Tuesday, July 5, 2011

Forgot to Announce This

Though I'm late on this, I've been in the habit of announcing presentations of things I have been working on recently. At the WEAI conference, I am a co-author on two presentations (one of which I have put together the majority of the analysis). Unfortunately, I was unable to get funding for WEAI because I am attending a bunch of other conferences this summer, including the Joint Statistical Meetings in Miami at the beginning of August. Anyway, here are some recent presentations (they were given by Dr. Rodney Fort and Dr. Jason Winfree, respectively). You can get the full Western Economic Association International conference program right here.

Attendance Time Series and Outcome Uncertainty in the NBA, NFL, and NHL
Brian Mills and Rodney Fort

Discrimination Among MLB Umpires
Scott Tainsky, Brian Mills and Jason Winfree

The first paper simply looks at the long-run stationarity of attendance in the three leagues and assesses--at a very simple level--the influence of competitive balance (playoff, game and consecutive season uncertainty) on these attendance levels. This is part of my dissertation, and there are a number of issues to be dealt with (not the least being the censoring issue for NFL sellouts). I think this paper might bore most of the readers here--unless you're really into Lagrange Multiplier statistics for a unit root with breakpoints.

I imagine that the latter paper would be of more interest to those here. I can't divulge the entire paper (or much of it really), but we tend to find that there is very little going on in the strike-calling data with respect to umpire race. The data go back through 1996 (I think), and I update the study with some Pitch F/X analysis. There's much to do, though.

In addition to these recent presentations, my fellow graduate student Steve Salaga and I will be presenting on Language-Based Discrimination in NHL Hall of Fame Voting at the Joint Statistical Meetings. There is a whole section on sports statistics there, with a presentation by Shane Jensen on fielding metrics. It sounds like a lot of nerdy fun. For this paper, we implement a technique called Random Forests (spoiler alert, we don't find any evidence in the analysis of discriminatory behavior). This is a parallel analysis to our forthcoming paper on MLB Hall Voting Discrimination in the Journal of Quantitative Analysis in Sports. When I know the issue, I will link it here. If anyone is dying to read it, let me know.

Lastly, I would encourage anyone interested in sports statistics to attend the New England Symposium on Statistics in Sports. For those interested in soccer (futball, football), there is a soccer analytics competition being run by StatDNA. The winner gets a trip to the conference to present their paper and a $500 prize. I am currently working on some things with some people you may know, but I won't be mentioning anything until later on. It's been fun.

Okay, off to get some work done. Sorry that I have been somewhat MIA of late. Been really bogged down with a lot of different projects. Hope to get back to sab-R-metrics soon.


  1. Brian, when you say "very little going on" with strike-calling and umpire race, does that mean you find even less than what Phil Birnbaum reduced the Hamermesh paper to, which was that maybe only one or two minority umpires are biased in favor of their own race?

  2. In the horse racing world there is a ton of information

    and different systems available to improve handicapping.

    Throughout years of reading and research the only true way

    to become very successful at handicapping boils down to 3

    key principles. With knowledge, the ability to adapt to

    different situations, and discipline to do the work

    spending time on your selections you will dominate any

    track you come across. There are so many factors that can

    affect any given race that some races you can look at a

    hundred different times and still be dumbfounded.