A Duke Loss Was Always Going To Look Like This


After a boring first round and an exciting round of 32, we’re through the first weekend of March Madness. Here’s a rundown of four major storylines that jumped out at me and the numbers behind them:

Happy trails, Duke

That Duke’s run as the No. 2-seed in the East would ultimately end against No. 7 South Carolina, was undoubtedly surprising. But the way it lost wasn’t. Duke struggled this season to pressure opposing guards into turnovers and is not a strong rebounding team. Against the Gamecocks, Duke lost the turnover battle 18-11, and the Blue Devils were outrebounded 37-34, grabbing a measly 62 percent of the rebounds available on defense. Both performances were among the worst of Duke’s season, and they happened to be in metrics that had correlated with disappointing Duke games all season, particularly at the defensive end of the floor.

No wonder South Carolina on Sunday posted the fifth-highest offensive rating Duke had allowed all season long.


VIDEO: How the Villanova and Duke losses shook the bracket

FiveThirtyEight: How the Villanova and Duke losses shook the bracket

The Gamecock guards also destroyed what had been a strength of Duke’s defense: Keeping opponents off the free-throw line. Against the Blue Devils, South Carolina’s backcourt drew 26 free-throw attempts, with Sindarius Thornwell and Rakym Felder making 12 and 10 trips to the line, respectively. The Gamecocks’ 32 total attempts tied for the most Duke yielded in a game all season, and South Carolina made 84 percent of them — including several crucial ones down the stretch to extinguish a potential Duke rally.

For Duke, it was the kind of flat performance associated with most of its midseason struggles. It was also a game that wouldn’t have come as much of a surprise if you hadn’t tuned in for Duke’s dominating ACC tournament run.

The favorites play it close (again)

Although the second round of the NCAAs wasn’t as upset-free as the first, it’s still been a good year for favorites at the tournament. Through Sunday night’s games, better-seeded teams have won 38 of 48 matchups, or 79.2 percent — tied for the best rate for favorites through two rounds since 1991. (Going into this year,1 the average winning percentage for favorites through the first two rounds was 73.2 percent.) Although we’ve seen a No. 1 seed (Villanova), two No. 2s (Duke and Louisville) and a No. 3 (Florida State) be dispatched at a relatively early stage of the tourney, the rest of the bracket’s chalk is entirely intact, with only one double-digit seed (West No. 11 Xavier) still alive to play Cinderella.

That being said, this year’s favorites have been pretty lucky to win so often. Their average margin of victory is currently 7.9 points per game, which is below the historical average of 8.1 points per game even though they’ve won more games than usual. In other words, the games have been close enough to allow for slightly more upsets — it’s just that things have broken the better seeds’ way more often than we’d expect. The good news is that we’re now left with an unusually strong, balanced set of teams for the Sweet 16, which should lead to even tighter, more exciting games.

Weekend winners

Every team that punched a ticket to the Sweet 16 is now significantly closer to winning the tournament than they were at this time last week. But some teams have seen their chances improve more than others, based on what happened in the bracket around them. For each of the remaining teams, here are the swings in their chances of getting to the Elite Eight, making the Final Four and winning the championship as a result of this weekend’s games:

CURRENT CHANCE OF ADVANCING TO CHANGE SINCE END OF ROUND 1
REGION SEED TEAM ELITE 8 FINAL 4 CHAMPS ELITE 8 FINAL 4 CHAMPS
East 3 Baylor 56% 23% 4% +27 +14 +2
East 4 Florida 54 34 6 +41 +28 +5
East 7 S. Carolina 44 16 1 +34 +14 +1
East 8 Wisconsin 46 27 5 +35 +22 +3
Midwest 1 Kansas 70 53 15 +15 +15 +5
Midwest 3 Oregon 43 11 2 +20 +4 +1
Midwest 4 Purdue 30 19 4 +10 +7 +1
Midwest 7 Michigan 57 17 3 +38 +11 +2
South 1 N. Carolina 68 34 9 +10 +5 +2
South 2 Kentucky 68 40 11 +22 +11 +3
South 3 UCLA 32 14 4 +12 +5 +2
South 4 Butler 32 12 2 +7 +2 +1
West 1 Gonzaga 66 46 18 +4 +4 +4
West 2 Arizona 68 26 9 +30 +10 +5
West 4 W. Virginia 34 20 6 +8 +5 +3
West 11 Xavier 32 8 1 +25 +7 +1
How our tournament predictions changed during the round of 32

Florida won big this weekend — both literally (they crushed Virginia 65-39 on Saturday night) and figuratively. With both the No. 1 and 2 seeds in the East now eliminated, the Gators’ chances of making the Final Four skyrocketed. Then again, so did Wisconsin’s: Our model now thinks the eighth-seeded Badgers are the second-most-likely team to represent the East in Phoenix.

The other regions saw less movement, which makes sense considering that they haven’t been as chaotic. Midwest No. 1 Kansas’s path got slightly easier with No. 2 Louisville out of the picture; by contrast, West No. 1 Gonzaga’s odds were relatively unchanged after No. 2 Arizona (and No. 4 West Virginia) stayed on a collision course with the Bulldogs. And the same was largely true for North Carolina in the South, after both Kentucky and UCLA moved on.

No tournament magic for Izzo this time

I’ve written before about how Michigan State coach Tom Izzo squeezes more NCAA tournament victories out of his teams than just about any other coach ever, and it seemed like he was set up for more of the same this year. Contrary to what our model predicted before the tourney, Midwest No. 9 Sparty knocked off No. 8 Miami with ease in the first round. And in their second-round game, MSU nearly pulled even with No. 1 Kansas with about 12 minutes left. Visions of another overachieving Izzo Final Four run were surely dancing through heads in East Lansing — but, alas, it was not to be. KU switched on the afterburners down the stretch and turned what had been a close game into a 20-point blowout.

Taking a No. 9 team to the Final Four would have been a remarkable run even by Izzo’s standards. He’s gone that far seven times as Michigan State’s coach but never with as low a seed as he had this season. Izzo’s Final Four squads have carried seeds No. 1 (1999, 2000, 2001), No. 2 (2009), No. 5 (2005, 2010) and even No. 7 (2015). But only five teams seeded ninth or worse have made the Final Four since the tournament expanded to 64 teams in 1985, and only one — Wichita State in 2013 — was specifically a 9-seed. No. 9-seeds have a particularly hard mission: They must upset a first-round opponent and then immediately face the No. 1 seed in the second round. Taking that difficult a route to the Final Four is just about the only thing Izzo hasn’t pulled off yet.


You have three days to rest up before the games begin again. Take a breath, take stock of your bracket, and start prepping for another four days of nonstop basketball. Come Thursday, those win probabilities are going to start changing again.

Check out our March Madness predictions.