I don't do real work during Michigan week. I peruse box scores and advanced stats and try to get an angle on the game. I figured I'd share it here.

Common Opponents

Ohio State and Michigan have four common opponents this year -- Michigan State, Oregon, Indiana, and Northwestern. Some general notes:

• With conference expansion and no divisions, there are less common opponents than usual. There used to be at least 6, but now there are just 4.
• Interestingly, the teams had opposite home fields for each of these matchups (e.g., Ohio State was at Oregon, Michigan was at home; Michigan was at Indiana, Ohio State was at home, etc.). 
• Ohio State was 3-1 in these matchups with a total score of 138-61. Average score of 34.5 to 15.25.
• Michigan was 2-2 in these matchups with a total score of 106-81. Average score of 26.5 to 20.25.

Common Opponent Box Score Notes

I took a look at the yards per play in each of these games. I think this is the best metric we have in today's modern game given the different paces of games.

• Offensively, Ohio State averaged 8.7 yards per pass and 4.8 yards per rush against these four teams -- a pretty solid output given that two of these four teams are in the top five.
• Michigan averaged 6.0 yards per pass and 3.8 yards per rush. Anything more than a yard per play difference is huge. The rushing number is obviously positive for Ohio State, but that difference in pass production is an astronomical difference when it comes to a yards per play basis.
• Defensively, Ohio State gave up 6.9 yards per pass and 2.7 yards per rush. Michigan gave up 7.3 yards per pass and 2.7 yards per rush.
• Again, their defense is pretty darn good. Their offense puts them in a lot of bad spots but their front four are ferocious and keep them in games.

Michigan on the Road

Before looking at the advanced stats, it is worth noting that in the above games, Michigan was the home team in 3 of the 4 games, while Ohio State was the road team in 3 of the 4 games. One would assume this helps tilt the comparative play-by-play numbers in Michigan's favor.

• Michigan has been awful on the road this year. They are 0-3 (L @ Washington by 10, L @ Illinois by 14, L @ Indiana by 5). Three things are import from these games.
  • First, Michigan's offense is averaging 13 points per game on the road. That is obviously not even in the galaxy to be able to beat Ohio State in Columbus. 
  • Second is that Michigan averaged 2 turnovers per game in these three contests. Do that Saturday and all of these numbers don't mean anything -- OSU will blow their doors off.
  • Third, Michigan's offense started slow in all three of these road contests. 
    • UM 1st Half Drives at Washington: Punt, Punt, Punt, Touchdown, FG
    • UM 1st Half Drives at Illinois: Punt, Punt, Fumble, Fumble, TD
    • UM 1st Half Drives at Indiana: FG, Punt, Punt, Fumble, Punt
      ​

So they are averaging 5.66 points per first half. Do that Saturday and it's probably over at halftime.

The Advanced Numbers

The darker the green the more elite you are. The more purple you are the more terrible you are. Note that Ohio State is dark green in essentially everything.

EPA = expected points added, which generally means offensively you are moving towards scoring (and thus increasing the probability you will score, and thus increasing your expected points), while defensively it means that you are preventing the other team from scoring (and thus decreasing their expected points).

• Of note is that Ohio State is awesome in all of these metrics on both offense and defense.
• Offensively, Michigan is god awful in passing EPA and mediocre in rushing. Somewhat surprisingly (given their DL) is they are pretty average at rushing defense and passing defense.
• The most interesting thing to me is that Michigan's EPA on early downs is absolutely awful. Coming into last year's game they were 22nd in early downs EPA on offense, this year they are 117th. This means they are getting crushed on first down on offense, thus putting them behind the chains, and then they don't have the QB play or the weapons to get out of it.
• Overall Michigan last year was #2 in net EPA per drive, meaning they were elite at adding expected points on offense and preventing expected points on defense on a drive-to-drive basis. This year they are 79th. With a defense as good as theirs, this reflects how bad their offense is more than anything.

Success Rate = focuses more on down and distance, and essentially measures whether you are staying ahead of the chains on offense, or keeping the other team behind the chains on defense.

• Michigan was absolutely elite in these categories on both sides of the ball from 2021-2023. This was the trademark of the Harbaugh era. For reference, coming into last year's game, Michigan was #11 in offensive success rate and #15 in defensive success rate, meaning that they were pretty darn good at keeping opponents behind the sticks.
• This year Michigan is #57 in offensive success rate and #31 in defensive success rate. Those are pretty mediocre numbers.

My summary of the advanced stats:

• Michigan's offense is not good on early downs and thus not very successful. Stuff them on first down and they will have a very long day.
• Michigan still has a very good defense. Not as elite as last year but they are just a touch behind Oregon and Penn State in my opinion. 
• It's interesting that Michigan's net field position is so bad (87th nationally). 

Conclusion -- Keys to the Game

1. Offense: Michigan's front four is arguably one of the three or four best in college football. I'd like to see us play fast to make them rotate out and then ram it right at them. In the passing game, Will Johnson is out, so attack his backup. Most important thing is no catastrophic plays. Make their offense beat us.
    
2. Defense: Stuff them into a blender on first down and make them one dimensional. They don't have the pass protection or receivers to get them out of these jams on a consistent basis.
    
3. Maybe Nervous?: There's 25% of me that thinks maybe Michigan is like Florida, Kansas, or Oklahoma -- a team with a lot of talent that took way too long to figure it out, but now that they have, they are dangerous. In the last four games they beat Michigan State at home, played better than I expected against Oregon, stayed in the game at Indiana, and then beat the doors off Northwestern. They are certainly playing their best complimentary football of the season right now. They should never have benched Warren, and now that he's getting reps, he looks like your typical Michigan QB. 
    
4. Game Management. Given the above we have to knock this confidence out of them early in the game. They are averaging just 5.66 points per game in the first half. Hit them in the mouth and get a couple touchdowns on offense. But, Saturday has the potential to be a slow start and then a Penn State-like grind, which is what Michigan wants. Michigan will script a decent opening drive and will try to limit possessions in the first half. If Ohio State can jump on them the game has a chance to be over at halftime. Hopefully we come out fast. But if not, stay calm and do what we've done all year -- dominate the third quarter and put them to bed.