A New Voting System for the Hall of Fame

The results of the Hall of Fame balloting this year were followed by a now predictable round of articles criticizing the Hall of Fame voting process. The apparent decision by some voters to withhold support for Jeff Bagwell because of rumors linking him to steroids, the unexplainably, to some, poor performance on the ballot on players with sterling statistical credentials such as Tim Raines, the absence of any clear policy on steroid use in general and the quick departure from the ballot of candidates for whom a decent Hall of Fame argument could be made, such as John Franco and Kevin Brown were just some of the topics covered in the blogosphere after the vote. This year there was less dissatisfaction than usual because the two players who were voted in, Bert Blyleven and Roberto Alomar, were relatively non-controversial selections around whose election a solid consensus exists.

Every few years the Hall of Fame seeks to address some of these issues, usually by adjusting the process for electing people outside of the annual ballot through changes to the veteran’s committee and the like. However, the method for voting on the players in the annual ballot has remained largely unchanged since the early days of the Hall of Fame. The requirement that a player must get 75% of the votes cast to win induction is something of an unusual voting system which differs from the BBWAA voting system for baseball awards, which is done through rank order voting and the fan’s vote for the All Star Game which is based on a simple plurality.

The reasoning behind the system would seem to be that by setting a threshold at 75%, only players around whom there is general agreement will win election to the Hall of Fame. Additionally, this electoral structure makes it possible for nobody to get elected in years there are no players who are very strong Hall of Fame candidates, as will likely be the case next year.

There are, however, some disadvantages to this system. The first is that all votes are equal. There is no way to rank one player higher or to cast a vote that sends any kind of nuanced message. A vote for a player of historic greatness, such as Ken Griffey Jr. in a few years cannot be differentiated from a vote for a player who the voter simply thinks should be remain under consideration, such as Brown or John Olerud this year. The current system captures preference not intensity at the individual level, but ultimately seeks to measure intensity of support at the aggregate level. One result of this conundrum is that many voters vote for far fewer than the maximum of ten players because they are wary of treating a borderline candidate the same way as all time great. This leads to fewer inductees, which may be fine, but also to more players being dropped from the ballot without sufficient consideration, which is more of a problem.

One way to solve this problem would be for voters to rank order their preferences and award points based on how many voters each candidate receives in each of the first ten slots, and candidates who receive above a fixed number of points winning election. It seems that this idea might work, because it would make it possible to distinguish between the best candidates and the others, but it would also force candidates to rank players which could create problems at the top and bottom of the ballot as they seek to determine whether, for example Alomar was better than Blyleven, or Alan Trammel was better than Edgar Martinez. This system treats differences between rankings as all being equal, even though some voters feel very strongly about, perhaps the top three candidates and are less certain about the rest of their ballot.

Any new voting system for the Hall of Fame should, like the current system, make it possible for any number of candidates from zero to ten or so to be elected in any given year, so not all voting systems would work. There is one, albeit somewhat complicated, system that would still make it possible for no, one or some candidates to get elected every year, while giving voters the option to vote for as many players as they wanted and which would reflect intensity of preference. This system would be to give each voter, a fixed number, say 100, votes which could either be concentrated all on one candidate, divided evenly, or unevenly, between two or more candidates, or not cast at all. Those candidates who win a fixed number of votes, perhaps 25% would be inducted, while those receiving a significantly smaller number of voters, perhaps as few as 3% would remain on the ballot. Determining what these thresholds should be is difficult, so these numbers are just examples.

This system would allow a voter who believes only the very best players should be inducted to concentrate his or her votes on the top candidate or candidates, if there are any on the ballot. It also allows voters who want to have some more time to consider candidates, vote in a way that reflects that. It also makes it possible for a voter to award more points to candidates who they believe are the most deserving. A ballot for this year that reflects this might have awarded 25 points each to Blyleven and Alomar, ten points each to Bagwell, Trammell and Barry Larkin and five points each to Franco, Brown, Olerud and Martinez. This is, of course, just an example of one possible ballot but it demonstrates the versatility of this electoral system.

There are two obvious critiques of this system, but neither of them is particularly strong. The first is that this system is easily gamed. Conniving voters could cast most of their votes not for the best candidates, but for those who they believe are likely to be overlooked while assuming the best candidates will get enough support and get elected anyway. This is clearly possible, but every electoral system can be gamed in one way or another. In the current Hall of Fame voting system, voters can similarly leave the strongest candidates off their ballots to vote for ten players they believe are in danger of being overlooked. The growing trend towards publicizing Hall of Fame ballots makes gaming the system more difficult, or at least not without cost.

The second critique is that this system is new and a little strange. There is not much that can be said in defense against such a comment, because the system is new and a little unusual. However, the current Hall of Fame voting system is also a somewhat unusual. This proposed reform would require more from the voters, but they could more or less opt out of it by dividing their votes evenly among a small number of candidates. Reforming the Hall of Fame voting system will not resolve questions about how to treat suspicion or knowledge of steroid use or what really defines a Hall of Famer, but it would give voters more tools with which to answer these types of questions.