The Experts

Espn's Power16 has even the skeptic's in the voting (Patick Forde) picking us a number 1. http://sports.espn.go.com/ncb/news/story?id=1983208.

I like to do these little arithmetic of probability exercises. Assuming each game is an independent event (not a bad assumption) and that the probability of winning any one game is the same across games (a horrible assumption but I can't do the arithmetic otherwise where the quality of the other team and home or way matters) then what probability of winning a single game makes it and even money event that we would win 22 games in a row. If that probability is "p" then it solves the equation: p^^22 =0.5. Using Excel, I got to three decimal places p = 0.969 That seems like a high probability to me even if we're playing schools with RPI less than 100. This brings to mind Steven Jay Gould's article about Joe D's hitting streak. http://www.nybooks.com/articles/4337. Are the Illini hot (meaning the independence assumption is wrong and that prior wins make subsequent wins more likely)? Or are the Illini just really good? Gould says the data don't support getting hot in most cases where the expression is applied (by sportscasters) but the idea that performance depends on confidence and confidence depends on prior success at least suggests an obvious mechanism.