How I Would Vote on the 2014 Hall of Fame Ballot

Today is something like the last day for Baseball Writers Association of America (BBWAA) members to submit their votes for 2014 induction into the MLB Hall of Fame. This year's ballot is perhaps a uniquely crowded one, because the past few years have seen far more deserving players enter the ballot than depart it through induction, mostly because people have been failing to vote for suspected steroid users or, in some cases, for anyone at all from the "Steroid Era." The names on the ballot can be seen here (note that that page will be updated with the actual vote tallies once those are released), along with their history on past ballots. Now, I obviously don't have a vote, but I have an opinion, so this post will say who I would vote for if I could vote for anyone. This is an unusually complicated question. Normally it's just a matter of looking over the names on the ballot and deciding for each of them whether they deserve induction. But pretty much everyone who isn't a hypocritical curmudgeon admits that there are more than 10 deserving players on this year's ballot, which is a problem since one can only vote for 10 players. So first I'll say which of the eligible players I think are deserving, and then I'll say who I'd actually vote for, as a strategic/game-theoretic matter.


Returning Players

Barry Bonds
Y'know. My philosophy about the Hall of Fame and various moral failings and/or cheating would be to put everyone in and then acknowledge their controversial aspects on their plaques. Maybe to devalue a bit the performance of just plain known cheaters. But Bonds minus steroids is an obvious Hall of Famer. With the steroids he's the best player ever; without them he's one of the best ever, no question. So I'd vote for him, in an unlimited world.
Roger Clemens
Same thing I said about Bonds. If you give him full credit for his actual performance, he's, well, one of the best pitchers ever, maybe the best by aggregate production. There's no reasonable steroid discount you can apply that makes him miss the cut. Obviously if you felt that any known users should be flatly excluded, which would I suppose follow the (terrible) precedents of excluding Shoeless Joe Jackson and Pete Rose. But I don't feel that way, so I'm a Clemens voter.
Curt Schilling
He got near 80 WAR. I kind of think I couldn't possibly think that someone who got 80 WAR in the post-integration era, at least, wasn't deserving of induction. There's no reason to think about applying a steroid discount to him, and one doesn't generally give these guys a "jerk discount," though Schilling is quite the jerk. He pitched for 20 years, had at least three Cy Young-worthy campaigns, was all kinds of dominant in the post-season, etc. He's in, clearly.
Jeff Bagwell
One of the best first basemen ever. The steroid standard that would keep him out, "flatly exclude anyone who played in the 90s or early 2000s who hit lots of home runs and had a very muscular physique," is just plain ridiculous. There is no question here.
Mike Piazza
He's in the group of the six greatest catchers all-time (along with Bench, Fisk, Berra, Carter, and Pudge Rodriguez). That, combined with the abject lack of any incriminating evidence against him (and the falseness of the charge that his power came out of nowhere, he had been an impressive power hitter since he was a teenager), makes this an easy case. Plus, he's my favorite player, so I'm not just voting for him, I'm going to end up casting a very high-priority vote for him.
Larry Walker
Coors Field is a funny thing. Hitters put up absurd numbers there. Since it opened, a Rockie has won the NL batting title eight times and the NL home run crown three times in twenty-one years. Early on, awards voters had a limited sense of just how much those numbers needed discounting, as in the time Dante Bichette finished second in the MVP voting despite having a below average season. Then they figured it out a bit better. But that doesn't mean no numbers anyone puts up there count for anything, and Larry Walker was always one of the best at actually producing at a dominant level adjusted for altitude. His unadjusted numbers are insane; his adjusted numbers are still Hall-worthy.
Craig Biggio
Biggio had 3000 hits. Arguably he shouldn't have, since he was a below-average player for the latter part of his career. But one way I like to evaluate players who hang on for too long is to look at their Wins Above Average peak, i.e. how high they got before falling off. Through the 2001 season Biggio was worth 35.4 WAA, eminently Hall-worthy. He also had 2149 hits. Gaining his next 911 hits cost him 6.8 WAA, but I think he should get credit for his high tide line. Also I do think that 3000 hits and 500 home runs are just categorical qualifiers, though I understand that that's kind of irrational. He's in.
