It’s that time of the year again, the time when the baseball world looks to Cooperstown. This Wednesday, at 2 p.m. EST, the Hall of Fame will announce the results of its annual BBWAA election, and we’ll see who goes into Cooperstown.
This marks the seventh straight year I’ve tried to give you an advance warning of what will happen. Each January, I write this column predicting not only who will go into Cooperstown and who won’t, but I’m foolhardy enough to predict exact voting percentages for all major candidates.
I’ve had some success over the years. In the last six columns, I made 96 predictions on what level of support a candidate will receive. I’m proud to say that in 32 cases—exactly one-third of them—I was within one percent of a players actual support. And 79 times I’ve come without five percent of the real total. My average margin of error has been 3.5 percent.
Admittedly, I’ve missed some big ones. I said Craig Biggio would enter Cooperstown last year, and he didn’t. I made the same mistake with Roberto Alomar in his first year on the ballot (and he just missed with 74 percent). But no one bats 1.000.
My system is based on 10 guidelines, and here they are.
All other things being equal, assume guys will do about as well next year as they did last year. However, all other things aren’t equal, leading to the other rules.
2. Strength of ballot
During years in which there are a lot of strong candidates entering the ballot, the backlog candidates see their support collapse. For example, in 1999, Nolan Ryan, George Brett, Robin Yount, Carlton Fisk and Dale Murphy joined the ballot, and every single member of the backlog declined.
More recently, in 2013, the backlog also suffered when Biggio, Mike Piazza, Curt Schilling, Barry Bonds, Roger Clemens and Sammy Sosa joined the ballot. The 2013 backlog wasn’t as badly affected—four returning candidates actually saw their support increase—but the previous ballot set a record for fewest names per ballot (5.10), so there was plenty of space available.
Alternately, when there is a weak rookie crop, the backlog rises up. But that doesn’t really apply this year, as Maddux, Glavine, Thomas, Mussina and Kent go on the ballot. Based on this guideline, we should expect most of the backlog to have a rough time of it this year.
3. Candidates per ballot
The previous guideline bleeds into this one. The Hall of Fame limits BBWAA members to voting for no more than 10 candidates per election. Technically speaking, the previous guideline shouldn’t be an issue. There is almost always enough space on a ballot to let new guys come in.
Remember how every single member of the backlog had his vote total go down from 1998 to 1999? Well, there were only 6.74 names per ballot in 1999. There was space available to let others rise up.
The 6.74 names/ballot in 1999 is the most the Hall has seen in the last 20 years. The BBWAA ballot hasn’t averaged seven names per ballot since 1986. It hasn’t topped eight names per ballot since 1983. It last averaged nine per ballot in 1955. Technically speaking, there is nothing stopping the BBWAA from letting the new guys come in while still keeping their older guys on the ballot.
Aye, but that’s the thing. Clearly, many BBWAA members are averse to filling up their ballots. It’s more a self-imposed limit than anything else. If you typically vote for three or four names per ballot, listing seven at once seems cheap. There is a reason why the BBWAA hasn’t averaged seven players per ballot since the Cold War.
Last year, the BBWAA had 6.60 names/ballot, one of its highest totals in decades. And, of course, no one was elected, so the retuning backlog is the highest in decades: 6.29 names per last year’s ballot on this year’s ballot. Combine that with the arrival of Maddux, Glavine, Thomas, Mussina and Kent, and at the very least, it’s clear we’re going to have more names per ballot this year than any time in decades.
One of two things is going to happen this year. Either the backlog is going to be completely and utterly slaughtered, or the BBWAA members are going to have to get over their self-imposed limitation and start filling out their ballots en masse with a full complement of 10 names. In fact, even if the BBWAA starts voting like it’s 1955 all over again, many backloggers will still go down this year.
4. “Over the top” surge
Those last few points talk in general about what will happen to the ballot as a whole, and to backloggers specifically. But not every member of the backlog is affected the same way. Some guys are especially likely to see their vote total rise: guys at the top of the backlog.
