A closer look at Michigan penalty stats
As there have been several posts in the last couple weeks complaining about refs and asserting that there is a systematic bias against Michigan, I thought I would take a closer look at penalty stats. Specifically, I wanted to go through each opponent and compare their penalty stats against Michigan to their average penalty stats.
The stats come from Team Rankings. They do have the limitation that results against Michigan comprise 1/9th of the data, especially consider Michigan is ranked very low in both penalties against and penalties drawn. This breakdown also doesn't say anything about the legitimacy of penalties called for and against Michigan, or potential missed calls for and against Michiga. It is simply a closer look at opponents' penalty stats against Michigan compared to their average.
---------------------------
Opponent Penalties Per Game Stats:
Opponent | Rank | Penalties/Game | Opponent Penalties vs. Michigan | Difference |
Florida | 126 | 8.4 | 5 | -3.4 |
Cincinnati | 111 | 7.6 | 4 | -3.6 |
Air Force | 5 | 3.8 | 3 | -0.8 |
Purdue | 85 | 6.7 | 10 | 3.3 |
Michigan State | 86 | 6.8 | 11 | 4.2 |
Indiana | 77 | 6.4 | 5 | -1.4 |
Penn State | 19 | 4.6 | 1 | -3.6 |
Rutgers | 64 | 6.0 | 3 | -3 |
Minnesota | 3 | 3.7 | 3 | -0.7 |
- Michigan rank: 102
- Opponent average rank: 64
- Average difference: -1.0 Penalties/Game
Opponent Penalty Yards Per Game Stats:
Opponent | Rank | Penalty Yards/Game | Opponent Penalty Yards vs. Michigan | Difference |
Florida | 88 | 60 | 45 | -15 |
Cincinnati | 114 | 70.6 | 30 | -40.6 |
Air Force | 11 | 35.5 | 29 | -6.5 |
Purdue | 80 | 56.8 | 82 | 25.2 |
Michigan State | 90 | 60.4 | 81 | 20.6 |
Indiana | 95 | 62.4 | 55 | -7.4 |
Penn State | 30 | 44.6 | 10 | -34.6 |
Rutgers | 56 | 50.9 | 20 | -30.9 |
Minnesota | 3 | 30.8 | 10 | -20.8 |
- Michigan rank: 113
- Opponent average rank: 63
- Average difference: -12.2 Penalty Yards/Game
Opponent Penalties Drawn Per Game Stats:
Opponent | Rank | Penalties Drawn/Game | Michigan Penalties | Difference |
Florida | 56 | 6.4 | 7 | 0.6 |
Cincinnati | 16 | 7.6 | 7 | -0.6 |
Air Force | 130 | 3.6 | 7 | 3.4 |
Purdue | 19 | 7.4 | 7 | -0.4 |
Michigan State | 97 | 5.1 | 7 | 1.9 |
Indiana | 13 | 7.8 | 16 | 8.2 |
Penn State | 64 | 6.1 | 6 | -0.1 |
Rutgers | 17 | 7.6 | 3 | -4.6 |
Minnesota | 67 | 6.0 | 9 | 3 |
- Michigan rank: #113
- Opponent average rank: 53
- Average difference: 1.27 Penalties/Game
Michigan Penalty Yards Per Game Stats:
Opponent | Rank | Penalty Yards Drawn/Game | Michigan Penalty Yards | Difference |
Florida | 73 | 50.6 | 55 | 4.4 |
Cincinnati | 6 | 75.5 | 68 | -7.5 |
Air Force | 128 | 33.4 | 72 | 38.6 |
Purdue | 39 | 61.2 | 57 | -4.2 |
Michigan State | 87 | 47.8 | 53 | 5.2 |
Indiana | 14 | 70.2 | 141 | 70.8 |
Penn State | 58 | 56.4 | 59 | 2.6 |
Rutgers | 28 | 64.2 | 25 | -39.2 |
Minnesota | 52 | 58.0 | 85 | 27 |
- Michigan rank: #111
- Opponent average rank: 54
- Average difference: 10.86 Penalty Yards/Game
Some Additional Michigan Penalty Stats:
- Penalties/Play: #123
- Opponents Penalties/Play: #88
---------------------------
- Penalty Yards/Penalty: #71
- Opponent Penalty Yards/Penalty: #107
---------------------------
- Penalty First Downs/Game: #119
- Opponent Penalty First Downs/Game: #100
tl;dr:
- On average, opponents are called for 1 fewer penalties and 12.2 fewer penalty yards against Michigan than their average opponents.
- On average, opponents draw 1.27 more penalties and 10.86 more penalty yards against Michigan than their average opponents.
- Michigan ranks at the bottom at both drawing and committing penalties.
- Overall, Michigan's opponents are squarely average at committing penalties, so that does not explain why Michigan draws so few penalties.
- Overall, Michigan's opponents are slightly above average at drawing penalties, but this may be attributed to Michigan, one of the most penalized teams in the country, contributing 1/9th of the data.
- Michigan's bottom ranking in penalties committed may simply be due to youth and poor execution.
- However, Michigan's bottom ranking in penalties drawn is harder to justify. One could argue that Michigan's anemic passing game decreases opportunities to draw penalties against defenses. But one would expect Michigan's strong defense to draw its fair shair of penalties. Maybe next time, I will look for a correlation between penalties drawn and Defensive S&P to see if that expectation is valid.
November 6th, 2017 at 4:23 AM ^
Great work! I definitely think youth and inexperience are factors but to this extent, doubtful.
