Hall of Fame, Hall of Mirrors (Part 2)

First, The Bad News

Despite a sea-change of historical proportions in a land disengaging from “fear itself”…

…“Black Monday” looms.

Why? The Hall of Fame will announce its voting results that day (January 12).

How? It seems all but certain that the BBWAA writers, who stand in for you and me—and are quite often seen as the Devil in the Deep Blue Sea—will give Jim Rice in excess of the 75 percent vote needed for enshrinement.

From our look at BBWAA voting patterns and the subsequent actions of the Veterans Committee, two scenarios are inescapable:

1) Rice will be elected by the BBWAA in his 15th and last “front door” opportunity.

2) If not, his 72 percent of the vote in 2008 ensures that some version of the VC will eventually admit him.

Despite this, there has never been a more pitched battle in the baseball world’s increasingly concentric (and contentious) circles. It makes Gore v. Bush look like a Sunday school picnic.

While I don’t think Jim Rice is the most deserving Hall of Fame candidate, I simply disagree about the consequences of such an election. We’ll return to this point later; for now, I want to digress (and, face it, what would a Don Malcolm article be without a digression? I know, I know: a whole lot shorter).

I want to present a prediction tool for Hall of Fame voting that puts a few issues into context—and boy, do we need some context on those issues!

Parameters for Prediction

The prediction tool requires at least five years of voting results for an individual player. There are too many possible perturbations otherwise.

And even then, we see that other circumstances can cause things to change course after a long-standing pattern has been in place—as we’ll see with the candidacy of Bert Blyleven, who, for our purposes here, is the “Anti-Rice.”

Let’s put the data on the table first:

Player 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 BTF
Baines 5 4 2
Bell 2 0
Blyleven 17 14 17 24 26 29 35 41 53 48 62 71 97
Cone 14 26
Dawson 45 50 50 52 61 57 66 70 37
Gant 0 0
Grace 8 0
Henderson 94 100
John 27 19 27 28 27 23 22 24 30 23 29 31 30
Mattingly 28 20 14 13 11 12 10 16 17 6
McGwire 24 24 29 85
Morris 23 20 21 23 26 33 41 37 42 46 3
Murphy 19 22 18 15 12 9 11 11 9 14 15 26
Orosco 1 1
Parker 16 21 16 14 10 11 13 14 11 14 15 16
Plesac 0 0
Raines 24 37 100
Rice 43 29 52 58 55 52 55 60 65 64 72 78 11
Smith L 42 37 39 45 40 43 45 23
Trammell 16 14 14 17 18 13 18 20 94
Vaughn G 2 0
Vaughn M 4 0
Williams 9 0
1st bal HOF 0 3 0 2 1 1 2 1 0 2 0 1 1
Vote/Bal avg 5.4 6.7 5.6 6.3 6 6.6 6.5 6.3 5.6 6.6 5.4 6.3 6.5
V/B Avg BTF 6.5 5.2 6.8 5.8 6.5

We have here 11 years of HoF voting by the BBWAA, with all of the 2009 candidates shown. Three men have been on that ballot in every year covered: Rice, Blyleven, and Tommy John (about whom, more later).

In 2008, Rice came very close to the 75 percent threshold. The prediction tool, which looks at the pattern in voting results over the past five years and makes an extrapolation, figures that he will cross that threshold in 2009 with about 78 percent of the vote.

A Hardball Times Update
Goodbye for now.

While I doubt that any sabermetric types will flee to Canada as a result (besides, Ron Johnson simply doesn’t have room for all of you…), for many it will be a Day of Darkness akin to the one I experienced back in 1972 when I “won” the only lottery I ever participated in—the U.S. Draft Lottery, where my birthday was literally “No. 1 with a bullet.”

For those who want to understand (and critique, even) the prediction method itself, a more detailed description can be found in the Resources section. Of course, this method is just one portion of Chris Jaffe’s estimable prediction method. I’m just going into a bit more detail.

Along with the predicted 2009 voting results, I’ve posted the vote percentages from the recent Baseball Think Factory Hall of Fame vote (far right column), which can safely be said to epitomize the sabermetric viewpoint.

Rocking the Boat That Rocked The Vote

Before we look at contrasts and contradictions in BBWAA and sabermetric voting patterns, however, I want to note a couple of things that came up in the process of assembling the data.

First, there’s the number of players voted for per ballot. Neither voting group comes close to finding 10 suitable Hall of Fame candidates. The BBWAA averages, shown in the bottom row, average out to 6.1 players per ballot over the past 11 years. The BTF group in 2009, with its “small-Hall” and “big-Hall” factions, has a very similar average (6.5). Its five-year average, in fact, is almost identical to the BBWAA’s 11-year average.