Rafael Palmeiro
Palmeiro has both of those categorical milestones. Precisely three other people can say that, and they (Mays, Aaron, and Murray) are all in the Hall. Now, Palmeiro used steroids. Also, he's barely over each of those marks, and his WAR/WAA numbers are also marginal, so if one took the "discount" approach to steroids it would be quite defensible to exclude him. Accordingly I think he's quite a low-priority vote, but since I'm a pretty emphatic PED dove I'd vote for him, all else being equal.
Alan Trammell
Twenty years at shortstop being an above-league-average hitter. An incredibly well-rounded player, being above average at hitting, running, and fielding (at one of the toughest positions). It ain't flashy but it adds up, making Trammell exactly the kind of player sabermetrics is designed to highlight. It's interesting that either a worse fielder with stronger hitting stats or a worse hitter with more incredible defense, i.e. Ozzie Smith, would get noticed more than Trammell, holding overall career value constant. Perhaps someone with flashy speed, i.e. 600 or more steals, might also attract more notice. The very well-balanced players seem to slip through all the nets, unless you make a point of balancing all the various factors. Once you do that, Trammell is deserving.
Tim Raines
Raines was basically a more mortal version of Rickey Henderson. Both leadoff-hitting left fielders. While Rickey hit .279/.401/.419 with 1406 steals over 25 years, Raines hit .294/.385/.425 with 808 steals over 23 seasons. That's not as good, especially since Rickey had more playing time per season than Raines. But as Bill James said, if you split Rickey Henderson in two you'd have two Hall of Famers, and Tim Raines is one of them.
Edgar Martinez
He was mostly a DH. Lots of people seem not to want to vote for those. I don't like that the DH exists, but I think the voters kind of have to accept that it does. Also, WAR includes a positional adjustment that hits DHs pretty hard, so if a DH looks deserving by WAR, they probably are. And he is, with 68.3 WAR and 38.4 WAA (which included 40.8 WAA from his first full season through his penultimate year, excluding his dreadful last season). He was a .300/.400/.500 player, along with the following people since 1947: Joey Votto (so far), Todd Helton, Manny Ramirez, Larry Walker, Frank Thomas, Ted Williams, Chipper Jones, Albert Pujols, and Stan Musial. Yeah. There's almost no level of bad defense you can have to not be deserving if you were that good a hitter.
Mark McGwire
Anyone who wants to apply a steroid discount should not vote for McGwire. The only reason one would vote for him, I think, is his 583 home runs. Take a good chunk of those away and he's Dave Kingman. So again, he's low-priority for me, but I think if I were the Imperator of the Hall, I'd put him in, warts and all.
Sammy Sosa
See above. If anything he was more of a one-tool player than McGwire.
Lee Smith
The same way I think those of us who don't like the DH can't punish those who played the position the best of anyone, those who don't think modern relief pitching is particularly valuable should, I think, vote for the best relievers anyway. I tend to think that roughly ten relievers should be in the Hall right now, and Smith is clearly one of the top-10 relievers ever. Also, his 29.6 bWAR is unimpressive, but if you take his 13.8 WAA and multiply it by his 1.8 career leverage index, which I feel is appropriate for late-game relievers, you get 24.8 WAA*, which is not exactly the world's most exemplary figure but is at least borderline. If we made relievers compete equally with starters for every spot in the Hall, Mariano Rivera would barely make it in; as soon as we acknowledge that short relief is a legitimate position of its own, Smith is an obvious pick.

Uncertain
Fred McGriff
On the one hand, he wasn't that good. 52.6 WAR, 19.8 WAA. My concept of peak WAA isn't even that helpful; he peaked at 21.7 after his age-30 season and spent the next ten years as a slightly-below-average player. And he doesn't meet any of the automatic qualifiers. But on the other hand, he comes damn close, with 493. There's a part of me that thinks it's cruel that, had he held on long enough to hit seven more bombs, I'd think him eligible. There's another part of me that thinks that a 500-HR cutoff isn't really a 500-HR cutoff if 493 is close enough. So I'm not sure whether I think he deserves to be in. That makes him a very low priority vote, I guess.

