Bayesball -- Stats Amazing

The name of the paper is amazing enough to warrant reading:

Bayesball --A Bayesian hierarchical model for evaluating fielding in major league baseball

When you combine two of my favourite things in this world -- sports and statistics -- I am a happy hedgehog. If you can add food and a 20-hour sleep in there as well, I am a veritable purring kitten of joy.

The introduction may confuse some but basically all they are saying is that evaluating fielding in baseball, the success and failure of fielders, is subjective and through using a mathematical model based on Bayes Theory.

Again, simply put, this method is: "based on the assumption that probability is operationalised as a degree of belief, and not a frequency as is done in classical, or frequentist, statistics."
(Original Source: The Handbook of Market Research: Uses, Misuses and Future Advances)

The authors Shane T. Jensen, Kenneth E. Shirley and Abraham J. Wyner of the University of Pennsylvania discuss at length that

"a player’s fielding ability is more difficult to evaluate, because fielding is a nondiscrete aspect of the game",

and that;

"an inherently complicated aspect of fielding analysis is assessing the blame for an unsuccessful fielding play."


Anyway, I won't ruin it by telling you anymore. It is a good read for all those baseball fiends out there who want to trump their mates at the pub next time they bring up a fielding error.