A couple years ago I learned something simple and profound about analytics in general: new information is better than old information. When projecting a player to the NBA from college, your data set is limited; this is why busts happen. When a player gets to the NBA (or whichever the "big show" is) you learn a lot more about their abilities pretty quickly. After a year of that player being in the league, you have a much clearer picture of who that player will become. But people hang onto their old presumptions, which is why people were convinced that they should keep guys like Bismack Biyombo or Jan Vesely around much longer than they probably should've.
I believe that the "new info > old info" adage is true more generally, also. Before the college football season starts, you have some limited set of data (returning starters, previous team quality, etc) and you learn a lot more about a team as the year goes along. For instance, we know that Regular Michigan is better than we expected given preseason data.
But in the rush of week-to-week polls and narratives, we overestimate the amount of new information we've got. Take, for instance, last year's AP Top 5 at this point in the season: 1) Mississippi State 2) Florida State 3) Ole Miss 4) Baylor 5) Notre Dame. Exactly one of those teams finished the season in the Top 5 (Florida State). Notre Dame finished unranked. You get the picture.
And Mark Dantonio essentially said as much in his press conference on Tuesday:
Q. Mark, halfway through the season, I'm curious in you've gotten closer to understanding the identity of this football team, especially with the close wins and the way that they played down the stretch?
COACH DANTONIO: Yeah, you get closer. I think games like this coming up start to define you and games throughout the rest of the season. You're only halfway through, but I think what you know about Michigan State right now is that we will hang in there and we will hang tough and we will play through some pain, and that we will play through some adversity and that we've come out on the other end, and that's the positive. Are we perfect? No. Do we have some playmakers? Yes. That's what you know, halfway through.
As somebody who likes to pick things apart and create a #narrative, it pains me to say this: have faith. Our sample size is just 6 games. I think Dantonio's earned that much.
When Regular Michigan Has The Ball
De'Veon Smith has clearly emerged for RM (Regular Michigan) as the lead back, but they've already got four different backs who have received at least 26 carries: Smith (5.0 yards per carry), Derrick Green (3.5 ypc), USC transfer Ty Isaac (7.1 ypc), and Drake Johnson (4.9 ypc). Smith got injured during the BYU game and Green filled in; Johnson got the bulk of the work against Maryland with Smith out.
The offensive line for RM has improved in this area, but much of the rushing game improvement for them is attributable to the backs themselves; UM ranks 81st nationally in adjusted line yards and 71st nationally in opportunity rate (Bill Connelly's offensive line metrics).
I have to point out here that 140 of the 781 yards given up by MSU this year came on two plays: the 68-yard touchdown by Markell Jones against Purdue, and the 72 yard run by Paul James against Rutgers. That's 18 percent of the total. And obviously those plays need to be cleaned up, but MSU is averaging a much more respectable 3.1 yards per carry defensively absent those two plays. I'm hopeful in this regard.
Quarterback Jake Rudock will get involved to some degree here as well. He's averaging 5.6 ypc on 23 carries.
The classic stoppable force versus the movable object. Who knows if Darian Hicks will be able to go or not; either way, MSU's secondary will be a little shaky. If Hicks doesn't play, expect to see a true freshman in this game, whether that's Josh Butler or David Dowell.
The good news is that Jake Rudock and company have not been world beaters themselves, largely because they haven't needed to be. Rudock has only needed to attempt more than 30 passes twice, in the Utah and Maryland games. Stats suggest that Rudock is sort of a poor man's Chris Laviano from Rutgers. Laviano at least has a positive touchdown to interception ratio and a completion rate north of 70%. Anyways, RM has been relatively efficient (45th nationally in passing success rate) but not particularly explosive (80th nationally in passing IsoPPP).
Oh, and RM also doesn't have anybody like Leonte Carroo. Jake Butt, Amara Darboh, and Jehu Chesson are all adequate players, though none of them average more than 7.8 yards per target. I'm a little less into Butt than most Regular Michigan fans, considering his unspectacular catch rate of 62.9 percent.
