February 25, 2003
STATLG-L Hall of Fame Results
Sorting the Ballots
Well, sort of. The voting patterns on the two ballots (Players and Composite) were rather similar in some respects. On both ballots, only one person received the support of as much as half of the voters. On both ballots, the average voter cast votes for only a small number of candidates. On both ballots, nearly half of the candidates were able to attract the votes of fewer than 10% of the IHOF voters.
It turned out to be a pretty good day for Chicago at the top of the Players ballot. If the Hall of Famers cast their ballots in a manner similar to what our 1,789 participants did, the long, long wait is finally over for Ron Santo. The great Cubbie third baseman made it past the 75% plateau with 42 votes to spare; he was named on just over 77% of the ballots. With just over 40% of the vote, Minnie Minoso, who spent much of his career playing for the White Sox, finished a distant second to Santo. The only other man to garner as much as one-third of the votes on the Players ballot was Dick Allen, who spent three years on the South Side (among them, his 1972 MVP season). As a long-suffering Phillies phan, however, I will forever remember him in red pinstripes.
The complete tally on the Players ballot is displayed below:
Player Votes Percent Ron Santo 1384 77.4% Minnie Minoso 731 40.9% Dick Allen 638 35.7% Joe Torre 559 31.2% Gil Hodges 394 22.0% Tony Oliva 388 21.7% Curt Flood 361 20.2% Roger Maris 353 19.7% Joe Gordon 302 16.9% Carl Mays 262 14.6% Maury Wills 225 12.6% Ken Boyer 213 11.9% Bobby Bonds 213 11.9% Thurman Munson 179 10.0% Don Newcombe 139 7.8% Wes Ferrell 135 7.5% Vada Pinson 128 7.2% Mickey Lolich 100 5.6% Elston Howard 91 5.1% Rocky Colavito 88 4.9% Mike G. Marshall 87 4.9% Ted Kluszewski 84 4.7% Allie Reynolds 74 4.1% Marty Marion 54 3.0% Ken R. Williams 43 2.4% Bob Meusel 36 2.0% TOTAL 7261 Total Ballots Cast: 1789 Votes Per Ballot: 4.06What strikes me immediately is the very last number in the table--the average voter chose just over four players. That's a good bit lower than the typical voter in our simulated BBWAA balloting, where the average has ranged between 5.2 and 6.5 in the last three years. Another observation is that the distribution of vote percentages is much narrower than that of a typical IHOF simulation of the BBWAA ballot. In the just-completed version, for example, nearly 40% of the candidates (13/33) received fewer than the two percent of the vote that Bob Meusel picked up.
It's possible that the low votes/ballot count results in part from the voters' unfamiliarity with these players. I don't know the demographics of the BP readership, much less the subset of that readership who took the time to cast ballots, but I think I'm pretty safe in assuming that they're a good bit younger than I am (I was born near the end of the season of the Whiz Kids). I have definitive memories of only 17 of the 26 men on the Players ballot, and I bet that most of our electorate remembers the on-field exploits of far fewer of them. To test whether there was any sort of association between vote totals and the time during which a player was active, I recorded each man's "best" full season, using OPS+ for non-pitchers and ERA+ for pitchers. For instance, Santo's best OPS+ season was 1964, while Minoso's was 1954 and Allen's was 1972. Plotting vote percentage against best season and applying a linear trend line to the scatterplot, we find: The farthest-left point is Carl Mays, whose best season was 1917, while Bobby Bonds had the most recent peak season, in 1975. Although the slope of the trendline is positive--indicating that the vote percentages increase for more recent player peaks - the association isn't particularly impressive. By the way, that one outlier observation (Santo's 77% vote) doesn't affect the relationship all that much. In fact, without that one point, the R-squared value is actually slightly higher, 0.08.
For the record, I was a fairly typical IHOF voter, in that my ballot had checkmarks next to four names. As it happens, my choices were the first four finishers in the overall race; I had voted for them on STATLG-L ballots when they were still eligible for the BBWAA vote, and saw no reason to deny them on this one.
