FAME and the Hall of Fame

A few months ago, my colleague Joel and I introduced a new statistic called FAME (Fanfare and Acclaim Metric Extraordinaire) to measure not how good a player was, but how highly he was thought of during his playing days. In my first article, I compared the FAME results of great players to their WAR. The results: Yogi Berra was the most overrated player of the past eighty years, and Tony Phillips the most unsung. (After expanding my numbers for today’s article, Berra remains atop the leaderboard. Second place, amazingly, belongs to Manny Trillo.)

Today, my inquiries center on the upcoming 2013 Hall of Fame ballot. Having read the first seventeen pages of The Signal and the Noise, I believe I’m ready to turn my mental powers toward the art of prognostication. My goal: to try to predict the outcome for the first-year eligible hitter’s on this year’s ballot.

The natural first step would be to use each hitter’s WAR to estimate their chances at Cooeprstown. Remember that FAME only concerns itself with position players, because pitchers don’t get enough awards to quantify them properly, and thus will be omitted. Here’s a graph that correlates the percentage of every first-ballot hitter from 1993-2012 with their career WAR:

The scatterplot gives us an R-squared of 0.52, meaning that we get a fairly strong (but by no means definitive) correlation. However, let’s look at the same batters, when comparing their vote totals to their FAME rating:

The correlation is somewhat stronger. In other words: lazy writers are lazy, and they tend to vote the same people for the Hall that they do for MVP. This is perhaps not altogether shocking.

Included now is a list of all first-year eligible hitters, including their fWAR and FAME ratings (click to embiggen):

Of course, there is the steroid factor, which basically ruins any attempt to predict the fates of Barry Bonds and Sammy Sosa (although less so for Rafael Palmeiro). But for those who are relatively free of suspicion and with reasonable election arguments, FAME and regression give us the following baselines for vote totals:

Mike Piazza: 58%
Craig Biggio: 30%
Kenny Lofton: 15%

Given the inner turmoil of the BBWAA, and the glut of quality candidates on the ballot, these numbers are likely to be somewhat inflated. Lofton in particular looks dangerously close to the Whitaker Zone, his case never to be truly heard.

Finally, every once in a while there seems to be a guy who collects multiple votes for no discernible reason: Hal Morris (5 votes), Vinny Castilla (7 votes) and Ozzie Guillen (5 votes) each received more acclaim at the end of their careers than John Olerud and his 61.3 WAR. If you’re looking for a candidate this year, look no further than Sandy Alomar, Jr. Alomar’s FAME Ratio comes out at 1.656, the third highest ever of any hitter I’ve bothered to calculate. Alomar went to the All-Star game six times, including a season in 1991 where at the end of June he was hitting .200/.232/.233 and was easily below replacement level.

I’ll conclude with my final predictions for the vote percentages of the first-year hitters:

Mike Piazza: 46%
Barry Bonds: 37%
Craig Biggio: 24%
Sammy Sosa: 12%
Kenny Lofton: 7%
Sandy Alomar: 1% (5 votes)
Shawn Green: 1% (3 votes)
Steve Finley: 1% (1 vote)
Julio Franco: 1% (1 vote)
All Others: 0%

Feel free to ridicule these numbers or provide your own predictions in the comments.