My preview before the tournament began said that the women’s tournament isn’t the place to go looking for upsets, at least not in the early rounds.
I hope you took that advice to heed. The higher seed won 31 of 32 games in the Round of 64. That’s right: all but one game. In the Round of 32, a couple of 5-seeds (Baylor and Colorado) prevailed over 4’s (Kansas State and Virginia Tech). And #7 Duke beat #2 Ohio State. But even that wasn’t too surprising. The model’s composite power ratings thought Duke was underseeded and should have been a 5, while Ohio State was overseeded and should have been a 3.
So really the only question has been who looked most dominant in winning the games they were supposed to win. And the answer is South Carolina, which so far has wins by margins of 52 (!) and 47 (!) points. The Gamecocks are undefeated, and their only loss in the past two seasons came to Caitlin Clark and Iowa in last year’s national semifinals. Now, South Carolina had a 96 percent chance of reaching the Sweet Sixteen before the tournament began, so this isn’t that much of a shock. Still, the model, being a good Bayesian, updates its ratings based on margins of victory and the Gamecocks were even more dominant than it expected, leading #8 seed UNC 56-19 at halftime before laying off the gas pedal. Even a modest improvement in a team’s power rating compounds over four remaining games, so they’re up to a 53 percent chance to win the tournament from 41 percent before the games began.
That improvement in win probability needs to come from somewhere, and it mostly chips away at the chances for the other top contenders. Clark’s Iowa Hawkeyes are down from 11 percent to 9 percent chances of winning the title, and UConn is down from 16 percent to 13 percent despite also winning out. Technically speaking, Baylor, Duke and USC saw their overall tourney-winning odds improve, but all are long-shots and it was only by fractions of a point.
Odds follow for paid subscribers, beginning in the so-called Albany 1 region. For men’s Sweet Sixteen numbers, see the separate post from early Monday morning; I promised I’d also update the men’s numbers if there was any new injury news, but there hasn’t been yet.
Now, Portland 4: