Bert Blyleven Got In Because Voters Are Irrational, And Other Hall Of Fame Revelations (Follow-up from Deadspin Post)

By David Roher

We’re very excited to announce a new weekly-ish column for Deadspin called “Regressions.” I asked for “The John Salley Story Corner,” but it was apparently taken. The pieces won’t be too different from the ones you see here. Thanks to everyone at Deadspin, especially Tommy Craggs, for giving us the opportunity. As always, if you have any post ideas, feel free to shoot us an email or leave a comment.

I tackled the first column, an analysis of Baseball Hall of Fame voting habits that was an extension of last year’s post. We’ll be cross-posting all of our articles to the HSAC blog, so subscribers won’t miss any. Sometimes we’ll just reprint the post as is, but in many cases there will be an addendum based on comments and responses. Also, if relevant, we’ll go deeper into the math or provide the data.

The first piece, and a short follow-up, are after the jump.

The conventional wisdom is that Bert Blyleven’s Hall of Fame election represents a victory for rational thought, that at last more than 75 percent of voters were swayed by the righteousness of the smart people’s cause. While there’s plenty of truth to that, I think there was another factor in play. Blyleven didn’t get in only because baseball writers started thinking rationally; in large part, he got in because they’re consistent about voting irrationally.

The issue is a fake lower limit on vote totals that BBWAA members seem to impose upon themselves. Last year, curious as to why so many players had received underwhelming totals, I found a weird trend: the average number of votes doesn’t move much, regardless of the ballot’s quality. What follows is a list of the average number of votes per voter in the last 11 years. (Remember: a voter can select up to 10 players.) In parentheses are the players inducted that year (first-balloters are starred) and the number of first-balloters who received more than 15 percent but weren’t elected. You can get a sense of the relative ballot strengths this way: a year with no competitive first-balloters or obvious holdovers is weaker than the year before.

2011: 5.98 votes per voter (Alomar and Blyleven elected, 2 strong but unelected first-years remaining)
2010: 5.67 (Dawson, 4)
2009: 5.38 (Henderson* and Rice, 0)
2008: 5.35 (Gossage, 1)
2007: 6.58 (Ripken* and Gwynn*, 1)
2006: 5.64 (Sutter, 0)
2005: 6.32 (Boggs* and Sandberg, 0)
2004: 6.55 (Molitor* and Eckersley*, 0)
2003: 6.60 (Murray* and Carter, 2)
2002: 5.95 (Smith*, 2)
2001: 6.32 (Winfield* and Puckett*, 1)

Since you can’t vote for a single player more than once, you’d expect a deep ballot (2010) to draw more votes than a top-heavy ballot with just one great player (2002). You’d be wrong. The only thing that appears to change the average total significantly is the appearance of a new, no-doubt guy on the ballot. This in no way suggests that all voters keep their total the same each year, but it might indicate that more than a few keep their total within a cozy low and high.

The voting tic may just explain some of the nonsensical voting trends we’ve seen. Take the 2006 ballot, which had no compelling new players. My guess is that a lot of BBWAA members, used to voting for three-four or six-seven players in the five strong years previous, voted for the same number that year. As a result, Bert Blyleven (up 13 percentage points over 2005), Bruce Sutter (10 points and elected), Goose Gossage (nine points), and Andre Dawson (nine points) all got a big bump and thus the illusory but all-important “momentum” needed to reach the Hall, even though they had all been on the ballot for years. It’s worth noting that Blyleven’s biggest bump (14 points) came in 2008, another weak ballot and perhaps another instance of, “Well, I gotta use that sixth vote on someone.”

But what about when the opposite happens — when a lot of worthy new guys pop up on the ballot? That’s when we get the spoilers. Now, we shouldn’t have to talk about spoilers here. Unlike citizens in a presidential election, of whom only Chicagoans may vote more than once, the writers can pick up to 10 players. Given at least a decent consensus on qualifications, 10 should be more than enough to avoid one player’s interfering with another’s candidacy. But look at someone like Lee Smith, a guy with enough votes (45.3 percent this year, 47.3 last year) to make a difference for another player, even if his own case doesn’t appear to be improving. It’s impossible to say who played spoiler to whom. But it’s reasonable to think that a writer with rigid habits who felt compelled to vote for someone last year might’ve thrown his support to Smith and not, say, Blyleven or Alomar (or both, or all three). In other words, Smith might have had a lot to do with why Alomar and Blyleven had to wait one more year for induction. We can add election spoiler to the list of things Big Lee and Ralph Nader have in common, alongside plus fastballs and tireless consumer advocacy.

Or consider this year’s voting. Of the players on the ballot in 2011, there were probably around eight who had strong Hall of Fame support from some constituency, and maybe three or four others who still had decent support. That’s not a problem when the vote limit is 10. But when the limit is whatever a particular writer is used to, it gets tricky. Edgar Martinez, Fred McGriff, Mark McGwire, and Don Mattingly all saw their totals slip, and the new guys — Jeff Bagwell, Larry Walker, and Rafael Palmeiro — had disappointing figures as well. Harold Baines, Kevin Brown, and John Olerud, who at least have vaguely comparable statistical cases to some players still in the running, dropped off altogether.

Granted, vote total is far from the only factor influencing HOF voting. Some writers are more selective than others about who they’ll vote in, and there are an unusual number of solid-but-not-Ruthian cases right now. And of course, there’s that whole steroid thing. But I think the role PEDs are playing in the voting is amplified by the vote-total problem. BBWAA members are used to writing about the story, not making it. In a few cases, the explanatory narrative they construct around their voting is probably not the same as the actual thought process it took to make their ballots. I can’t help but think that if Bagwell had been the only potential Hall of Famer on the ballot, fewer writers would have brought their suspicions into his case. And unless significantly more people were turned off by McGwire’s apology, why did he lose votes?

Maybe this isn’t a major problem. After all, good cases aren’t shut down completely, only delayed. Next year, without any huge names coming up (Bernie Williams is the only one with any chance), the current crop should see a 2006-like boost. The year after, though, will offer an interesting litmus test: Bonds, Clemens, Sosa, Piazza, Schilling, and Biggio become eligible, and if voting habits don’t change, we might be in for a massive logjam. Nader help us all.


Follow-up: As Sky Kalkman pointed out, the piece doesn’t provide enough evidence to conclude that this is what voters are doing on a large scale. The data isn’t available — ideally, I’d want the individual ballots of every Hall of Fame voter. However, the means are close enough, and the inferences we can make about voting patterns are decent enough, that this problem is at the very least a strong possibility.

About the author


View all posts

Leave a Reply

Your email address will not be published. Required fields are marked *