January 10, 2013

Baseball Therapy

Lessons from the Hall of Fame Vote

by Russell A. Carleton

the archives are now free. All Baseball Prospectus Premium and Fantasy articles more than a year old are now free as a thank you to the entire Internet for making our work possible. Not a subscriber? Get exclusive content like this delivered hot to your inbox every weekday. Click here for more information on Baseball Prospectus subscriptions or use the buttons to the right to subscribe and get instant access to the best baseball content on the web.

Subscribe for $4.95 per month Recurring subscription - cancel anytime. Subscribe for $39.95 for one year a 33% savings over the monthly price! Purchase a $39.95 gift subscription a 33% savings over the monthly price! Already a subscriber? Click here and use the blue login bar to log in.

So... the Hall of Fame vote happened. And no one got in.

There have been plenty of threads of analysis and a few angry fingers being pointed to explain how an immensely talented class of potential inductees failed to produce anyone who will get a plaque in Cooperstown next year. The 10-person limit on the ballot is to blame. Steroids are to blame. The vague suspicion of knowing someone who might have at some point taken steroids is to blame. Aaron Sele is to blame. People not taking the process seriously are to blame. The Tea Party (huh?) is to blame. Kirk Radomski is to blame. The difficulty of rallying 75 percent of any group of people anywhere around one definition of "famous" is to blame. Jack Morris and Tim Raines are to blame. A golfing website is to blame. Bacne is to blame. The DH is to blame. The fact that Kid Rock once "wrote" a "song" called BBWAA (or something like that) is to blame. The statheads are to blame. The dinosaurs in the media are to blame. The space aliens are to blame. Bias against first-timers is to blame. Vitamin supplements are to blame.

Okay, so mostly it's been blaming. We get it. The system is broken, and there will be a series of conversations that take place over the next few days about that. But let's drill a little deeper on the Hall of Fame voting and see what lessons we might learn from this year's Cooperstown class:

1) Expanding the ballot might not have gotten anyone elected.
According to the BBWAA, there were 569 ballots submitted, and they contained a grand total of 3,756 names, meaning that each ballot contained an average of 6.6 names. Much of the discussion leading up to the vote centered around how many well-qualified candidates there were. There were plenty of people casting fake ballots who found themselves saying, "I would have voted for Todd Walker, but I ran out of room." The BBWAA voters (the ones who count) weren't so generous with filling out their papers all the way to the bottom.

As of this writing, 114 ballots (20 percent of the total) have been made public on the BBWAA website. These list an average of 7.0 names. Logically, this means that the private ballots contained an average of 6.5 names. Our public voters were a little more likely to write down an extra name, but still, there were an average of three blank spaces on each of their ballots. In addition, only one quarter (29 of the 114 public ballots) were completely full, listing 10 names. Recall that the ballots kept private had even fewer names on them. This means that there was even more extra space out there on those ballots that just went unused. If all ballots were made public, there's a good chance that the number of completely full ballots would actually go down.

Number of Names on Ballot	Percentage of Public Ballots
1-5	28.1%
6	15.8%
7	7.9%
8	12.3%
9	10.5%
10	25.4%

On top of that, there's little evidence that the top five vote-getters (Craig Biggio, Jack Morris, Jeff Bagwell, Mike Piazza, and Tim Raines) were done in by the 10-vote limit.

Let’s take Biggio (who came the closest to election at 68.2 percent) as an example. The average writer who voted for B-G-O (and who made that ballot public) wrote down 7.4 names. Among the writers who did not vote for Biggio, the number was 5.8.

Player	Avg. number of other names on ballots that included player	Avg. number of names on ballots that did not include player
Biggio	6.4	5.8
Morris	6.1	6.8
Bagwell	6.8	4.8
Piazza	7.0	4.8
Raines	6.8	5.5

When writers did not vote for Biggio or Bagwell or Piazza or Raines, the trend was not that they were generally voting for a bunch of other people and someone got crowded out, but that they were the types who just preferred a smaller Hall and may have been nitpicky with everyone. The notable exception here is Morris, who might have been squeezed out by the limited ballot, and even then you have to make some assumptions.

