An open letter to Baseball-Reference and the statistical powers that be

To whom it may concern:

On the heels of a pair of great Baseball-Reference blog posts this past week ranking the best pitchers and position players not in the Hall of Fame based on their Wins Above Replacement data, I may have created a new baseball statistic and found another way to gauge worthiness for Cooperstown.

This statistic is called Hall of Fame +/- and it measures how many future Cooperstown members a ballplayer finished ahead of in Hall of Fame voting compared to how many fellow non-inductees got more votes than them, divided by the number of years they were on the ballot. As I’ll explain momentarily, it’s a great tool for discovering forgotten players. Also, it appears many recent Veterans Committee picks have positive Hall of Fame +/- ratios so my metric could be a good way for predicting future honorees. In fact, the old-guard players and managers of the committee could care less about modern sabermetrics like WAR, so this might predict more future picks.

Baseball-Reference helped with the creation. I recently learned it’s possible through the site to view the results of Hall of Fame voting by the Baseball Writers Association of America for every year dating back to the first vote in 1936. Looking at the results of the 1983 election for a post I did in April on one-and-done Hall of Fame candidates, I noticed Gil Hodges, whose Hall of Fame credentials I’ve looked at before, got more votes than six future Cooperstown members that year but wasn’t enshrined. In fact, Hodges exhausted his 15th and final year of BBWAA eligibility in 1983 and still doesn’t have a plaque.

For each year Hodges was on the ballot, he got more votes than an average of 9.67 men who were later enshrined. Only once, in his first year of eligibility, did anyone finish ahead of him in the voting who doesn’t have a Cooperstown plaque now. So, by taking the 145 times Hodges got more votes than a future Hall of Famer, subtracting the three non-members who beat him in 1969 and dividing by the 15 times Hodges was on the ballot, we get his Hall of Fame +/- of 9.47.

I wanted to see if this was an anomaly or the norm for the 32 other men besides Hodges who’ve gone the full 15 years and failed to make the Hall of Fame with the writers since the advent of modern voting procedures in the 1960s, men like Ron Santo, Roger Maris and Tommy John. Thus, I started going through the voting records and tabulating the Hall of Fame +/- for each player.

I didn’t look at all 32 others, but the ones I saw didn’t approach Hodges’ ratio. Maris, Santo, and John all have negative Hall of Fame +/- ratios– that is, the number of non-members who got more votes than them was higher than the number of future Hall of Famers they beat out. Bill Mazeroski never finished ahead of a future Hall of Famer in his 15 years on the ballot, though the Veterans Committee later enshrined him. Other committee picks like Jim Bunning and Red Schoendienst appear to have positive ratios at quick glance, though I haven’t calculated them yet, and I’m guessing the numbers are lower than Hodges’ ratio.

I dug through old ballots to find Hodges a peer. Some may argue it’s not a valid comparison, since the older the Hall of Fame voting year, the more time that’s transpired to allow a larger number of players to be honored by the writers and Veteran’s Committee. Old ballots also sometimes teemed with more than 100 players, including active stars and managers. That’s where the ratio comes in: It means little for a player to have finished better than 50 future Hall of Famers in 1938 if that many non-members finished in front of him. My stat rewards non-members who finished consistently better than other non-members.

Hodges doesn’t have the best Hall of Fame +/- ratio among all non-inducted players. I found three with better ratios: Lefty O’Doul with 13.8, and a pair of Deadball Era catchers, Hank Gowdy with 14.59 and Johnny Kling with 13.11. Hodges also doesn’t have the record for most future Hall of Famers beaten out on one ballot, even though he bested 13 in 1970. Gowdy, who I recently wrote belongs in a starting lineup of combat veterans, got more votes than 33 future members in 1956. Kling beat out 32 in 1937 and 31 the following year, while O’Doul, an amazing player a short time in my book, got more votes than 27 in 1960. Gowdy, Kling and O’Doul weren’t bested by any non-Hall of Famers those years, either.

I hadn’t heard of Kling prior to my research, and it illuminated others like Babe Adams, Duffy Lewis, and Bucky Walters. To me, that makes this stat valuable. If for no other reason, it could help honor forgotten players. With that said, there’s another reason I’d love to see this stat added to Baseball-Reference: All I’ve done this weekend, it seems, is pore over old Hall of Fame ballots.

Anyhow, with that, I’m out.

Yours truly,

Graham Womack

(Postscript: Not everyone liked this idea.)