For example, remember how I said that four backloggers in 2013 had their support go up last year despite the big rookie candidate crop? Well, three of those four were at the top of the backlog: Jack Morris, Jeff Bagwell and Tim Raines.
I studied this years ago and found that if a candidate’s support hits 50 percent or higher, more voters start joining the bandwagon. Candidates who already have the support of most writers pick up the most new supporters the next year. The higher his vote total, the things start moving in his direction. But candidates under 50 percent are more likely to just tread water or fall under.
Logically, this doesn’t make sense. If you have one guy at 52 percent and another at 25 percent, the latter should find it easier to pick up supporters because there are so many more possible converts out there. True, but the guys at the top of the ballot are the ones at the forefront of people’s minds.
Once most of the BBWAA supports a candidate, the other members start wondering what they’re missing. The question shifts from, “Why should I vote for this guy?” to, “Why not?” And the logs start rolling his way.
This year, there are five returning backloggers who received more than half the votes last time. That’s huge. The last time five candidates received between 50 and 75 percent of the vote was 1983. They were Harmon Killebrew, Luis Aparicio, Hoyt Wilhelm, Don Drysdale and Gil Hodges. All but Hodges wound up getting elected by the BBWAA, because people in that top 50 percent tend to keep rising, often up to 75 percent eventually.
The top five from last year—and their support from last year—are: Biggio (68.2 percent in 2013), Morris, (67.7), Bagwell (59.6), Piazza (57.8) and Raines (52.2). If anyone in the backlog is going to rise (or at least hold his own), it’s these guys.
5. Comparable candidates
What can really kill a backlogger is if a new candidate emerges who is similar to the backlogger but clearly superior. Or, even worse, if multiple similar candidates emerge. Heck, even if the new guys aren’t clearly superior, having some similar candidates can diffuse support away from the backlogger.
The all-time great example of this came in 1989. That year, Gaylord Perry, Fergie Jenkins and Jim Kaat joined a ballot that already featured Jim Bunning, Mickey Lolich and Luis Tiant. The upshot is that Bunning, Lolich and Tiant had their support crater.
This year’s crop of new candidates is extremely pitcher-heavy. Three of the top four ballot rookies are pitchers: Maddux, Glavine and Mussina. This doesn’t bode well for pitchers in the backlog: Morris, Schilling, Lee Smith and Clemens.
While Clemens is a special case (see the next guideline for more on him), a good case can be made that the three other backlog pitchers are clearly inferior to the three big rookie pitchers. On a crowded ballot, some guys will get crowded off.
In 2007, Mark McGwire joined the ballot, and the BBWAA Hall of Fame voting process joined the steroid era. Votes about steroid guys are often less about what you think of him as a player and more about what a voter thinks of steroids.
For instance, last year Clemens and Bonds both received under 40 percent of the vote (38 percent for Clemens, 36 percent for Bonds). It’s safe to say that the remaining 60-plus percent of the voters just won’t support someone associated with steroids.
Looking at the older PED-associated candidates, clear trends have emerged. First, their support tends to be unusually stable. Second, when their support changes, it only goes down, not up. Those who won’t vote for them have remained resolved, while some supporters start to give up on them as either a lost cause or because there just isn’t enough room on the ballot, especially with a ton of steroid-associated candidates now on it.
For example, look at the support Mark McGwire has had since going on the ballot:
2007 23.50% 2008 23.60% 2009 21.90% 2010 23.70% 2011 19.80% 2012 19.50% 2013 16.90%
That is remarkably stable, especially in his first four years. Thus, we can assume that the steroid guys won’t see increases in their support this year. Here is what happened to them last year:
Clemens 37.6% Bonds 36.2% McGwire 16.9% Sosa 12.5% Palmeiro 8.8%
There have been some whisperings about PEDs with other candidates, but those five are the ones that many/most voters believe used PEDs.