Among publicly available ballots that did not contain a vote for Morris (there were 50 such ballots), 10 were completely full. There were 40 public ballots that had space for the hero of 1991 World Series Game 7, but where the writer just... said... no. Morris missed induction by 42 votes. Since one in five ballots is public, and if we assume that the public ballots are a good proxy for the ballots as a whole, we might estimate that there were 50 total ballots on which Morris was not listed, but there were 10 entries (again, I'm guessing that the number is probably lower; see above). If that's right, 84 percent of those full-ballot, no-to-Morris people would have to say "I would have voted for him, but I ran out of room" for Morris and his mustache to get a plaque.

The reverse pattern that Morris shows compared to the other four also underscores the polarizing nature of his candidacy. Morris's supporters prefer a smaller hall than his non-supporters, and also a smaller hall compared to supporters of the other four. This suggests that while his supporters view him as closer to the inner circle, his non-supporters are bigger hall people (6.8 names among his non-supporters) a full unit above the next closest (Biggio), and thus consider Morris not good even good enough to get into a Big Hall.

The other four top vote-getters fare even worse. There were only three full ballots that didn't include Biggio: three for Bagwell, two for Piazza, and five for Raines. Assuming that we can multiply those numbers by five (again, a generous assumption), and that they were the 11th or 12th choice of all of those voters, Biggio hits 403 votes, while no one else would break 360.

The very likely reason that no one was elected had more to do with either the Small Hall tendencies of the voters or the fact that, for some reason (perhaps several reasons) 75 percent of them could not agree that any of these players were Hall of Fame material.

The people who may have been squeezed out by the 10-man ballot were Kenny Lofton and Bernie Williams, both of whom failed to clear the five-percent support necessary to keep them on the ballot. They were always going to need Big Hall voters to help them stay up, and many of the roughly 100 projected full ballots were (somewhat necessarily) the product of Big Hall voters. Still, that means that about 350 writers had an extra space on their ballot and could have voted for either. A few extra votes would have kept them around. No such luck.

2) Steroids matter. A lot.
Performance Enhancing Drugs (PEDs) and the Hall of Fame are one of those sure-fire troll-baits that will start an argument. I'll sidestep the should-we-or-shouldn't-we issue for a moment and look at what the voters are telling us about what they think. Roger Clemens and Barry Bonds have both been accused of using PEDs during their playing days. If that taint were erased from the record (and assuming the same numbers), both gentlemen would be posing for the bas relief sculptor guy.

The record vote share is held by Tom Seaver and Nolan Ryan at 98.8 percent, followed by Cal Ripken (98.5 percent), Ty Cobb (98.2 percent), and George Brett (98.2 percent).

Wade Boggs got 91.9 percent. I think it's safe to say that Bonds and Clemens would have finished north of 90 percent under different circumstances, and they might have challenged for that 98 percent bracket. Bonds and Clemens got 36.2 percent and 37.6 percent of the vote, respectively. I'm guessing that most of that gap is attributable to anti-PED backlash.

Tellingly, of the 114 publicly available ballots, 54 had votes for Clemens and Bonds, 60 had votes for neither, and there were no split ballots. That Clemens scored eight more votes than Bonds suggests that there were some split ballots behind the scenes. Even more telling were the votes for fellow PED-linked players Mark McGwire (who admitted using) and Rafael Palmeiro (who tested positive during his playing days). There were no votes for either first baseman that did not also include a vote for both Clemens and Bonds. Clemens and Bonds had better numbers, so it's not surprising that they out-polled McGwire and Palmeiro, but the message seems pretty clear. The PED users are something of a bloc.

More interestingly, ballots containing a vote for Clemens and Bonds generally contained 6.6 other names. Those that did not contained an average of 5.4 names. Big hall voters tended to be more forgiving of PED allegations. Again, this is the same pattern that was observed above for the Great Polarizer, Jack Morris. So Clemens and Bonds (and presumably McGwire and Palmeiro) need to convince a group of voters who tend to like their hall smaller anyway... and who probably don't like them. And at that, they would need to convince roughly half of these people who didn't vote for them to have their hearts soften a bit before they could be enshrined.

3) The public/private ballot split has some fascinating cultural data hidden in it
We are what we do in secret. Take a look at this tabulation (stolen from here) of the percentages of public votes for a player vs. the percentages of private votes.

