EDIT: I’ve started a Yahoo! bracket pool for the procrastinators among us — see here. Notice the funky scoring rules — lots of bonus points for picking upsets!
These aren’t my ratings per se, but instead a compilation of six different sources: simulations based on computer ratings from Ken Pomeroy (source), Jeff Sagarin and Kvam/Sokol; a meta-simulation from WhatIfSports; Vegas (actually dodgy offshore casino) odds — these need to be adjusted to take out the “juice” the casino builds in — and the composite picks in Yahoo! Tourney Pick ‘em (not technically probabilities but just the number of people who picked each team).
Quite a few differences of opinion! All of the models are way down on Kentucky, which is the second-favorite selection in Vegas. The computers, on the other hand, have several funky darkhorses like Wisconsin, BYU (very dangerous if they make it to the Sweet 16 since the regional is held in Salt Lake City), and Baylor. They also like Duke a lot, in part because their region is very weak through the Elite Eight. BTW, I don’t think the computer models are foolproof — it would probably be advisable to take into account things like how well a team plays away from home, free-throw shooting (definitely important in close games), travel, and injuries — but they should provide for a good general guide.
If you’re still filling out your bracket, you probably want to tend to bet against the Yahoo! picks, since there’s definitely some incentive to be contrarian in a winner-take-all pool where high-variance strategies are rewarded. This is especially so if your pool has a lot of entrants, weights late-round games disproportionately, and does not give a bonus for picking upsets. If you pick Baylor or something and get it right, odds are that just one or two other people in your pool will have them as the winner and unless you really screwed up everything else, you should be in the money. But if you go with a Kentucky or a Kansas, you need to get a lot of other things right. In the one pool where I took Kansas, I had three relatively low seeds along with them in the Final Four, figuring that if I get Kansas and any of the longshots right, I should have a pretty good chance to money.
This should be a good tournament from a gambling perspective. Lots of the 4/13, 5/12, 6/11 games have serious upset potential — check out that West regional in particular.