thoughts on the Mariners, MLB draft, and more homelinksdraftabout me

Just How Cramped is the 2015 Hall of Fame Ballot?

It's that magical time of the year where Hall of Fame voting occurs, which in recent memory has been the annual battle to get Edgar Martinez's incredible hitting accomplishments recognized. Edgar is my favorite player ever - a title he has a good chance of holding my whole life - so I am heavily biased. My love of him has probably only increased with the Hall of Fame battle he finds himself in. Others, such as the incomparable Jay Jaffe, bang the drum for Edgar every year, so I won't repeat my case for him.

However, Edgar's predicament has driven me to investigate the Hall of Fame in much more detail than ever before. This year's ballot is noted for how bloated it is, and I decided to see if it really is as clogged as many claim.

I started my investigation with a basic premise. There are two basic ways to get to the Hall of Fame: compile all-time great numbers over a long career, or without the counting stats, have an amazing peak that compares with the all-time greats. Career WAR can measure the compilers, and a simple best single-season WAR goes a long way towards describing a player's peak. Essentially, I approximated Jay Jaffe's JAWS stat by splitting it into its two core parts and treating those parts as an ordered pair. Below are the results:

Career bWAR vs. Single-season best bWAR
Current HOF (blue) and 2015 Candidates (red)

To start with, I was surprised at how compact the data is. I expected less of a relationship; in other words I expected more diversity (some players with mediocre peaks that played forever and some players without great careers that had absurd peaks). I do not know how much of this behavior is due to voting preference, but I would expect quite a bit of it. Anecdotally, different voters prefer different Hall of Fame sizes and different kinds of Hall of Fame careers, but a general consensus is forced by requiring 75% of ballots to name someone a Hall of Famer.

It's somewhat hard to see with the naked eye, but the relationship between season best and career WARs is more logarithmic than linear. That means career bWARs increase exponentially faster than single season bests with increasing larger bWAR totals. This phenomena is likely caused by a couple factors. First of all, it might be possible for a player to only be so phenomenal in one season, and all-time greats stay at that level longer than other players. This would result in a career bWAR that compiles every larger without creating a new standard for a season best bWAR. The second option is that Hall of Fame voters favor career accomplishments over phenomenal peaks at a predictable rate. I haven't looked into non-Hall of Famers, but that might be an interesting study to do at some point. My guess is that the relationship in this graph is caused more by the limits of how good a player can be than any voting bias.

Simple looking at the graph with a naked eye it's easy to identify four obvious Hall of Famers - Barry Bonds, Roger Clemens, Randy Johnson, and Pedro Martinez. Bonds and Clemens remain mired in PED purgatory, which creates the initial problem with this year's ballot. In normal circumstances Bonds and Clemens would have sailed into the Hall of Fame on the first ballot, but now their vote totals (or lack thereof) provide a barometer every year for how tolerant the voting bloc is of steroid use at the turn of the millenium.

Johnson and Martinez are newcomers to the ballot, neither with PED controversy around them, so they should theoretically sail through this year, emphasis theoretically.

The Big Unit's career, at least by peak and career WAR totals, is most similar to the likes of Tom Seaver, Lefty Grove, and Christy Mathewson, which is quite the trio. Mathewson went in as part of the first HoF class ever. Tom Seaver was also a first-ballot Hall-of-Famer and still holds the record for highest percentage of ballots (98.8%). Lefty Grove, though comparable to these pitchers, made it to the Hall in his fourth year of eligibility. So, who knows what kind of vote total Randy Johnson will end up with, but he's a strong candidate for enshrinement on his first try.

Pedro Martinez has two remarkably similar peers in the Hall of Fame already, Bob Gibson and Fergie Jenkins. Gibson went in on his first ballot with 84% of the vote, and Jenkins made it on his third try. It seems like many think Pedro is a given to make it on his first try, but I am not so certain. I agree that he should go in on his first try, but his case rests on a tremendous peak with nice, but not legendary, career totals. Neither Gibson nor Jenkins seem like borderline Hall of Famers at this point, but when they were voted on neither garnered phenomenal vote totals. Perhaps Pedro gets more votes than either of them because his iconic seasons came in the heat of the PED era, making him the first to get a bump from the era instead of a penalty, but Hall of Fame standards also keep escalating. I am curious to see Pedro's vote total, and I don't consider him a shoo-in to make it on his first try.

