Welcome

I am a PhD candidate in the Department of Political Science at the University of Alabama. You can learn more about me on my Linkedin page.

Hockey

Ranking hockey fighters: I used user rankings of wins, losses, and draws for more than 5,000 fights from hockeyfights.com to rank hockey fighters.

Testing the Cherry Hypothesis: Don Cherry of Coach's Corner has said teams that fight more win more (e.g. here). Is that true? It appeared to be on February 18, 2012, the date of the above clip, but a better test would look at the hypothesis' predictive power for data Cherry has not observed (i.e. data after February 18, 2012). Below is a plot of points per game against fights per game, and it is updated daily to reflect data from the 2012-2013 season. I included a standard linear-regression line, and the equation near the top gives the number of points per game a team can expect given how many fights per game it has. '(S)' means fights per game is a statistically significant predictor (at .10 significance level), and '(NS)' means it is not.

College Football

I have rated all FBS football teams and their offenses and defenses using points. I also rated each FBS team's rush offense, rush defense, pass offense, and pass defense using yards per game (which are good predictors of points) through December 15, 2013. The results are here. The segments represent 95% confidence intervals, which are a measure of uncertainty. The narrower the confidence intervals, the more confident we can be in the estimate.