fjords

12/3/2019 – Michigan 43, Louisville 58 – 7-1

The John Beilein era was not exactly thick with offensive debacles, so everyone was probably thinking of the same two games as Louisville constricted Michigan's offense last night. One was last year's Sweet 16 game against Texas Tech. That ended Michigan's season, after which three guys and the coach left. Also that was a historically great and very weird defense. So there's no data on how Michigan recovered from that.

The other, though, was Michigan doing this at South Carolina three years ago:

points, twos, threes

This, too, came after Michigan had just lodged a couple of encouraging wins in a tournament. They'd hammered Marquette and SMU in New York. The trip to Columbia was Michigan's first road game of the season. Like Louisville, South Carolina had played no one of note, winning a couple of games against fringe top 100 foes and mixing in some cupcakes. Any spare prep time they'd had all season got applied to Michigan since it was the first real test on the schedule. And Michigan got blitzed.

The good news is that result had close to no bearing on the rest of the season. This was the Maverick Morgan year. Until mid-January Michigan's defense did a fair job of obscuring what would end up the #4 offense on Kenpom. The South Carolina game (and a Texas game a couple weeks later Michigan actually won) were so far out of trend that they look like a different team took the court for a couple nonconference games. Torvik's chart of adjusted offensive efficiency that year:

red: ncaa average efficiency, thick yellow: moving average, thin yellow: linear trend line, dotted: 5 game moving average

I don't have to tell you which one is South Carolina. For their part, they had an elite defense all year, finishing third, and went on a Final Four run after landing a seven-seed. Louisville is going to have an elite defense as well.

It'll be fine. If you'd offered me three of four against Iowa State, UNC, Gonzaga, and Louisville—with none of them at home—you'd lose your arm because I shook it so fast. Eating a schedule loss at the end of that sequence is hard to watch but less indicative of what's going to happen down the road than Atlantis.

[After THE JUMP: Phil Martelli on fjords]