After that, there are lots of red dots on the chart well within the pack of "typical" Hall-of-Famers. Easily more than 10 dots, which speaks to the crowded ballot. It will simply be hard for players to get to the 75% line with so many viable candidates to choose from. So, theoretically, the players most "central" on the graph have the best chance - those who had both high peaks and lengthy careers. They have the kind of candidacies which should appeal to a broad enough swath of voters to cut through the backlog and make it into Cooperstown.

I found the median career bWAR for a Hall of Famer, and the median peak bWAR for a Hall of Famer, and then divided each individual player's WAR totals by the median to get a scaled score of how good they are relative to other Hall-of-Famers. Any number over 1 represents a player with a WAR total better than the Hall of Fame median. Below is the list of players who have above-average peak and career bWAR totals on this year's ballot, with their overall score (both peak and career scaled totals added) and their percentage on last year's ballot:

  1. Barry Bonds (4.1, 34.7%)
  2. Roger Clemens (3.8, 35.4%)
  3. Randy Johnson (3.0, N/A)
  4. Pedro Martinez (2.8, N/A)
  5. Larry Walker (2.4, 10.2%)
  6. Curt Schilling (2.4, 29.2%)
  7. Mike Mussina (2.4, 20.3%)
  8. Jeff Bagwell (2.3, 54.3%)
  9. Craig Biggio (2.2, 74.8%)
  10. Alan Trammell (2.2, 20.8%)
Only Pedro Martinez and Randy Johnson are newcomers to this list, which is mind-boggling when you think about it. These are players who had more productive careers AND better individual seasons than the CURRENT MEDIAN HALL-OF-FAMER. None of these 10 players should even be borderline cases - no matter a voter's preference for long-term success or a high peak - yet only two of the eight on this list that were on last year's ballot got over 50%. In fact, both Walker and Trammel seem likely to fall off the ballot altogether without sniffing enshrinement. Furthermore, only Bonds and Clemens have legitimate PED implications, though it seems that Jeff Bagwell has been lumped in with the PED crowd just because a bunch of voters feel like it.

It's lazy to blame the current glut on PED use. The steroid era certainly contributes, but the voting bloc has largely passed judgement already. Bonds and Clemens would already be Hall of Famers with well over 90% of the vote if not for steroid use. So, based on their vote totals, roughly 2/3 of voters won't even consider a steroid user no matter their career accomplishments - more than enough to keep any confirmed steroid user out of the Hall of Fame, and enough to have already cleansed the ballot of many PED users.

For fun, let's say that absolutely nobody voted for Bonds and Clemens on the 2015 ballot and instead put all their votes towards another player on the ballot. Only Biggio and Bagwell would get big enough boosts to make the Hall. Everyone else would still fall short,* even though nobody else with more than 20% of votes has PED suspicions in their past.

*Assuming that everyone who voted for Clemens also voted for Bonds, meaning there aren't 70% worth of votes to go around, but more like 35%. I would argue this is an extremely reasonable assumption.

The BBWAA, subconsciously, has turned into a club characterized by grumpy, cantankerous old men who I am convinced did not even watch the careers of most players on today's Hall of Fame ballots, and certainly does not bother to look at the production they accumulated over their illustrious careers. How else can some of these vote totals be explained? In what world do Mike Mussina and Curt Schilling have to fight to stay on the Hall of Fame ballot? Especially Schilling, with his Cy Young awards and fabled bloody sock to go with his Hall of Fame caliber stats. What exactly are 80% of these voters looking for?

As much as I want to argue in favor of Edgar Martinez, I am not sure I would even keep him on the Hall of Fame ballot with how crowded it is at the moment. The BBWAA, for whatever reason, has decided the Hall of Fame is pretty much closed. The current debates are pathetic and should not even be debates. There are blatant Hall-of-Famers who won't even sniff enshrinement this year, or the foreseeable future, until the complexion of the voting body changes. Good luck fixing the Hall of Fame with this group of voters.