Why Michigan-Wisconsin is Week 4's biggest game and more

In terms of rankings, college football's Week 4 headliner is Georgia-Notre Dame. That's fine. It should be a hell of a spectacle. But to me, Michigan-Wisconsin is the biggest game of the week.

We have so many questions regarding both teams. Can Michigan snap out of its early offensive funk? Was Wisconsin's early domination of lesser foes a sign of big things to come? Is Badgers QB Jack Coan really as good as he has looked so far? Is Michigan's O-line really as bad? Can Jim Harbaugh finally win again as an underdog (he's 0-6 at Michigan, and it has been a while since Stanford-USC in 2007)?

Even the weather is cooperating. Michigan-Wisconsin games should be chilly, dreary gray or both. And the forecast currently calls for rain. Perfect.

I love that this game pops up as early on the calendar as it does, and I love that the early Saturday slate is mostly clear. All eyes on Madison.

1. The year so far

Both the Badgers and Wolverines are coming off bye weeks and have therefore played only two games. But they've left us with both limited and lasting impressions.

If we ranked teams purely on how they've looked this season, ignoring any previous impressions or assumptions, Wisconsin would be the No. 1 team in the country right now. The Badgers humiliated USF and Central Michigan, outscoring them by 110 points and outgaining them by 819 yards. The next points they allow will be the first since the first quarter of last year's New Era Pinstripe Bowl. They have been devastating, albeit against teams ranked 86th and 119th, respectively, in SP+.

Michigan, meanwhile, is 2-0, but the Wolverines are still threatening to steal the early-season Existential Crisis title from USC. They looked decent enough in a 40-21 win over Middle Tennessee but needed a missed field goal attempt at the end of regulation and an overtime turnover to survive Army. Harbaugh announced his intention to modernize his offense this offseason, bringing in offensive coordinator Josh Gattis, but results have been lacking so far.

It's still early, though, and Michigan is still unbeaten. It's not too late to find traction, but it might be if they don't find it Saturday.

2. What's Michigan doing differently?

Michigan finished 25th in offensive SP+ last season; that's not exactly destitute. The Wolverines scored more than 30 points in nine of 13 games last season, including seven of nine conference games and three contests against teams that finished in the SP+ top 20.

The season was bookended with meek offensive performances, however: a 24-17 loss to Notre Dame (307 total yards, 4.5 per play) and a 41-15 blowout at the hands of Florida (326, 4.9). Harbaugh has brought Michigan back to the land of the near elite, at the least, but it seemed like predictability, along with the inability to make big plays against good defenses, was holding him back.

Through two games, the Harbaugh-Gattis marriage has indeed produced some changes in the Wolverines' offense.

They've picked up the pace. Michigan currently ranks 28th in my adjusted pace rankings, which compares your tempo to what's expected based on your run-pass ratios. The Wolverines are snapping the ball nearly one second faster than this expectation, a clear change after taking 3.2 seconds per play longer than normal last year (123rd).

They're disguising their intentions better, at least when it comes to formations. Last year, Michigan lined up in 11 personnel (one running back, one tight end) 313 times and threw 66% of the time from that set. Meanwhile, it lined up in either 12 (one RB, two TE), 21 (two RB, one TE), or 22 personnel (two RB, two TE) 511 times and ran the ball 68% of the time from that look.

It's pretty easy to prepare for that. Sure, the Wolverines had enough talent to move the ball anyway, but you can see how that might be an issue against teams with similar talent.

This year, Michigan has moved away from 21 and 22 personnel (i.e. the use of a fullback), sticking almost entirely to 11 and 12. And the Wolverines' run-pass rates from both are near 50-50 (56% pass from 11 personnel, 52% run from 12).

When they pass, they're also more likely to use the entire field. Whereas last year's passing game was defined primarily by outs, screens and go routes, Shea Patterson is throwing more intermediate routes this year -- 20 passes per game between 11 and 20 yards downfield, compared with 16.2 in 2018. He's also doing a little better within this range, averaging 8.4 yards per pass compared to last year's 7.8. And that's without last year's leading receiver, Donovan Peoples-Jones, who hasn't played yet this year because of injury.

Of course, you can still pretty easily identify what Michigan wants to do. All you need to know is the down and distance.

I define standard downs as first downs, second-and-7 or less, and third- or fourth-and-4 or less. Passing downs are everything else: second-and-8 or more, third- or fourth-and-5 or more. Obviously teams run the ball more on standard downs (58% of the time on average) than passing downs (34%), but the Wolverines are far more extreme than the national averages: They're running 67% of the time on standard downs (24th most overall) and only 20% of the time on passing downs (14th least).

