 110

This next table shows these same stats for our opponents with the totals from each team's Michigan game removed.

Team Total Defense National Rank Scoring Defense National Rank Total Offense National Rank Scoring Offense National Rank
Utah 399.4 77 20.4 38 414.8 55 39.6 14
Oregon State 391.3 66 28.8 84 356.9 98 22.4 106
UNLV 430.4 102 26.6 72 412 59 32.6 44
BYU 391.2 66 25.4 65 468.8 23 33.2 43
Maryland 473.6 111 36 111 388.6 79 29.2 69
Northwestern 246.4 3 7.64 3 391 77 25.4 90

Read this table as follows: If the national rank of a team decreased with Michigan's game removed, than Michigan underperformed against this team in that stat. The opposite goes for a national rank increasing.

For example: Utah's Total Defense rank fell from 66 to 77 without the Michigan game stats in there. This means that Michigan's offense had less than 392 yards of offense against Utah. The same logic follows for all stats.

The third and final table shows the differences between Michigan's totals against these teams and their averages with the Michigan game removed.

Team Total Defense Scoring Defense Total Offense Scoring Offense
Utah -44.4 -3.4 77.8 15.6
Oregon State 13.7 6.2 218.9 15.4
UNLV -53.4 1.4 177 25.6
BYU 56.8 5.6 363.8 33.2
Maryland -95.6 -8 283.6 29.2
Northwestern 133.6 30.36 223 25.4
Average 1.78 5.36 224 24.1

Read this table as follows: If there's a negative sign in front of the number, it means Michigan performed that much below average against a team in that categorie. The same logic follows for the positive numbers.

For example: In the Utah game, Michigan has a -44.4 for total defense. That means Michigan gained 44.4 yards less than Utah's defense gives up a game on average.

The average total shows how much Michigan out performs or under performs in a categorie each game. So according to this table, Michigan will gain 1.78 yards more a game than the opposing defense normally gives up. So if a team normally gives up 300 yards a game, you can expect Michigan to gain 301.78 yards.

Similarly, Michigan will score 5.36 more points a game than the opposing defense allows, allow 224 yards less than what the opposing team normally gains, and allow 24.1 points less than what the opposing team normally scores.

If I had more time/motivation, I could put more time into the analysis of these stats. Instead of doing a raw average total in the last table, I would much rather see the percent difference between Michigan yards gained and average yards allowed by a team, etc. So there is some more analysis that can be done and some stuff that's left to be desired, but it does paint the general picture pretty well.

The reason I did this is because I think this paints a better picture of how good a team is. If a team averages 500 YPG that looks great in terms of stats, but to me a team that averages 400 YPG while playing defenses that allow 300 YPG is a better offense than one that averages 500 YPG while playing teams that allow 475 YPG. An average offense isn't one that is ranked 65/130 in YPG, but one that gains as many yards against a team as that team normally lets up.

So that being said, the results in the third table above show pretty much what we expected. Michigan has an average to slightly above average offense. Michigan will gain as many yards as their opponent normally allows. The scoring stats are higher than you might expect as we score almost 5.5 more points a game than our opponents allow, but some of that comes from our non offensive scores against NW. Similarly, because of our great special teams and defense we generally have good field position, increasing our point output while decreasing our yard output.

Defensively however, we are great. Teams gain 224 less yards agains us than they normally average, and 24.1 points less per a game. That's incredible. This is a great defense, arguably the best in the country, and the stats back it up in any scenario.

Comments

leftrare

October 13th, 2015 at 10:54 PM ^

Well done. I've been poking around at stats trying to see this but don't have nearly as much dedication as you. Specifically with regard to how opponents have done when not playing Michigan. You nailed it. OTOH, S&P pretty much bears out, I think, how good Michigan is: #3 in the country. So, in some ways this analysis is only about confirming what we know from a stat guy that's worked even harder at it on every team in the country. Edit: sorry, Bambi, that sounds Dickish. I really do appreciate what you've done here.

Silver Fox

October 14th, 2015 at 4:25 AM ^

I originally read this as our offense and defense against the Avengers. I know I wouldn't mind seeing Lewis pick off Hawkeyes pass to Ironman, while the rest of the D shuts down the Hulk up the middle. In my head I picture Smith bulldozing over Thor, but that is about all I see from the offense.  I don't think there is much else they could do going up against Captain America at safety and assuming the Hulk is a 2 way player and is also in as the entire defensive line.

bcnihao

October 14th, 2015 at 7:20 AM ^

The stats confirm what our eyes tell us:  UM's peformance has been average on offense and great on defense. 

An observation re. the opponent's "scoring defense" outlier for the Northwestern game:  NW's "scoring defense" was affected by the TDs that were scored by UM's defense and kickoff return unit.  Taking that into account, UM's offense can be credited with scoring 16.36 points more than the average NW's defense had previously given up (if I'm reading the third table correctly), which is still a respectable overperformance against the average of NW's defense. 

(UM's "scoring defense" comparisons are similarly victimized by Utah's pick-6 but aided by NW's fumble return for TD against Minnesota.)

Another wrinkle affecting the stats would be the injury status of the opponent, such as Utah's quarterback.  But good job, OP.