Saturday, February 17, 2007

Alternative Hall of Fame -- Methods (HoF Monitor)

Bill James created the Hall of Fame Monitor test in his "Baseball Abstract" to address an issue not addressed by the Hall of Fame Standards test. Whereas the Standards measured career performances, the Monitor evaluates how many individual seasons a player hit a certain benchmark, like a .300 batting average, 25 wins, or winning an MVP award. It does so for purposes of predicting whether a player will make the Hall of Fame based on what seems to impress the voters. For instance, James included number of seasons of 200 hits because that seems to impress the voters, even though it is essentially meaningless.

The Standards is a 100 point system with an average Hall of Famer getting 50. The Hall of Fame Monitor has no limit, but most players who hit 100 are likely Hall of Famers. A player's Hall of Fame Monitor score is reported on Baseball Reference.com.

I took the system as it is, with a few modifications. First, all of the numbers I plug into the Hall of Fame Monitor are normalized. And second, instead of measuring how many All-Star games a player actually played in, I evaluate how many times the player was worthy of being an All-Star by reference to his Win Shares and WARP1 scores (adjusted for season-length). Win Shares seasons of 25 or better qualify. WARP1 scores of 8.0 or better qualify. This makes up for the pre-All Star period, where players would be at a disadvantage in the system and for the fact that some players continue to be elected to the All-Star game well after they should be.

The last thing I did was develop a grade scale for each position. Like the Hall of Fame Standards, I evaluated the Hall of Fame Monitor scores for current Hall of Famers to develop the grade range. A player who gets an "F" might be a very good player -- he is not a failure relative to all baseball players -- but the "F" represents how he compares to solid Hall of Famers.

The Hall of Fame Monitor test gets normal weight in the GPA.