1
Hall of Fame Voting So Predictable, a Computer Can Do it
By Kevin Spak
|
Posted Jul 27, 09 7:37 AM CDT
|
Share
(Newser)
–
The Baseball Writers of America turn out to be a predictable lot. A new computer program from a Missouri State computer science professor can accurately predict who’ll join the Hall of Fame, the Wall Street Journal reports. Given a set of 1,592 players who retired between 1950 and 2002, it accurately identified 98.7% of those selected to be Hall of Famers.
The program doesn’t even use particularly arcane statistics to make its predictions. For pitchers, plain ol’ wins, saves, ERA, and winning percentage are all it worries about; for hitters, it considers hits, home runs, and OPS. All Star appearances count for both as well. But don’t worry, there’s still room for barroom debate: some cases, like Mike Mussina or Edgar Martinez, even the computer can’t guess, giving them a 50/50 shot.