Clemens and Bonds should remain pretty stable. They, along with Maddux, are all clearly and easily the best candidates purely in baseball terms, so if someone was willing to vote for them last time, they’ll still vote this time. The other trio should see a notable loss of support in this crowded ballot.
7. Last year on the ballot
Players in their last year on the ballot typically receive a bump, averaging around three percent. For example, remember how I noted that only four backloggers had their support go up last year? Well, three were experiencing the “over the top” surge, and the fourth was Murphy, who was in his last year on the ballot. In fact, Murphy had the biggest rise in support of anyone, going up 4.1 percent from 14.5 to 18.6.
This year, one man is in his final year on the ballot: Morris. He is a very weird candidate this year because so many factors affect him, and they pull him in different directions. He’s in the “over the top” surge territory, and he’s in his last year on the ballot. In most years, those things would often be enough to raise him from last year’s 67.7 percent to the Promised Land of 75 percent and a Hall of Fame plaque.
True, but Morris also is hurt by the strength of the ballot and comparable candidates. Last year, no one went up by more than 4.1 percent, and Morris needs to go up by nearly double that. Last year, Morris’ support rose by just one percent. Can he really do much better on a far more crowded ballot with four pitchers who have more career wins than him (and only one, Clemens, associated with steroids)?
I think Morris is going to the Veterans Committee. Those voters probably will put him in when they look at their more recent candidates in 2017.
8. Repoz’s BBWAA ballot tracking Gizmo at Baseball Think Factory
Each year, BTF website editor Repoz does a valuable service for those of us interesting in the Hall of Fame debate. He tracks every single BBWAA ballot made public (and even has some other voters privately let him know how they voted) and posts the ongoing tally. It’s called the HOF Ballot Collecting Gizmo, and it brings in tons of results.
Last year, Repoz set a personal best by tallying 194 votes (though many were made publicly available only the day of the Hall’s announcement itself).
It isn’t a scientific sampling, and the Gizmo does tend to have some bias. For example, Morris and Don Mattingly do better with the overall electorate than in Repoz’s sample. Raines does better in the Gizmo than in reality. But it does provide plenty of pointers.
The Gizmo does two things especially well. First, it lets us know how rookie candidates are doing. This is especially important for me because all of the other items on this list deal primarily with returning candidates. Second, the Gizmo gives us an idea how many names are being listed per ballot.
Normally, that second point isn’t too huge a deal. Normally, I can figure it out on my own. But the 2014 election is proving to be anything but normal. The results in the Gizmo are so far shocking and unprecedented to me.
As I write this, Repoz’s Gizmo contains 130 full ballots (over 20 percent of the total electorate), and they are averaging 9.26 candidates per ballot. Damn near the entire BBWAA is filling out ballots fully. Remember, the last time this body averaged over nine names per ballot, there were only 48 stars on the American flag. None of the current voters has taken part in an election that has averaged over nine names per ballot, but so far they are this year.
Okay, let’s back up for a second. How accurate has the Gizmo been in years past predicting this sort of thing? Well, the 2013 Gizmo averaged 6.59 names/ballot. The actual result was 6.60. Admittedly, the earlier returns from the Gizmo were higher: 6.73 from the ballot when the Gizmo was only half-full. In 2012, the Gizmo was actually low, with 4.70 names in its tally versus 5.10 in reality.
So it can be off, but it shouldn’t be too off. For now, I’m guessing the 2014 Gizmo is on the high side—because I can’t imagine it would be low. But I’ll guess we’ll have a little over nine names per ballot. I never would’ve guessed that otherwise, but there you go.
Actually, the question then emerges, why would the vote total go up so drastically this year? I have a few thoughts. First, last year Cooperstown was skunked, with no one elected. As an added bonus, the Veterans Committee elected in only people who’d been dead for decades, so no one really cared about any of the inductees.
