I don't see why not. Could just as easily get blown out, too.
It's kind of weird... The way I look at it, our best chance of picking up 2 wins in this thing is probably to lose to IU, then beat St. John's and (potentially) UNLV.
If we beat IU, then we have Vandy. If we lose that one, we get loser of KU/(UCLA/UNLV winner).
Someone with Kenpom might have to pull up our chances against each team, but just an initial estimate and assuming the favorites win every game will produce the following:
If we lost to IU, we would most likely play St. Johns (id guess we'd have a 67% chance of winning this game) and then UNLV (probably a toss-up, say 50% chance of winning) or Chaminade (for simplification I'll say he have a 100% chance of winning, if you lower it to the true percentage of approx. 95%, it wouldnt change our final expected value too much)
If we upset IU, assuming we would play Vandy in the semis, and then KU in the finals or UCLA in the third place game, with the following: P(beating Vandy) = 20%, P(beating UCLA=30%), P(beating Kansas 10%).
Given these assumptions, if we
LOSE to Indiana we can expect 1.33 wins with a 33% chance of winning two games. If we
Beat Indiana, our expected wins is 1.46 with a 44% chance of winning 2 or 3 games.
If we lose to IU, our remaining games become much more winnable, but not enough to make up for the game we already lost (IU).
Of course all of these could completely change depending on if Hudson plays/doesnt play. I've also assumed the favorite wins every other game, which is incorrect. Working on a new model now with every game included.
Which of my winning percentages seem too high? Too low?