New Players

Greg Maddux
355 wins, 5000 innings pitched with a 3.12 ERA that included the height of the Steroid Era. There is no coherent argument against Greg Maddux's Hall of Fame candidacy.
Tom Glavine
305 wins. Yeah, okay, I know, wins ain't everything, but as with the hits and home runs milestones I think I believe that 300 wins should be an absolute cutoff. I don't think I know of anyone that rule gets wrong, anyway, and Glavine certainly isn't one of them. He was apparently someone who overperformed his decontextualized, defense-independent numbers (i.e., his strikeout, walk, and home run rates), but for retrospective evaluation as opposed to projection I think there's no good argument for using FIP instead of ERA, maybe plus or minus some adjustment for defense. And when you look at his raw run prevention, he's up over 80 WAR, which, as I said before, is just plain enough to convince me.
Mike Mussina
Mussina did not get to 300 wins. That's okay. He pitched over 3500 innings with an ERA+ of 123. It adds up to 82.7 bWAR and 48.6 WAA. That latter is a figure I just find incredibly hard to overlook, in this day and age at least. He retired after a 20-win, 34-start, 200-inning, 3.37-ERA season. During his last five years with the Yankees, he averaged 14 wins per season, so if he had felt like hanging around for another three years, he almost certainly could've gotten to 300. I don't think players should be penalized for the damage their WAR/WAA stats take if they hang around to get a milestone, but I also don't think they should be penalized for failing to get a milestone by not hanging around. Oh, and he should go in as an Oriole, and that's not just out of spite.
Frank Thomas
Remember that thing I said about how Edgar Martinez was one of the very few .300/.400/.500 hitters of the integrated age? Yeah. .301/.419/.555, over 19 seasons, with 521 home runs. He was primarily a DH, but that's okay, because he was just so good at hitting. If you had to designate someone to hit, he was an awfully good choice. I think it's fair to say that literally the only argument against Frank Thomas, unless someone wants to start suspecting him of steroids, is that he was a DH (and a bad first baseman when he tried being one). WAR, as previously stated, takes that into account, though, and even with -246 defensive runs he comes out to 73.6 bWAR and 39.1 WAA. Good enough for me.

Uncertain
Jeff Kent
He was surprisingly good. .290/.356/.500 is pretty impressive for a second baseman, as are 560 doubles and 377 home runs (most all-time at the position if we don't count Alfonso Soriano). His career was a bit short, however, at 17 seasons, he was a mediocre fielder, and it adds up to 55.2 WAR with 26.4 WAA. His peak WAA was 27.8, which is below the 30 mark I think of as the threshold. He just kind of doesn't seem quite good enough. On the other hand, he seems never to have used steroids, and a clean player in a world of cheaters suffers a double penalty in their context-adjusted stats. So I can see the argument for why he should be in, but don't think I'm convinced by it. That would be different if we wanted to exclude the known steroid users, but I don't, so I'm not in on him.

Voting Strategy
There are eighteen different players I would like to vote for, plus two more I'm on the fence about. I can only vote for ten of them, so I need to cut half of the players I've named. I can eliminate the two on-the-fence people right off the bat; if I'm not even sure I want them in, I shouldn't waste a vote that could go to someone else on them. Similarly I think Sosa, McGwire, and Palmeiro are out as low-priority votes, since I wouldn't feel particularly outraged if the Hall acted on the "steroid discount" approach that would keep them out. That leaves fifteen more of whom to pick ten, and now things get properly strategic: I want to put my ten votes to the most efficient uses, i.e. I want to cast them so as to maximize their impact. That means favoring people who are near the 75% threshold, and maybe also people who are on their 15th and last ballot, or close to it, or who might be in danger of falling below the 5% threshold to make it onto next year's.