And remember, five players received support from a majority of BBWAA writers. They think there are several deserving candidates—6.60 names/ballot is huge for the modern BBWAA—and the combination of all of these factors has caused may voters to break through any self-imposed limitations on how many candidates to vote for. Most voters are filling out complete ballots or leaving just one spot blank.
9. Beware five percent
If you fall below five percent, you fall off the ballot. People hovering just above five percent typically fall under it over time as just enough of their few backers give them up as a lost cause.
Rafael Palmeiro is our lowest returner in the backlog with 8.8 percent of the vote last time. Normally, that would plenty enough to ensure he returns for 2015 (a man isn’t likely to lose half of his voters in one year, after all). But this isn’t a normal year, not with this strong ballot.
10. Guidelines ain’t rules
These are all just trends. You can’t take any one of them too seriously. Predictions are more art than science, so I don’t have any clear mathematical formula for what comes below.
Enough of that already. It’s now time for the main show, how I think guys will do on Wednesday. Here are my predictions for the 22 main candidates, as well as a little blurb at the end to cover all other candidates. For comparison’s sake, I’ll include last year’s vote total so you can see who I think will go up or down.
Candidate 2014 2013 Greg Maddux 99 Tom Glavine 94 Frank Thomas 85 Craig Biggio 76 68 Mike Piazza 71 58 Jack Morris 70 68 Jeff Bagwell 62 60 Tim Raines 50 52 Roger Clemens 38 38 Barry Bonds 36 36 Curt Schilling 34 39 Lee Smith 34 48 Mike Mussina 29 Alan Trammell 26 34 Edgar Martinez 23 36 Larry Walker 16 22 Fred McGriff 13 21 Jeff Kent 11 Mark McGwire 11 17 Don Mattingly 10 13 Sammy Sosa 7 12 Rafael Palmerio 5 9 Others 4
That’s 9.04 names/ballot.
No matter how crowded a ballot is, if everyone agrees who the best candidate is, he gets tons of support. Thus, in 1999—on that super-crowded ballot—both Ryan and Brett sailed in with some of the highest support percentages ever. (Ryan actually tied Tom Seaver for the all-time record with 98.8 percent). Maddux should sail in similarly this year.
The BBWAA normally doesn’t elect four guys in one election. In fact, it hasn’t done so in decades. But the writers also haven’t had nine names per ballot in decades. With so many strong backloggers over 50 percent and three hugely qualified rookie candidates, having four inductees is likely. Even five inductees are possible. That’s happened in BBWAA voting only once, in 1936, the very first year.
Biggio and Morris were practically tied last year, but the 2014 election looks better for Biggio. First, comparable candidates really hurt Morris more. Second, Repoz’s Gizmo currently has a huge split, about 20 points between the two. Third, as a rule of thumb, if one candidate takes 14 years to win two-thirds of the voters, while and another starts off with two-thirds, it’s going to be a lot easier for that second guy to win more voters. That’s also why I think Piazza will leapfrog Morris.
The voters are trying to fill out their ballots and trying to help guys at the top of the backlog while also making way for the new candidates, so someone has to pay the price. It won’t be Bonds or Clemens, because they are still overwhelmingly qualified for those willing to vote for PED guys.
One guy above Bonds and Clemens last year should take a big fall this year, Smith. He’s always been an odd candidate. The Hall has never quite known how to handle relievers. Smith’s case was always based on being the all-time saves leader, but he no longer is.
He’s also been aided in recent years by a general lack of strong pitching candidates. That started to change last year with Schilling and Clemens debuting, and it really falls apart this year. Smith should take a serious hit.
Palmeiro might fall under five percent, but my hunch is he’ll manage to stay above.
There you go. Clip ’n save, and come back on Wednesday to find out just how dumb I am.
References & Resources
The Gizmo is huge. I also have a database with all previous Hall of Fame voting results.