Bracket Lesson #1: Which Upsets To Pick?

I was going to wait until tomorrow morning before starting to analyze the brackets, but a quick glance at the tsunami of web traffic right now suggests that people are looking to fill out their brackets NOW. I guess it makes sense that most brackets will be filled out right after the selection show, or at the last possible moment on Thursday morning. Depends what type of person you are.

Anyway, the first thing I want to talk about is picking upsets. And by "upsets", I'm talking about the 7/10, 6/11, 5/12 and 4/13 games. Picking a 9 over an 8 is not an upset, in fact, the 9 seeds have won 54% of the time since the bracket expanded to 64 teams in 1985, so it's an "upset" to pick the 8 seed. Wins by 14, 15 and 16 seeds are so rare that I generally strongly recommend against picking any of those in your brackets. This year is no exception: the only upset I'd even consider is North Dakota State over Kansas. And again, don't pick it unless your bracket challenge has a scoring system that really rewards upsets, otherwise it's a high risk, little reward play.

Now, I've talked about 5/12 and 6/11 games here, and I linked there back to an analysis I did a couple of years ago on 5/12 games and another one on 6/11 games. The lessons from both analyses are the same, and can carry over the 4/13 and 7/10 games. The lesson is: bet on underdogs from major conferences, and bet against favorites from mid-majors and small conferences. In fact, the stats are so remarkable that if you have a major conference 11 or 12 seed against a small conference 6 or 5 seed, just take the 11/12. Don't even look at the team names. This year we've got a number of major conference teams with 12+ seeds, but no 11 seeds. The 12 seeds are Arizona and Wisconsin. The 13 seed is Mississippi State. The formula also suggests that Gonzaga and Xavier would be vulnerable 4 seeds. If we look specifically for small conference teams favored against BCS conference teams, I see only one example: Utah against Arizona.

The Short Version:

Slam dunk picks: Arizona over Utah.

Not-so-slam dunk picks: Mississippi State over Washington, Wisconsin over Florida State.


In other words... this rule isn't going to do a lot of good for you this year. In fact, a more likely upset pick than Mississippi State and Wisconsin is taking Utah State over Marquette. And that has nothing to do with player psychology, and everything to do with the fact that Marquette just hasn't been very good without Dominic James.

It's unfortunate that this analysis, which is often very useful, isn't so useful this year. I'll be back in a moment with some more useful things to look for in your brackets.