Second, there’s the hard-to-quantify impact of “first-ballot HoFers” on the rest of the candidate pool. What is clear from the player-per-ballot data is that the more “in-on-the-first-try” players there are in a year, the higher this average will be. In years with no “first-ballot” electee in the mix, the player-per-ballot average is down by about 20 percent—in short, one fewer player per ballot.

Interestingly enough, this same pattern exists in the BTF voting records (bottom row of table).

For the 2009 vote, there will be one “first-ballot” selection: Rickey Henderson. There will be little disagreement between BBWAA and stathead here. The pattern in the voting is that with one first-ballot type, long-time candidates tend to get helped a bit; with two or more, they tend to be hurt.

Three long-time candidates—Rice, Blyleven, and Andre Dawson—have been moving up in the voting over the past few years. Each of them weathered the 2007 election, when there were two first-ballot winners, and posted strong gains in 2008.

While the prediction tool suggests that Blyleven will get the biggest boost of the three this time, it concludes that it won’t be quite enough to get him over the 75 percent threshold. Rice and Dawson project to have more modest gains (about 5 percent of the vote instead of Blyleven’s 10 percent), but Rice is close enough that this will push him over the line.

And, as the table shows, for upwards of 90 percent of the sabermetrically-inclined voters at BTF, this will be a singular catastrophe, a blot on existence, a tear in the fabric of the universe, a…

OK, enough already.

What’s notable about the BTF voting is how two players who are much closer to one another in terms of their Hall of Fame worthiness than they are apart can be on such completely opposite ends of the ballot results—97 percent for Blyleven, 11 percent for Rice.

The only possible conclusion available is that ideology has taken over.

The history of the BTF voting indicates that it wasn’t always this way—at least to an extent. Blyleven has been a sabermetric favorite for a long time, with voting percentages in the BTF HoF polls in the 90s ever since the first year of voting (2005).

In that first year, Blyleven received 96 percent of the sabermetric vote. (His BBWAA percentage that year: 41 percent)

Rice received 31 percent. (The BBWAA gave him 60 percent of the vote.)

In subsequent years, Rice’s BTF vote total has, of course, gone into the toilet. He received 23 percent in 2006, fell through the floor to just 8 percent in 2007, 7 percent last year.

This is what a man far wiser than yours truly (but who will nonetheless remain nameless…) termed “the groundswell of groupthink.”

Groupthink in the form of more than a dozen proselytizing articles for Blyleven’s candidacy over the past several years.

And at least twice as many essays dismissing Rice.

So many this time last year that one wag termed it the “winter of dissed content.”

In the world of the BBWAA, though, Jim and Bert are both on the rise. The prediction tool suggests that if Blyleven gets 70 percent of the vote this time, he’ll make it across the 75 percent mark in 2010.

That will be a wonderful victory for “our side.”

But it will have been won by the use of the same tactics employed by everyone else who has been involved in the Hall of Fame selection process over the past 60 years.

It will simply be playing the age-old “power and influence” game with a new set of tools.

Meanwhile, a pitcher virtually equivalent to Blyleven, Tommy John, has been ignored by the saberites—and has seen his support slip in that community and flatten out in the BBWAA voting. As a result, John will wind up in the black hole of the Veterans Committee.

In 2001, John received a higher percentage of BBWAA votes than Blyleven. Once Bert became the “anti-Rice,” however, it has been mostly a case of “Tommy Who?”

You can call this “marshalling one’s forces” if you like. But, whatever it is, it’s not objective analysis. A more complete analysis would have concluded that both Blyleven and John were worthy, and would have urged the BBWAA to see them as linked.

This massive effort will produce the same result—and the same problem—that Bill James identified in The Politics of Glory: two players with very similar credentials for HoF membership discovering that the door opens for one, but not the other.

Can This Be Fixed?

Instead of focusing on such blatant partisanship, a different approach needs to be envisioned and implemented. Bill James suggested a new approach for Hall of Fame voting in Politics of Glory, but it was too cumbersome and had no practical chance for being accepted by the powers-that-be.

No, the place to fix the HoF voting process (to the extent that it’s possible to do so) is in the Veterans Committee. Keep in mind that from the inception of the HoF until the dawn of sabermetrics (1936-1979), the various incarnations of the Veterans Committee (warts and all) had enshrined 56 percent of the hitters and pitchers (note: those inducted solely for on-field achievements).