Of the non-eliminated holdovers from last year, Biggio led the way with 68.2% of those votes, followed by Bagwell at 59.6%, Piazza at 57.8%, Raines at 52.2%, Smith at 47.8%, Schilling at 38.8%, Clemens at 37.6%, Bonds at 36.2%, Martinez at 35.9%, Trammell at 33.6%, and Walker at 21.6%. Of that crowd, only Lee Smith and Alan Trammell are nearing the end of their eligibility. For the others, then, let's assume the 1.1 multiplier Nate Silver found for the change in a player's vote share from one year to the next holds. That would get Biggio precisely to the 35% threshold, Bagwell and Piazza to around 63%-65%, Schilling, Clemens, Bonds, and Martinez to around 40%, and Larry Walker maybe closer to 25%. Probably Piazza and Bagwell are at the spot on the curve where an extra vote matters the most, because Biggio is projected to get in and once he gets in he's got all he needs.

The criteria, then, should probably be a blend of how important the marginal vote would probably be for a given candidate, how close to the end of their tenure on the ballot each candidate is, and, I think, how outraged I would be for that player not to get into the Hall. That puts Lee Smith in, for example. I dunno if votes for him have particularly high impact, but he's running out of time (especially with all these giants coming onto the ballot over the next few years) and I have to say that it's a travesty that he's not in the Hall, almost as much as it was that John Franco got kicked off the ballot after one year. Also, Mike Piazza is obviously in, because he's my favorite player. Plus he is unquestionably deserving, his exclusion would indicate an insane approach to the steroid issue worthy of anyone's outrage, and he's at the most critical spot on the vote curve. That's two in, five out, and five more needing elimination.

And I think Craig Biggio is next in line to go. I do think he deserves to get in, but honestly his case is kind of weak compared to many other people on the ballot and there's essentially no doubt that one of these years he will get in, after so strong a start. Greg Maddux, I gotta say, falls prey to the same phenomenon. He's going to get in, this year, and it should be pretty close to unanimous. As I said, there is no coherent argument against his induction. That means that voting for him would be a waste, in the way it's a waste for urban Democratic Congressmen to get 80% of the vote in their district. That leaves three to eliminate, and I'll spend them somewhat frivolously. One on Curt Schilling, mostly because he's a jerk and I don't like him and if I'm in a world where votes are scarce there's no reason to give one to someone I don't like. The second on Mussina, who is a bit marginal on the merits and who, well, was a Yankee. Hell, there's a serious chance he could go in as a Yankee, and we can't have that.

The final one, I think, in Roger Clemens, and this is an interesting one. I definitely think he should be in, for the same reason I think Bonds should be in. And I actually feel pretty strongly about the principles involved there. But I am going to vote for Bonds, and I think it's pretty clear from the overwhelming correlation between votes for those two players that if one gets in eventually, so will the other. And, well, I don't like Clemens much either. He was a bully, on the field and off it. His behavior around the (admittedly unfair) steroid investigations was pretty tawdry. And, well, he was a Yankee for a long time. I don't think he would ever go in as a Yankee rather than as a Red Sock,* but since I have to pick someone to unjustly snub I might as well go with the irrational Yankee hatred.

That leaves a ballot of, in alphabetical order, Jeff Bagwell, Barry Bonds, Tom Glavine, Edgar Martinez, Mike Piazza, Tim Raines, Lee Smith, Frank Thomas, Alan Trammell, and Larry Walker. Hopefully as many people as possible get in; as of this writing, a survey of publicly-released ballots shows that three of the four newbies, Glavine/Maddux/Thomas, plus Biggio, should get in, leaving us with cast of holdovers featuring as many deserving Hall of Famers as we had after last year's debacle. Lee Smith's vote share appears to be plummeting, maybe by as much as half, which is a damn shame. Piazza, meanwhile, is just below the threshold. I hope as many people as possible get in, 'cause it's not like the logjam will get any better next year, when Randy Johnson, Pedro Martinez, and John Smoltz join the party. All of this could have been avoided by putting in three or four people over each of the past several years. Hopefully the Hall will start some real catch-up growth beginning this year.


*How the hell does one spell this?