November 6th, 2017 at 7:36 AM ^
I think youth and experience ARE heavy factors. The offensive penalties are either false starts or holding by an inexperienced OL. The receivers are being called for holding or illegal blocks and that is a very young group.
On defense, we seem to be getting an unusual number of hands to the face penalties and a large number of offsides calls. This just points to aggressiveness. At a glance, I don't feel outraged at the pass interference calls. We have received a number of phantom calls, but we have gotten away with plays that should have been called.
I am upset at the lack of respect our DL gets. They are held on nearly every play. Rashan Gary has to wonder what he can do to get a call. Maybe he needs to just go to the ground like other teams do to draw a call. Hurst, on the other hand, doesn't let holding stop him - he simply trucks the holder and runs through him anyway.
I long for the day when the game is called consistently and a hold for one team is a hold for the other.
January 7th, 2018 at 1:48 PM ^
onward
November 6th, 2017 at 8:18 AM ^
November 6th, 2017 at 12:30 PM ^
I don't think there is much room for fault in terms of the penalty stats against Michigan. Youth, poor execution, and the eye ball test all verify that most of the calls against Michigan are valid (within normal ref error).
The paucity of penalties drawn by Michigan is the real head scratcher, especially considering the numbers are so far below average for our opponents.
November 6th, 2017 at 1:36 PM ^
That is true. But they did get 11 against MSU, so they have gotten weird spurts.
I looked back a couple of years and this is about half a penalty less called against their opponents than Michigan's historical averaged of around 5.4. So it definitely is a "thing", but with such small sample sizes it's hard to tell if, say, they get 10 against Maryland they'll be right on average. The distribution has been weird, absolutely.
November 6th, 2017 at 9:14 AM ^
November 6th, 2017 at 9:23 AM ^
I just get tired of seeing our defensive line get blatantly held with no flags thrown, and then we get called for marginal chop blocks or ineligible man downfield or whatever penalty wiped out a TD against Florida. If the refs are watching close enough to call obscure or questionable stuff, how do they miss our DL getting held constantly. Last season, Michigan did not have an opponent called for offensive holding during the last 5 games. That is unbelievable.
November 6th, 2017 at 11:48 AM ^
November 6th, 2017 at 9:44 AM ^
Nice job. I wonder what the numbers would look like if you ran the same analysis over several seasons?
I suppose all of this could really be explained away as mere incompetence. But if it's more than that, here is one theory FWIW:
I do a lot of civil rights litigation. With discrimination, there are two key theories: "disparate treatement" and "disparate impact." Disparate treatment is a fancy term for intentional discrimination--it means we are going to treat people less favorably because of their race, sex, ethnicity, etc., and we are going to do it on purpose. That's pretty easy to understand.
Disparate impact is a bit more complicated. With disparate impact, that means you have have a policy that superficially treats everyone the same (i.e., it does not discriminate based on race, sex, ethnicity, religion, etc.). Yet the policy tends to disproportionately harm members of a particular race, sex, etc. To take an extreme example, let's say you have a Ann Arbor company that will only hire employees whose famiies have lived in Washtenaw County for at least three generations. Well, that policy is obviously going to exclude immigrants and the children of immigrants, and may also dispropotionately exclude people belonging to racial or ethnic groups that have historically been underrepresented in Washtenaw County.
A policy that has a disparate impact is not necessarily unlawful--even if it has a discriminatory effect, the person or entity using that policy can potentially justify it (e.g., perhaps the company in question is some kind of Washtenaw County historical society and they show that it's important for its staff to have enduring ties to the community or something). But that part isn't really important in connection with officiating.
If we look at officiating, I really don't believe that Big Ten crews go out onto the field with the intention of screwing Michigan or screwing Harbaugh. But often you see officials adopt certain styles or ways of calling the game that are kind of arbitrary in comparison with the rulebook. On offensive holding, for instance, it does seem that lots of officiating crews tend to swallow the whistle and only call egregious holds (such as where DL are tackled when about to sack QBs), or only call holding on outside plays, or only call holding once early to "send a message" and then swallow the whistle after that. Well, if you are not going to call every instance of holding that you observe, then that policy is not going to affect all teams evenly. It's going to disadvantage teams with faster, more athletiic defensive players--and work to the advantage of teams with stiffer, lower-skiller blockers. So while you may not have set out to purposefully screw Michigan, your policy will nonetheless have the effect--i.e., it will have a disparate impact on Michigan--a team that would likely draw more holding penalties with its fast and well-coached defense than it would likely surrender with its janky OL.
You can look for the same kinds of "policies" at different spots on the field. Pick routes are another area where M has tended to get hosed, as well as OL releasing downfield on passing plays.
One area that seems to have worked out well for the most part for M is on defensive pass intereference. We've all seen officials who call PI at the first sign of contact, but to their credit most Big Ten crews tend to let hand-fighting and incidental contact go so long as the defender is looking back for the ball.
November 6th, 2017 at 9:48 AM ^
November 6th, 2017 at 10:44 AM ^
blatant in fact
I believe a "Harbaugh factor" comes into play with the refs.
November 6th, 2017 at 11:42 AM ^
November 6th, 2017 at 12:26 PM ^
I accidentally used Michigan’s stats for both teams. Fixed it. Numbers are even more stark, now.
November 6th, 2017 at 1:12 PM ^
November 6th, 2017 at 1:20 PM ^
November 6th, 2017 at 1:55 PM ^
November 6th, 2017 at 6:57 PM ^
November 7th, 2017 at 6:13 PM ^
November 9th, 2017 at 11:06 PM ^
Fantastic analysis as usual!
Comments