Hokepoints: Law & Order in the Big Ten

Submitted by Seth on July 16th, 2013 at 5:04 PM

REPAIR NOTICE: I originally posted this article earlier this morning but accidentally had some bad data from a dreaded bad sort on Excel. Things should be better now, and the conclusions were affected less than I thought they would be. Biggest change is Ohio State was credited with a few that belonged to Oregon State (an avoidable vlookup error), and the old home-road stats were all screwed up. They are fixed below.



I've been slowly building and picking through an all-plays database built from NCAA.org's play-by-play data. The easiest thing to pull out so far has been penalties, so let's play with those.

The benefit of the all-plays is you can tell the difference between penalties, since a personal foul says a different thing about a team and does a different thing to them than, say, a delay of game to set up a punt. I broke the various penalties up into "Violent" and "Non-Violent" behaviors.

  • Acts of violence: Clipping, crackbacks, facemasks, illegal blocks, illegal use of hands, kick catch interference, pass interference (?), roughing the kicker (15), roughing the passer, tripping, and unnecessary roughness.
  • Non-violent behaviors: Delay of game, encroachment, false starts, holding, ineligible receiver downfield, intentional grounding, kickoff out of bounds, offsides, running into the kicker (5), sideline interference, substitution infraction, too many men, unsportsmanlike conduct, and illegal fair catch, formation, forward pass, motion, participation, shifting, and touching.
    Michigan last year was remarkably good at avoiding the latter type (in yellow in the chart below), leading the study at 2.3 non-violent infractions per game:



That's the Big Ten and the other 2013 opponents. I don't know if I want to count PI since its application can get downright chintzy, so that's broken out. Either way Ohio State managed to lead the conference in infractions per game, and was second in the study only to Terry Bowden's one-win (Morgan State) first season at Akron. Reason why this is? Online poll says:


Fact: 4.5% of people who take any online fan poll are Buckeyes

Yea, and Urban did steal "60 minutes of unnecessary roughness," previously committed to MSU. I was surprised that Michigan State appeared to have their pugilistic streak in relative check, i.e. they were only among the leaders, not far ahead as I supposed from watching them. It takes a while to gather all the data but minus the regular season Wisconsin game (data wasn't available) their 2011 penalty numbers were high but their personal foul quotient wasn't: 31 violent (11 of those pass interference) to 60 non-violent. Wanna guess where a disproportionate of those came from? Offsides. #JerelWorthyJumpsEarly.

Michigan vs. Average

We're dealing with smallish sample sizes so conclusions are shaky. That said there are things to see when you look at which penalties Michigan was getting called against them versus a typical team on their schedule.

Non-violent things per season:

Penalty Avg Tm* Mich OSU ND MSU'12 MSU'11
False Start 16.5 7 20 21 14 20
Holding 14.4 13 12 11 15 12
Offsides/Encroachment 7.3 6 4 4 8 20
Illegal Offensive Stuff 6.0 - 4 5 5  
Delay of Game 4.3 - 2 3 1 4
Coach Derps 2.7 3 3 2 2 2
Intentional grounding 0.8 - 1 - 1 -
Unsportsmanlike Conduct 0.8 1 3 - 1 1
Special Teams Derps 0.5 - 2 1 - -
TOTAL 53.3 30 51 47 47 59

* over13 games

Michigan's veteran offensive line was good for something last year: remarkably few false starts and none of those illegal formation/procedure things that plagued us in various offensive transitions. That's a feather in Al Borges's cap: the offense had their fundamentals down about as well as you can ask. Pre-snap penalty-avoidance may be correlated with offensive line experience, though I haven't proven this. Further study: is it experienced OL or just experienced tackles? Inquiring 2013 offensive lines want to know.

Violent crimes per season:

Penalty Avg Tm* Mich OSU ND MSU MSU'11
Personal Foul 15.0 11 22 9 21 16
Pass Interference 9.2 9 8 6 8 11
Various Illegal Blocks 5.8 8 7 4 6 -
Facemask 1.8 2 4 3 1 1
Roughing the Passer 1.3 3 2 1 - 3
Kick Catching Interference 0.6 2 - 1 2 -
Roughing the Kicker 0.3 - - - 2 -
Unnecessary Roughness 0.0 - - 1 - -
TOTAL 33.8 35 43 25 40 31

* over13 games

Michigan's ability to avoid the peaceful infractions meant the Wolverines were the most pugilistic in the study by percentage of penalties that were violent. Cue the Urban Meyer chart:


Forgot to add the 15 yards for logo infraction

Really the Wolverines were average, the only thing standing out being chopblocks. There were a few of these called against Michigan last year that I thought were horsecrap (Mealer's v. UMass and Gallon's vs. Minnesota), and here's one that was legit (on Gordon):

If you don't spot it in 10 watches, watch it 10 more times.

I'm declaring Michigan a very average team at this.

Home Field Advantage?