Both Middle Tennessee and Army took full advantage of this tendency, blitzing and stunting with abandon. And Patterson was dealing with plenty of second- and third-and-longs because MTSU and Army also knew when the Wolverines wanted to run. Michigan ranks 91st in rushing success rate so far, and in his first game as feature back, blue-chip freshman Zach Charbonnet gained only 100 yards in 33 carries against Army. The return of star lineman Jon Runyan from injury will help, but those are some pretty dreary numbers.

With a week off to reassess and self-scout, I'm guessing Michigan rights some of these extreme tendencies a bit, perhaps giving Patterson a few more first-down pass opportunities and giving Charbonnet some second-down carries, when maybe there aren't eight Wisconsin defenders in the box. If the Wolverines don't break tendency, the Badgers' defense is going to tee off.

3. Why has Wisconsin been so dominant?

With Wisconsin taking a step back in 2018 and young teams like Nebraska and Minnesota looking primed for a jump, it was easy to assume this year's Big Ten West race would be wide open. And maybe it still will be. But as things currently stand, Wisconsin enters conference play as the West favorite once again. Nebraska, Minnesota and Northwestern have all been underwhelming, and ... again ... Wisconsin 110, First Two Opponents 0.

What has made the Badgers so impressive so far? Let's start with the defense.

Success rate is a football efficiency measure that determines each play a success or non-success based on whether it gains 50% of necessary yardage on first down, 70% on second or 100% on third or fourth down. It functions as an on-base percentage of sorts, and Wisconsin's defense has been an efficiency machine thus far: first in success rate allowed, first against the run, second against the pass, first on standard downs, fifth on passing downs.

Do South Florida and Central Michigan have good offenses? Not even close. But there's predictive value in treating bad teams and units like truly horrible teams and units, and that's what Wisconsin has done.

The Badgers haven't been amazingly disruptive, though, and I'm curious what that means now that there's an upgrade in competition. Wisconsin ranks 42nd in sack rate and 22nd in havoc rate (total tackles for loss, passes defensed and forced fumbles divided by total plays); these aren't bad rankings, obviously, and they're improvements over last season, but Michigan doesn't give up many negative plays (fourth-down rushes against Army aside) and might be able to stay in manageable third downs without them.

Offensively, coach Paul Chryst and offensive coordinator Joe Rudolph have done exactly what Harbaugh and Gattis need to do: They've broken tendencies. Instead of leaning on running back Jonathan Taylor while breaking in a new starting quarterback (Coan) and mostly new offensive line, they've tried to make life easier on everyone by throwing more on run downs. It's a life hack for young QBs.

After running the ball 74% of the time on standard downs last season (seventh most in FBS), they're doing so only 63% of the time so far in 2019 (39th). That's still more than the national average, but it's giving Coan and his experienced receiving corps some easy completions against loaded defensive boxes, and the QB has taken advantage: 76% completion rate, 8.5 yards per pass attempt (including sacks).

Combined with the huge scoring margins, this tendency has also kept some wear and tear off of Taylor, who has only 35 carries in two games.

Was this simply a product of playing bad competition? Can we expect 40 carries from Taylor against Don Brown's ever-aggressive Michigan defense? Will standard downs passing even work against this D?

4. What can we learn from past meetings?

The past two games between these schools were decided by double digits. Two years ago in Madison, a limited Michigan team led in the third quarter, but Wisconsin pulled away to win by two touchdowns. Last year, Michigan bounced back, scoring 31 unanswered points and cruising 38-13.

There were similarities in the two contests. Michigan held a pretty significant field position advantage in both. The Wolverines also threw poorly both years (5.6 yards per attempt, including sacks, in 2017 and 4.2 in 2018) and didn't really slow down Taylor all that much (he had 19 carries for 132 yards in 2017, 17 for 101 last year).

The main differences between 2017 and 2018:

1. Michigan ran the ball well last year. The Wolverines had a 23% rushing success rate in 2017 and more than doubled that, to 55%, last year. That allowed them to both control the ball (6.3 plays per drive vs. 4.6 in 2017) and negate the fact they weren't throwing well.

2. Wisconsin's passing game was miserable last year. In 2017, quarterback Alex Hornibrook (who has since transferred) averaged 5.7 yards per attempt overall -- not great, but not terrible -- and went an impressive 7-for-13 for 121 yards on passing downs. When the Badgers fell behind schedule, Hornibrook was able to catch them up.

His trip to Ann Arbor last year, however, was a nightmare: 4.1 yards per attempt overall and 0.4 per attempt on passing downs before garbage time. He had no time to throw and no open receivers to throw to. Receivers A.J. Taylor and Danny Davis III caught four of seven passes for 106 yards in 2017, but only two of eight for 15 last year.

Knowing what we know about these teams, it would make sense if these two factors ended up being the most important once again.

