Risk: High

Results: 2006 ? 74th percentile in the ESPN Tourney Challenge; 2007 ? 58th percentile; 2008 ? 14th percentile; 2009 ? 94th percentile; 2010 ? 99th percentile; 2011 ? 73rd percentile. OVERALL AVERAGE ? 68.7 percentile.

Strategy: Advance Nos. 1 and 2 seeds two rounds and Nos. 3 and 4 seeds one round. Then award (or punish) each team by the PASE value for any of the top and bottom five attributes in the ?Top Factors? series. Now, add or subtract the average of these PASE values to each team's expected seed win totals. At the round they're expected to reach, evaluate their Factor PASE against their opponent. Choose the team with the better PASE. To advance a lower seed over a higher seed in the first three rounds, the low seed must have a positive PASE while the high seed's is negative.

Outlook: This has been a reasonably strong model for the last three tourneys, scoring well even last year. A 73rd percentile result won't win any dances, but it was the second best model in 2011. I still rate this as a high risk, mainly because I don't automatically advance any seeds more than a single round. (Plus, there's that whole 14th percentile result in 2008.) I?ve tweaked the rules a bit, applying the factor PASE calculations to the top 14 seeds instead of just the top six...and I'm evaluating factors by seed class. We?ll see how that works.