So the first two days.
1) Biggest surprise:
No question is Norfolk St. I had them rated around #195. I did have Missouri (and Duke) pegged as among the most vulnerable top tier seeds because they had such mediocre defence, but I don't think I would have expected a 30% 3-point shooting team to shoot the lights out. It's not like they stopped or slowed down Missouri either. Because the other 2-15 game involved Duke, people will say it was a bigger upset. I don't think it's even close. Lehigh was a top 75 team, and thus at least a plausible underdog (the same way UNC-Asheville was fairly solid for a 16 seed). If I were to assign a probability to Missouri losing it would have been less than 1%. Duke was at least around 10% in my mind.
For the record, I picked neither upset (and neither did any of you, more than likely). What I did do however was not have either team advancing very far. I (usually) had Florida upsetting Missouri and I (occasionally) had Xavier beating Duke and definitely had UNLV or Baylor winning after that. Both predictions look to be useful for scoring purposes (other than the UNLV option, which was poorly thought out). More pressing, I lost none of my final four teams in the first two days. Which is comforting. Mostly because the President did.
2) Belmont is really annoying. They should be able to beat slow plodding teams with efficient offences like Georgetown. Oh well. I would think it very possible NC State can still take care of business and advance, but it's a harder path than I would have expected.
3) Of the other upsets. Ohio beating Michigan was no big surprise, Ohio was pretty solid and Michigan was dramatically overrated (and I tended to pick it). South Florida beating Temple was much more so (I had Temple coming out of that pod, now it looks like USF). NC State was rated higher than San Diego State AND the game involved a west coast team traveling (called that one no problem). VCU had the annoying quality of being VCU (though admittedly a better VCU ranking than last year coming in and so a tougher out). Really the only surprises for me were USF and Colorado. I was impressed by Colorado after that but I don't think they're going anywhere.
4) 7-10/8-9 games do not involve upsets in my view. The margins by Gonzaga and Florida were impressive though. Based on public perception, Iowa State was an upset, but it really wasn't. Connecticut was living on a name. Of the 8-9, 7-10 games I missed only Memphis-St Louis, though I hedged a bit on St Marys-Purdue. And the Memphis game. They really got screwed with the seeding. I view that as a success. 7-1 on what are supposed to be tossups implies that the committee might have missed something (overrating Notre Dame and Southern Miss in particular, along with underrating Creighton, and both Memphis and St Louis), but mostly I got lucky (with Creighton especially).
5) Close calls. St Bonaventure was totally unsurprising that that was a close one, Florida St was way overrated and St Bonaventure was a tough 14 seed. Same with Texas-Cincinnati, though Texas made it very odd by getting blown out first then going with playing in the second half. (I managed to go 8-0 in that region). Long Beach was also another closer matchup on paper than a 5-12 should have been and it played out that way. Baylor also had a tougher matchup (South Dakota St rated higher than Colorado does).