Another Project I'll Never Complete

Back in November, I had an idea to create a database of statistics for General Managers, linking win shares data to transactions data to general manager tenure data to help evaluate the performance of GMs based on whom they signed, traded for and traded away. Despite the helpful tips from folks like Simon Oliver Lockwood, that project never got off the ground, given that I have a full-time job and three young kids.

Yesterday, I thought of another project that I'll never have time for when I stumbled on Retrosheet's Umpire Data. While Retrosheet tells you how many games the umpire worked at each base, they do not relate the umpire's presence to the result of the game. You can get the daily game log for each umpire, but it would be better if we got each umpire's "record" per team. For example, from this page you can tell that the Nats were 2-4 when Paul Schreiber worked their games. Some interesting stats to run would be (1) the home team's record when the ump worked the game; (2) whether teams fared better or worse when an umpire was behind the plate or in the field; and (3) number of strikeouts and walks in games when the umpire was behind the plate. Wouldn't it be interesting to find out that some teams performed better with certain umps? In the early 1990s, Barry Jacobs's Fans' Guide to the ACC use to publish data like this for all ACC referees -- it showed this Duke fan that we actually performed better with the villified Dick Paparo or Lenny Wirtz working the game.

I poked around Retrosheet but could not find data presented in this fashion. It does not seem like a difficult task to create a database from which these questions could be answered. If anyone knows of a place where this data can be found, I'd love to hear about it.