In this way, it's plausible to say Quintez Cephus will be the most important player on the field. Wisconsin's leading receiver missed the 2017 game because of injury and was off the team for all of 2018 after being charged with second- and third-degree assault. He was reinstated last month after being acquitted. He has caught nine of 13 balls for 169 yards and two scores so far in 2019. He's giving Coan a major plus target out wide, and he's making it so that defenders have too many weapons to properly cover.

SP+ projection: Wisconsin (-3.5) 27, Michigan 18. SP+ is reading a lot into the Badgers' early dominance. We'll find out Saturday whether that was fool's gold or a sign of a big year to come.

Week 4 playlist

Besides Michigan-Wisconsin, here are 10 games -- at least one from each weekend time slot -- you should pay attention to if you want to get the absolute most out of the weekend, from both an information and entertainment perspective.

All times ET

No. 10 Utah at USC (9 p.m., FS1)
The stakes here are high. Per SP+, this is Utah's most likely division loss -- the Utes have just a 59% win probability. Add that, plus the fact that Utah's dark-horse CFP dreams would mostly die with a loss, to the ongoing existential crisis that is USC football, and you have an exciting Friday night affair.
SP+ projection: Utah 31, USC (+4) 27.

Early Saturday
Tennessee at No. 9 Florida (noon, ESPN)
Tennessee can turn its season around with an upset over a Florida team that just lost quarterback Feleipe Franks to injury. Will it happen? Probably not, but keep tabs on it just in case.
SP+ projection: UF 35, UT 17*
*SP+ doesn't take Franks' injury into account

Saturday afternoon
No. 8 Auburn at No. 17 Texas A&M (3:30 p.m., CBS)
This is the afternoon headliner, and it should be a nice, physical affair. Auburn's defense is as good as ever, and two weeks ago, A&M became the only team in 18 games to hold Clemson under 27 points. The first team over 20, wins.
SP+ projection: Auburn (+4) 25, A&M 23

There are three exciting midday deep cuts here, if Auburn-A&M isn't doing it for you.

Appalachian State at North Carolina (3:30 p.m., ACC Extra)
UNC has played three nip-and-tuck games thus far, and this could be a fourth. SP+ ranks App State 33rd overall, ahead of the Heels and fourth among Group of 5 teams. A marquee win here could be the prompt for a run at the G5's New Year's Six bowl bid.
SP+ projection: Appalachian State 37 (+3.5), UNC 30

UC Davis at North Dakota State (3:30 p.m., ESPN+)
Oh, hell yes: FCS No. 1 NDSU against Dan Hawkins' fourth-ranked UC Davis squad. NDSU is coming off an easy road win over a top-20 Delaware team, but the Aggies gave Cal fits in Week 1. This one will give us an idea of whether anybody can hang with the Bison in 2019.
SP+ projection: NDSU 39, UCD 24

SMU at No. 25 TCU (3:30 p.m., FS1)
TCU's defense looks as good as ever, and the Horned Frogs' run game, led by Sewo Olonilua and Darius Anderson, is terrifying. The passing game is terrifying in a bad way, however, and now TCU hosts a 3-0 SMU team capable of luring just about anybody into a shootout.
SP+ projection: TCU (-9) 35, SMU 20

Saturday evening
No. 7 Notre Dame at No. 3 Georgia (8 p.m., CBS)
You probably didn't need my advice to watch this one. Georgia gets a major chance to prove its bona fides ahead of conference play, but the Fighting Irish are a top-10 team per both polls and computer rankings. No givens here.
SP+ projection: Georgia 36, Notre Dame (+13.5) 23

No. 16 Oregon at Stanford (7 p.m., ESPN)
Stanford's season has gotten ugly in a hurry following blowout road losses to USC and UCF, and Oregon's post-Auburn rebound has been demonstrative: a combined 112-9 score against Nevada and Montana. But would it be much of a surprise if the Cardinal made life difficult for the Ducks in this one?
SP+ projection: Oregon (-10) 30, Stanford 19

Oklahoma State at No. 12 Texas (7:30 p.m., ABC)
For the past eight months or so, I've written that Texas will prove it's back when it actually performs at a high level against the teams it's supposed to beat, not just against the two or three best teams on the schedule. The Horns did just that against Louisiana Tech and Rice, but now comes a formidable test from a team that hasn't lost in Austin in 11 years.
SP+ projection: Texas 34.2, OSU (+5) 33.8

Late Saturday
Utah State at San Diego State (10:30 p.m., CBSSN)
You should maybe keep tabs on UCLA-Washington State to make sure the Bruins aren't showing signs of life, but in the meantime, this is probably your best hope for an actually exciting late-night game. USU's star quarterback Jordan Love takes on what is almost certainly the toughest defense he's going to face all year.
SP+ projection: USU 24, SDSU (+4.5) 20