What is this diary? Each week, I take a deep dive into the box scores and team pages from Pro Football Focus, Game on Paper, College Football Data, and more for as many Big Ten teams as I can manage. Often, life makes it impossible to cover every team, but I try to always cover OSU, MSU, PSU, and whoever Michigan is playing in the coming week.

I try to generate a couple blurbs based on the statistics I find compelling from each game/team. Really, there’s no formula. This is just a very paper-based, statistical approach to evaluating teams. And, also, sometimes I watch a little bit of film if I don’t believe the statistics and want to know what the heck is happening (see: Drew Allar).

Note. I’ve been super busy recently at work (I’m a middle school teacher and have had a lot of parent-teacher conferences) and so I’m way behind on keeping up with this thing. Finally managed to throw this together. I’m hoping to put together something tonight on PSU prior to Saturday’s matchup, but we’ll see what happens.

(All images courtesy of MGoBlog's Flickr page; all images taken by Patrick Barron.)

MICHIGAN defeated PURDUE (41-13)

The leaders and the best.

• The gap between Michigan’s overall PFF grade (97.0) and the second best-graded team (Air Force, 94.4) is greater than the gap between that second best-graded team and the 16th best-graded team (Ole Miss, 91.5).
• Per College Football Data, Michigan’s net PPA (predicted points added, similar to EPA) is 0.556. There is only one other team in the whole country with a net PPA greater than 0.4 (Oregon, 0.413).
• Per BCF Toys, the gap between Michigan’s net yards per play and the second best team’s (Washington) net yards per play (1.66 net YPP) is equal to the gap between the second best team in net YPP and the 33rd best team (Penn State) in net YPP (1.66).

No problem on third downs.

Michigan converted 7-of-14 third downs in this game. On the season, they rank third in third down conversion rate at 55.6% (behind Georgia and Oregon). They held Purdue to one first down on their 14 attempts, too. The defense is nearly as stiff as the offense is productive, coming at 10th on the year with a third down stop rate of 29.4%. Only Georgia has a better net third down conversion/stop rate than Michigan.

Almost best at everything.

Through Week 10, J.J. McCarthy has the second best offensive grade on PFF amongst QBs with 100+ drop backs (92.5, behind LSU’s Jayden Daniels [93.0]), the second best passing grade (91.4, behind Oregon’s Bo Nix [91.6])*, the second best completion percentage (75.7%, behind Nix [77.9%]), the third best yards per attempt (10.4, behind Daniels [11.1] and Bama’s Jalen Milroe [10.5]), the second best big-time throw percentage (9.1, behind Milroe [9.2]), the fourth best adjusted completion percentage (81.5%, behind Nix [85.1%], UAB’s Jacob Zeno [84.0%], and Florida’s Graham Mertz(!) [81.8%]), the second best NFL passer rating (131.4, behind Daniels [136.4]), and the second best QBR (92.3, behind Daniels [93.1]). I am somewhat okay if J.J. finishes second in the Heisman to Daniels. To anyone else at this point would be utterly ridiculous, and if he puts together two strong statlines against PSU and OSU (think ~500 passing yards, 70% completions, 8.5 Y/A, four TDs, one INT, a handful of rushing yards) and wins both of those games, it would be utterly ridiculous if anyone but him won.

*After initially writing this, PFF updated their grades: Nix dropped to 91.4, and so he and J.J. are now tied for first in passing grade.

Trouble in the run game.

In Brian’s most recent Unverified Voracity, he included an image of a graph from CFB Data that showed adjusted offensive line yards and adjusted rushing highlight yards. The graph indicated that Michigan was very good at gaining consistent and positive yardage (see: line yards), but completely mediocre at breaking big runs (see: highlight yards). Explosiveness* is also a good metric CFB Data has that determines exactly how successful a team’s successful offensive plays are. Take a look at this graph:

It shows Michigan pretty far to the right (high line yards at 3.46) and pretty far to the bottom (low explosiveness at 0.86). Contrast this to last year when Michigan had 3.15 offensive line yards per run and a 1.03 explosiveness score on runs. The simplest takeaway from this is that the line is better this year and the running backs are worse. (Key caveat: Michigan has yet to play the toughest defenses on their schedule this year, and they also played a few other very tough defenses last year by this point in the season [see: Iowa, Illinois].) I still have yet to do my Alternate Equation Line Yards post, but at this point I’m pretty sold: Blake Corum 2023 is absolutely not Blake Corum 2022, and the numbers back that up.

*Explosiveness is the mean EPA on successful (positive EPA) plays.

More receiver runs, please.

The Wolverines ran two plays against Purdue that handed the ball off to a wide receiver: Semaj Morgan picked up 44 yards on his carry, and Cornelius Johnson picked up 14 on his. It’s the tiniest of sample sizes, but they’ve been incredibly successful on receiver runs this year: 5 carries, 98 yards, 1 TD. I imagine this works so well because teams are cramping themselves into the box to stop Michigan’s running back run game, and leaving themselves vulnerable to trickery on the edges. They should do this more, since it seems to work so darn well.

Which receiver would you rather have?

Let’s play a game. Wide Receiver Alpha catches 57.8% of the balls thrown his way, drops 8.8% of them, catches 37.5% of the contested balls, gets a first down on 41.1% of his targets, and gets a touchdown on 11/1% of his targets. Meanwhile, Wide Receiver Beta catches 76.6% of the balls thrown his way, drops 2.7% of them, catches 60.0% of the contested balls, gets a first down on 59.6% of his targets, and scores a touchdown on 21.3% of his targets. You’d choose Beta in a heartbeat, right? Well, Alpha is Marvin Harrison Jr. and Beta is Roman Wilson. To be fair, Harrison’s volume is nearly double what Wilson’s is (90 targets to 47 targets), and I’d wager he probably (definitely?) sees more defensive attention, too, but on a per-target basis, it’s pretty clear which receiver you’d rather throw at.

Dominating the trenches.

Michigan has the highest net line yards in the country at +1.19. Penn State is second at +1.13, and no other team is over +1.0. They are fourth in net power success rate at +0.337, bested by only Oregon (+0.357), UCLA (+0.346), and Oklahoma (+0.345). They also have the second best net stuff rate in the country at +0.149. Penn State is first at +0.151, and no other team is better than +0.120. No, those are not adjusted stats, and, yes, I wish they were, but, yes, that does mean that they are doing a very, very good job with the beefy boys. Amongst the Power 69* teams, they also have the fourth best run blocking grade (behind Oregon State, Wake Forest, and Alabama), the third best run defense grade (behind Clemson and Iowa), and the second best pass rushing grade (behind UCLA).

Wrapping them up.

Against Purdue, Michigan missed just two tackles (Mike Sainristl missed one and Amorion Walker missed the other). This brought their tackling grade on PFF to number one in the country at 92.0. In truth, they’re probably an even better tackling team than the numbers make them out to be, since the players who miss tackles with the most frequency are those who were spelling the injured secondary to start the season. Quentin Johnson has been the team’s worst tackler with a missed tackle rate of 22.2%. DJ Waller Jr. has missed two tackles this year, and has a missed tackle rate of 16.7%. Only one player I’d truly call a starter has a missed tackle rate higher than 11.0%, and that’s Derrick Moore at 13.6%. Junior Colson is the team’s most impressive tackler, as he leads the team with 29 tackles and has the lowest missed tackle rate amongst starters at 3.8%(!), which is lowest amongst any Big Ten LB with meaningful snaps