Consisting of managers, baseball executives, and an umpire, the Composite ballot is difficult to characterize. It was evidently also difficult for the IHOF voters to choose among the candidates. No one, not even Marvin Miller, was supported by 75% of the 1,183 voters. Miller fell well short of the 888 votes that it would have taken to be elected in our version of the Composite ballot, missing the mark by 69 votes. The entire Composite ballot looks like this:
Name Votes Percent Marvin Miller 819 69.2% Whitey Herzog 543 45.9% Charles O. Finley 475 40.2% Billy Martin 413 34.9% Walter O'Malley 408 34.5% Doug Harvey 247 20.9% Dick Williams 243 20.5% Bowie Kuhn 134 11.3% August Busch Jr. 116 9.8% Phil Wrigley 110 9.3% Bill White 93 7.9% Paul Richards 82 6.9% Buzzie Bavasi 77 6.5% Harry Dalton 42 3.6% Gabe Paul 42 3.6% TOTAL 3844 Total Ballots Cast: 1183 Votes Per Ballot: 3.25Behind Miller, the rest of the candidates arrayed themselves in several distinct groups. Herzog, Finley, Martin, and O'Malley received fairly significant levels of support. After them come the pair of Harvey and Williams; I wouldn't be surprised if nearly all of Harvey's voters are longtime National League fans. The other half of the field didn't draw very much support at all.
Like the Players ballot, the distribution of vote percentages on the Composite ballot was appreciably narrower than that of a typical BBWAA ballot. None of these men are what might be termed "courtesy candidates," like so many of those who go one-and-out on the BBWAA ballot. Another similarity to the Player ballot is the low number of names selected by the average voter. In fact, the mean number of names on a ballot (3.25) was even lower than we saw on the Players ballot. This phenomenon is indicative of the difficulty of assessing what it takes to reach the level of "Hall of Fame performance" in a non-playing role. We don't instinctively follow career managerial records (did you know, for instance, that Connie Mack was below .500 for his career?), much less whatever it is that general managers do. And umpires have to be assessed almost entirely on the basis of such non-empirical measures as "respect" and "authority."
As with the Players ballot, my own actions as a Composite voter were quite typical; I checked off three names on my ballot. From my comments above, two of those names should come as no surprise--in my opinion, Marvin Miller was the single most influential and important person in baseball over the last one-third of the 20th century, and Doug Harvey was unequivocally the best and most impressive umpire I have ever seen. My third vote went to Dick Williams, for reasons that I'm not sure I recall completely. I do recall him as a "player's manager," whatever that means, and of course he led the Impossible Dream BoSox in 1967. On the other hand, I might be partially conflating him with Davey Johnson.
What bearing these results might have on the actual results of the new Veterans Committee is anyone's guess. The conventional wisdom has been that it will be difficult for anyone to be elected to the Hall under this new process. Most of the 84 VC voters are older than the typical BP reader, and older than I am. They may remember the oldtime players from their own youth, and many will have played against the candidates who were active in the 1960s and 1970s. For every Ozzie Smith, George Brett, or Kirby Puckett who prospered under Marvin Miller's leadership of the MLBPA, there are many bitter old Bob Fellers who resent what Miller did "to" the game.
All in all, I don't think our vote can really help us to see into the minds of the HOFers voting on these two ballots. I'm still eager to see what their decision will be; our results might turn out to be completely off-base in comparison to the only votes that really count, or we might turn out to have done things similar to what they do. Or somewhere in between. Who knows?
We'll get another opportunity to mimic the Veterans Committee in two years. That vote will cover only the beyond-BBWAA players. The Composite ballot is planned as a quadrennial event, which means that we won't get another crack at them until 2007. In the interim, I hope the HOF will take the opportunity to examine the entire process they used to reach this point, and make changes as needed to produce a smoother and more reasoned methodology.
One final note: I have it on good authority that the STATLG-L Internet Hall of Fame balloting, in all its forms, will remain a part of the free content on the baseballprospectus.com website. You won't have to pay for the privilege of voting--which pleases me no end, as I've always wanted this project to be as open and available and widely-used as possible.
Neal Traven is the co-chair of the Statistical Analysis Committee of the Society for American Baseball Research (SABR).
To subscribe to STATLG-L, either (a) send a one-line email message consisting of: SUBSCRIBE STATLG-L [your name] to the address firstname.lastname@example.org, or (b) click here, and fill in the boxes. Either way, confirm your subscription and then join the discussion!