If you are interested in figuring out various strategies for filling our your bracket, I would highly recommend Peter Tiernan’s BracketScience.com. At his site, you can choose the strategy that you like and Tiernan churns out the bracket for you. He’s got about a dozen models that he uses.
As an example, here’s a couple of his approaches.
One that I like is the Ken Pomeroy Pythagarean model (no surprise there huh - it won me some nice money in a pool a few years ago). Here’s Tiernan’s descriptor.
Pomeroy Pythag model - Ken Pomeroy’s possession-based statistics have proven to be solid indicators of tourney overachievement. Pythag is a combined calculation of offensive and defensive efficiency. It has shown itself to be as strong as scoring margin in identifying tourney overachievers. With this in mind, I figured we do one bracket where we simply advance the team with the higher Pythag all the way to the championship. In 2008, this strategy was in the 99th percentile in the ESPN Tourney Challenge. Last year was nearly as good.
Here’s another he call’s Factors PASE.
Factors PASE model - Advance one and two seeds two rounds and three and four seeds one round. Then award each team the PASE value for any of the top ten attributes in KPI article. For the remaining matchups, advance the team with the higher cumulative PASE attribute value. Ties go to lower seeds in upset games. To resolve ties in toss-up games, give the nod to the team that possesses the most “top-three” attributes. If teams are still tied, opt for the higher seed.
This isn’t as immediately obvious as the Pomeroy model, but it gets to the heart of Kiernan’s approach. He’s got a list of about 15 “factors” that help distinguish successful tournament teams from the pretenders. They are things like:
Did they go dancing last year? No team goes from not being in the tournament to the champ the very next year.
How many tourney trips has the coach made? Coaches don’t usually win it their first trip either.
What percentage of points do they get from guards? Greater than 65% is not good.
… and so on.
PASE is what he calls Performance Against Seed Expectation. Tiernan tracks teams, coaches, and conferences to see how they historically perform in the tournament. Have they won more games than they would be expected to according to how they have been seeded in the past.
His models use PASE, key indicators, and Pomeroy numbers differently to come up with different projections.
It’s a pay site but there’s also lots of worthwhile free reading for geeks like me.