There was one for Michigan, not the other guys. Michigan was relatively clean at home and in limited samples got kinda duked in the neutral games (Brian gave the refs a composite –5 for the Alabama game alone, which is about the difference between a typical day of Obi Ezeh as a senior versus Kenny Demens as a senior). Overall I noticed very little difference in any type of penalty with regards to how it was assessed against home versus road teams. False starts are a little more common for road teams (like one every 10 games) but that's about it. Things broke out a bit more among the small samples of a single team's season:


Team Pen/G Home Away Neutral Home Field Adv.
Akron 8.2 6.7 9.7 - -45%
Ohio State 7.8 8.6 6.3 - +28%
Nebraska 7.1 6.7 7.8 7.0 -16%
Purdue 6.8 5.4 8.2 9.0 -51%
Michigan State 6.7 6.3 7.0 8.0 -11%
Minnesota 6.6 7.0 5.8 8.0 +17%
Illinois 6.2 5.9 6.6 - -13%
Indiana 6.0 6.7 5.3 - +20%
Northwestern 5.8 5.4 7.0 2.0 -29%
Notre Dame 5.5 6.4 6.0 1.5   +7%
Connecticut 5.4 5.0 5.8 - -17%
Penn State 5.3 5.0 5.6 - -12%
Michigan 5.0 4.3 5.2 6.5 -20%
Iowa 5.0 4.9 5.5 4.0 -13%
Wisconsin 4.6 4.4 5.2 4.0 -17%
Central Michigan 4.6 4.3 4.8 6.0 -12%
Other 6.6 7.2 6.2 6.8 +13%
AVERAGE 6.2 6.1 6.3 6.2   -3%
Undebated: the paucity of calls per game that went against the Wolverines in the Big House. Debatable: whether that's because Hoke's crew are well-behaved gentlemen who happen to get screwed when they go on the road, or because even referees' psyches don't' do well against 111,000 critics. Interesting that being on the road actually helped some teams , particularly Ohio State. When I broke it down with the violent/non-violent stuff Michigan held steady as the best team in the study at avoiding the little things (they were slightly better on the road than at home) but a fairly dramatic difference in how the big things were assessed.


Team Pen/G Home Away Neutral HFA
Ohio State 4.3 4.6 3.5 - +24%
Michigan State 3.6 3.1 4.4 3.0 -40%
Michigan 2.3 2.5 2.2 2.0 +12%
Avg Team 3.8 3.7 4.0 3.2   -7%


Team Pen/G Home Away Neutral HFA
Ohio State 3.6 4.0 2.8 - 31%
Michigan State 3.1 3.1 2.6 5.0 17%
Michigan 2.7 1.8 3.0 4.5 -64%
Avg Team 2.4 2.4 2.3 3.0 3%

Either they let the Wolverines get away with murder at home, we turn into Michigan State on the road, or those calls just went against us more often than they should have.


Everyone Murders

July 16th, 2013 at 6:10 PM ^

I was surprised that Michigan State appeared to have their pugilistic streak in relative check, i.e. they were only among the leaders, not far ahead as I supposed from watching them.

Your perception may be skewed because they play particularly  dirty with abandon on the games we pay particular attention to.  By which I mean, when they play Michigan (at least under Narduzzi's watch and, IMO, during the Perles years).

Interesting article - thanks.


July 16th, 2013 at 6:08 PM ^

To answer the last question, my guess is that with so few penalties overall, a couple of iffy PI's can wreak havoc with the percentages.

As for why MSU wasn't as violent as expected, the reason is probably that the rivalry games bring out their worst, but when you are playing Indiana or EMU you don't need the same level of aggression.


July 16th, 2013 at 6:46 PM ^


Really the Wolverines were average, the only thing standing out being chopblocks [sic].

Hey! That's not even...

I mean...


I'm really sorry, guys :-(

Undefeated dre…

July 16th, 2013 at 8:04 PM ^

Seth, great stuff. Are you able to look at number of penalties per total plays (ideally broken down by offense, defense, special teams)? Would also be curious of penalty yardage per unit / yards allowed/gained per unit. May not impact the home/road splits but could lead to a re-rank of penalty proclivity within the B1G.


July 16th, 2013 at 8:48 PM ^

Regarding the VLOOKUP error, I highly recommend INDEX MATCH instead of VLOOKUP.  INDEX MATCH is much more flexible (which might have helped with the error) and slightly quicker than VLOOKUP.  It's syntax isn't as intuitive, but it's worth learning. 


July 17th, 2013 at 7:53 AM ^

I love excel and have not heard of it [Index Match functions]. I'll migrate out of vlookup / hlookup for the index match formulas--way more functional.

Thank you!


July 17th, 2013 at 9:14 AM ^

on the charts Seth, as always...I think that Michigna got treated pretty well by the zebras, but at the same token I think it was fair because we were/are a disciplined team unlike O$U and little brother (most of the time, at least)


July 17th, 2013 at 12:33 PM ^

By all indications, OSU had 84 penalties last year, or 7.0 per game.  Did you add in some Oklahoma State penalties in there as well?

Also, Michigan had 62 penalties or 4.77 per game.  


July 17th, 2013 at 1:23 PM ^

I never have a problem with roughness penatlies as I am of the opinion that it is a side effect of good aggresive defenses.  It is understood that tactically any unneccesary roughness call is a judgement mistake.  You just hope you can minimize it.  Roughness calls are worse when they are occuring because of sloppiness or loss of cool. However, if some are generated because someone is attempting to make a play as long as you have multiple sacks, TFL, and TO's generated it is something that is worth the risk.

In the case of MSU I would say the roughness penalties are worth the final results.