Beware of the 10s

Lots of people will be figuring out their NCAA brackets over the next few days after the seeds are announced on Sunday. I think this year's tournament will be one of the hardest ever to predict since there seems to be so much parity in college basketball right now. After the top three teams (Nova, UConn, and Duke), the next 20-25 teams are pretty much even. So this seems to be a good year to predict some upsets, especially for double-digit seeds.

Without knowing the seedings yet, a strong - but risky - guess would be to pick some 10 seeds making it to the Sweet 16. I argue that a 10 seed is far superior to an 8 or a 9, especially this year. Why? Because an 8 or a 9 faces a toss-up game in the first round - its hard to earn any respect from a win (since neither team can claim an upset), plus an 8 or 9 is almost destine to loss in the second round (72-12 advantage favoring the 1s). Perhaps, some teams would relish the spotlight of playing a one-seed and the opportunity for a monumental win. Still, pragmatic teams should hope for a 10 over a 8 or a 9, since this seed presents an opportunity for a first round upset over a 7 seed. A 7 seed is probably only marginally better than an 8/9, but the 10 seed still earns an upset.

If the 10 seed pulls off the upset, they only have to play a 2 seed. A friend of mine tried to argue that this isn't much better than playing a one-seed. History indicates otherwise: 10 seeds have split with 2 seeds (according to EPSN, 2 seeds only hold a 16-14 advantage when matched with 10 seeds in the second round). This is a much better record than 8s have versus 1s (29-9) or 9s versus 1s (43-3), although a 10 has only a 19% of beating a 7 and earning that match-up with the 2 seed.

The difference in records of 8s versus 1s and 9s versus 1s is interesting as well, and perhaps suggests that the NCAA does a very good job of seeding, even for seemingly modest differences (....especially when you consider that the difference between an 8 or a 9 has absolutely no impact on the tournament!).

So this year, I suggest looking at the 10s and seeing if anyone looks promising to make it to the Sweet 16. 10 seeds from major conferences have particularly done well. I'll make one prediction: if UNC is a 2-seed, consider whether the 10 seed looks like it can beat the Tar Heels. UNC is too young and could be knocked off by a 10 seed from a major conference with experienced players.

Check out ESPN's seed analysis for other insights into seeding stats:

http://insider.espn.go.com/ncb/ncaatourney06/insider/news/story?id=2353126&action=login&appRedirect=http%3a%2f%2finsider.espn.go.com%2fncb%2fncaatourney06%2finsider%2fnews%2fstory%3fid%3d2353126