Since 1980, the Veterans Committee has accounted for only 31 percent of these inductees. Since 1990, that percentage is 27 percent (12 of 44). Since 2000, it’s down to just 16 percent (3 of 19).

So all this intervention at the front door of the process might result in a lone symbolic victory—enshrining Blyleven. Meanwhile, however, the Veterans Committee has turned into a bureaucratic nightmare straight out of Franz Kafka. The prime mechanism for redressing the imbalances and augmenting the inevitable omissions in the “front door” process is now stuck in a vat of molasses.

What can be done? That bureaucracy has to be torn away, and a new structure put in place that will break this escalating 30-year logjam.

How? By restructuring the Veterans Committee into autonomous working groups that specialize in periods of baseball history.

Wait a minute. Isn’t that what’s already been done, you ask?

Not the way I’d set it up. Instead of having 64 folks trying to reach consensus over a five-decade time frame, I’d create working groups for the 19th century, 1900-29, 1930-49, 1950-69, 1970-89. Those groups would have 15 members each, and, like the pre-1942 committee that just elected Joe Gordon, would have a good chance to reach a consensus.

To bring this to a boil, I’d suggest that BBWAA members be assigned to these working groups on the basis of specialized aptitude tests. By virtue of their demonstrated expertise, they would provide the historical/analytical background for each working group.

Take a guess as to which BBWAA members—those newly-minted BBWAA members who you’ve been hearing about of late—are likely to score highest on such an aptitude test.

The key to a return to a (mostly) reasonable Hall of Fame voting process is not division and self-interest within small enclaves. It is not an “us vs. them” mentality, mirroring the kind of demonization that has afflicted our culture and our politics over the same time period in which the two-party HoF voting mechanism broke down.

No, the key is a kind of enlightened subversion. By creating and exploiting a seam in the election process (as sketched out above), the new BBWAA members can establish a beachhead “behind enemy lines.” When this is accomplished, the Veterans Committee can fulfill its intended role: enlightened augmentation of the HoF selections.

A Silver Lining?

Rice’s impending election is seen by some as a slippery slope for an entire series of increasingly abominable HoF selections. This is part of a pervasive cultural negativism that has been attached to the sabermetric enterprise from the beginning (Bill James gave it sizzle with a dash of substance in The Politics of Glory, and the rest of us ran with it…) and is still in play today, even as advanced statistical methods have penetrated into baseball’s front offices.

In fact, though, Rice’s election may not be a negative. Rice is nowhere near the worst player in the Hall, and while he symbolizes “the wrong thing” to sabermetric types, there are aspects of his career that, if focused on properly, will permit future discussions and arguments to occur that have genuinely transformative possibilities.

Players with shorter careers but higher, more consistent peaks can now be directly compared to Rice, who is a contemporary or near-contemporary of many of these players (Ron Santo, Joe Torre, Dick Allen, Jack Clark, Reggie Smith, Frank Howard, Will Clark, Albert Belle). The fact that BBWAA voters see Rice as a “peak” player opens up an avenue of argument that can be applied to these other players, and that, with patient but relentless repetition, will prove persuasive.

All of this is simply a variation on the old saw that when life deals you lemons, make lemonade. To evoke another old cliché, it’s time to turn swords into plowshares (not Win Shares).

Disaster provides opportunity. So here’s your disaster, front and center. Now that it’s here, it’s time to step up to the plate with a new approach.

References & Resources
Nuts and bolts of the HoF projection tool

Use for players with five years of HoF voting results. For our example, we will look at Bert Blyleven.

Blyleven’s BBWAA vote percentages from 2004-08 are: 35, 41, 53, 48, 62.

1—Work backwards and calculate the differences in vote percentages for each adjacent year. These are: 14, -5, 12, 6.

2—Take the difference between year 5 (62) and year 1 (35), and divide it by 5.

Combine the following into the following equation: the sum of the four differentials in 1 with the five-year average change in 2 (double-weighted), divide by 5, and add the result to the year 5 voting percentage.

((14+(-5)+12+6)+(2*(27/5))+62 = 69.56.

Then, calculate the result when using only the measurement approach in 2:

(2*(27/5))+62 = 72.8.

Take the average of the two products: (69.6 + 72.8)/2 = 71.2.

The method projects that Blyleven will receive 71.2% of the BBWAA vote in 2009.

How accurate? In about 60% of the cases, the projection is within 2-3 percentage points of the actual vote. Errors of greater magnitude are almost invariably ones that are lower than the actual result. A more robust regression-based model might be able to account for the perturbations that occur when there are first-ballot HoFers in any given year.


Comments are closed.