The more interesting matchup is along the line, where RM is top-ten nationally in terms of adjusted sack rate, while MSU is 17th in terms of getting to the quarterback. If Rudock is pressured consistently then this phase of the game could be a major positive for MSU.
When Michigan State Has The Ball
My biggest concern of the game. The state of State's offensive line is up in the air, to say the least. Check out my guide to the possibilities here. In any case, 3.9 sack-adjusted yards per carry against a terribad Rutgers defense is a sign of danger, especially with any offensive line configuration where Jack Conklin is not playing left tackle. The good news is that I suspect MSU will have 2 out of 3 guys back along the offensive line (not bad).
Even with good health news, this Regular Michigan defense will be difficult to run the ball against. Since giving up 3.5 yards per carry to Utah, RM hasn't surrendered anything north of 2.6 ypc. RM likes to play a 4-2-5 base, with Jabrill Peppers as a STAR-ish sort of player. Peppers is adept at blowing up rushing plays off of the edge, which will make the jet sweep package with R.J. Shelton especially interesting.
RM is having their own health issues along the defensive line, with Mario Ojemudia and Brian Mone both out for the year. The difference is that it's easier to have depth with four players instead of five and cohesion is less important on the defensive side. DE Chris Wormley, DT Ryan Glasgow, DT Willie Henry, and DE Royce Jenkins-Stone have combined for 19 tackles for loss (!) and backups Taco Charlton and Maurice Hurst have combined for 5.5 sacks as well.
I expect some interesting playcalling with regards to running the football. WHYldcat would infuriate but not surprise, and also true read-option with Cook is on the table given the importance of the game. The goal will be to eke out 3.5 ypc in this game. That would be something of an accomplishment.
The matchup of the game, to me, is Aaron Burbridge against Jourdan Lewis. Burbridge is averaging better than 10 yards per target, while Jourdan Lewis is earning effusive praise basically everywhere. If Josiah Price is back, the remainder of the MSU receiving corps is solid but not spectacular, and Connor Cook may or may not have much time to throw given RM's nationally top-5 sack rate. A definitive win one way or the other between Lewis and Burbridge would go a long way for either team.
I'd also like to point out that RM gave up a 72.7 completion percentage to Utah's Travis Wilson. RM is top-10 nationally in basically any conceivable passing defense metric, but the only two semi-respectable quarterbacks they've faced are Wilson and BYU frosh Tanner Mangum (who is far from a known commodity). Their defense has been totally dominant against inferior passing competition; ask the 2014 Spartans how that goes against elite college quarterbacks.
I wouldn't be surprised to see some designed rollouts to combat the effectiveness of Regular Michigan's pass rush, depending on the health of the offensive line.
You know the woes: kick and punt returns early in the year, bad punt snaps against Rutgers and Purdue, Michael Geiger's inconsistencies, and exactly one punt return for exactly one yard on the season. The issues are myriad and pervasive.
But these are correctable things! Easily correctable, even. And given MSU's history of solid special teams work, there are reason for optimism in this realm. My bold prediction is that there are no special teams gaffes for MSU in this game, and they could be a real positive.
Here's what you need to know about R. Michigan's special teams: they've only got 5 kickoff returns on the year, and one of them was the opening touchdown against Northwestern. They're middling in the Big Ten in terms of punting and kickoffs, and kicker Kenny Allen is 7-for-9 this year. Field position has been a nice advantage for RM this year, but that has more to do with their defense than anything else.
Bottom Line and Prediction
Given the injury situation for MSU the location of the game, and the dominance against Northwestern, I understand why Regular Michigan would be favored. But 7.5 points is insane. This MSU team will be more competitive than that, it seems obvious to me.
If MSU miraculously has a full strength offensive line and a healthy Darian Hicks, I think the entire complexion of the game is changed. But even with just Conklin and Kodi Kieler, I think there are enough reasons to presume MSU can find ways to move the ball with the skill position talent, and has enough defensively to make Rudock beat them.
MSU 33 - UM 28