In all seriousness, as somebody who watched Virginia a lot, I was sort of shocked my alma mater got into the field. Like...really?

I do think, however, that the tournament selection process itself illustrates a very interesting tension: whether the committee should be using predictive measures or just results-based ones.

That tension is most obvious in the case of Virginia. The Cavs had by far the worst predictive metrics of any at-large tournament selection. However, they were without a doubt one of the most fortunate teams in college basketball, with every loss by double figures until their season finale (when the karma gods smacked them with a banked 3 at the buzzer in regulation before losing in overtime) but a 9-0 record in games decided by six points or fewer.

However, the committee can’t go by what it “knows;” it has to follow the rules set out for it. Predictive metrics all think Virginia has no business being in the tournament, and I have zero expectations of success against Colorado State on Tuesday, but Virginia’s resume had wins over Texas A&M and Florida and precious few losses anywhere. The losses that did happen were all by 90, but again, binary win-loss results have to matter for the games to have meaning. Sure, we can watch the Pitt game – the Panthers walked into Charlottesville and smoked the Cavaliers — and “know” Pitt was better, but Pitt didn’t have as good a tournament resume as Virginia.

So, Virginia got in over the Big East trio and Oklahoma and a few others that were almost certainly better, but didn’t have the game outcomes to back that up. That most of the big public bracket experts predicted 67 of the 68 field members indicates that the selection committee executed this part pretty faithfully. It’s a sign of a transparent process when outsiders can replicate your work.