Public   Private Player
70%     67%     Craig Biggio
59%     72%     Jack Morris
59%     60%     Jeff Bagwell
60%     57%     Mike Piazza
60%     48%     Tim Raines
38%     53%     Lee Smith
39%     39%     Curt Schilling
44%     34%     Roger Clemens
45%     31%     Barry Bonds
36%     36%     Edgar Martinez
38%     31%     Alan Trammell
16%     24%     Larry Walker
20%     21%     Fred McGriff
19%     19%     Dale Murphy
14%     18%     Mark McGwire
9%       15%     Don Mattingly
13%     12%     Sammy Sosa
13%     7%       Rafael Palmeiro

Some voters, most of them active columnists such as our own John Perrotto, wrote columns in which they disclosed their ballots and generally explained themselves. It's an easy column to write on a January evening when... well, there's nothing else going on in baseball. (On the BBWAA public ballot page, where it's available, there's a link to some of those columns.)

There are some clear patterns here, and they seem to break along certain fault lines. Sabermetric darlings Tim Raines (by 12 points) and Alan Trammell (seven points) fared much better in publicly published ballots. "Old School" favorites, such as Jack Morris (13-point spread) and Lee Smith (whose candidacy is based mostly on his save total; 15 points) were much more beloved behind the scenes. Larry Walker, who is a test case for how much voters believe in park factors, has an eight-point spread in favor of the private ballot. There are plenty of guys who show a virtual tie between public and private ballots, including gents such as Biggio, Bagwell, and Piazza, who, in general, fare well both in traditional counting stats and new-age value stats.

Then there are the steroid-tainted cases. Roger Clemens got 10 points more support from public ballots than from private ballots, while Bonds had a 14-point split. Curiously, Mark McGwire did better behind the scenes, but Rafael Palmeiro was better loved in public.

In general, the voters who released their ballots have previously been shown to be Bigger Hall voters, but also more sympathetic to the cases beloved by sabermetricians but suspected by traditional stat adherents. Public releasers also appear to be more willing to forgive potential PED use.

Shortly after the vote, Fox Sports baseball guru Ken Rosenthal disparagingly compared the Sabermetric movement to the Tea Party. (I suppose pending your political leanings, that might be a compliment.) Rosenthal's point was that the level of discourse around Morris (and presumably other issues) has gotten rather icky. There's a certain part of me that looks at these numbers and wonders if the writers who kept their votes for Morris and Smith (and not for Raines and Trammell) private just didn't want to put up with the comments section on the column. I know that Rosenthal caught a lot of flak for his comments, but maybe there's something to it. This won't be comfortable for some people to hear, but as sabermetricians, if we're trying to convince people to actually listen to us, we can't make things so uncomfortable that they'd just rather not deal with us. That might not be what happened, but I think we have to consider the fact that it may have.

Then, there are the steroid splits. This seems like a separate issue. Being willing to stand up and say "Yeah, I know that they might have used, but I don't care" is apparently easier to say out loud than "They might have used and I don't want that in the Hall." However, five of eight voters didn't vote for Roger Clemens or Barry Bonds. For all the "live and let live" messaging that is going on in public, there is a strong and apparently silent majority that has its objections. Again, the writers may just not have wanted to have that argument. But make no mistake: all the impassioned pleas that "everyone was doing it" and requests to stop being holier-than-thou fell on a lot of ears that wouldn't hear it and mouths that would not discuss it.

The question this raises is what would happen if all ballots were public, as has been suggested by many. Would accused PED users get more support from writers who do not want to appear holier-than-thou and who would be shamed into voting for the steroid bloc? Would the saber/traditional divide widen or narrow, and in which direction? People appear to behave differently under cover of darkness (or perhaps those that behave differently choose to remain quiet).

The bigger message is that as the baseball community inevitably discusses rule changes in the voting process, keep in mind that even the most simple suggestions might not be the panacea that everyone is hoping for. And they might not have prevented the shutout that happened today.

Russell A. Carleton is an author of Baseball Prospectus. Follow @pizzacutter4
Click here to see Russell's other articles. You can contact Russell by clicking here