A day off in Monte Carlo

Well, I'm taking a break from running today. However, we have the wonderful NCAA basketball tourney to keep us occupied. I've been doing some Monte Carlo simulations at work, and decided to see if a similar method could be used on our brackets.

It turns out, you get a decent correlation if you fit the average calculated point differential (for instance, sagarin rankings) , and historical win percentages for the different seeds.The parameters make good sense: 'a' should be close to one and 'b', the 'upset' factor (normally the temperature) is 2. Maybe next year I'll code-up the simulation and run a few million brackets to see who has the best chance of reaching the Final Four.