February 25, 1998

Abstract Progress: Rebuttal and Reply

Reader response on Bill James' influence

by James Kushner and Steven Rubio

Printer-friendly

Dear Steven Rubio and other Baseball Prospectors:

I'm writing in response to your article "Abstract Progress", which decries the lack of influence that Bill James and his theories have on "the minds of the average fan or baseball writer."

It seems to me that the source of Mr. Rubio's discontent is extremely narrow in focus: Hall of Fame voting. Well, of course the Hall of Fame voters are going to let you down - the makeup of the voting body practically guarantees it. Anyone who's been a member of the BBWAA for more than ten years is likely to have formed some pretty firm views about what determines quality in baseball, and to have had those views locked in long before the sabermetric revolution took hold. It takes a very courageous individual to publicly change one's mind, and sportswriters, as a body, are not typically known for their courage.

The Hall of Fame voters of today are, by and large, the same people who overrated Steve Garvey in the early 80's. They have nothing to gain as individuals by changing their minds now, so they probably won't. A quote attributed to physicist Max Planck seems apropos here: "A new scientific truth does not triumph by convincing its opponents and making them see the light, but rather because its opponents eventually die, and a new generation grows up that is familiar with it."

The problems faced by sabermetricians is a lifelong one, but it seems entriely likely that within ten years, the majority of the active sportswriting community will see the merits of sabermetrics as a tool. However, the Hall of Fame voting membership will probably have to wait until 2015 or so for the sabermetics-influenced generation to rule the roost. (2015 was chosen because it represents a date thirty years from the original publication of "The Bill James Historical Baseball Abstract." I figure that only by then could the voting membership possibly have a majority of voters who read Bill James before they became sportswriters.)

Also, the philosophy of Hall of Fame voting works to perpetuate contemporary illusions and misconceptions, not to rectify them. In most questions of the value of the best players (and with the Hall of Fame, we are theoretically talking about the best players), the opinions of contemporaries and the opinions of history are the same. (The really tough questions don't have to get asked--the Hall of Fame doesn't care whether Mickey Mantle was better than Willie Mays. They're both Hall of Famers, and that's that.) In the tough cases, the voters will defer to the opinions of contemporaries. (Indeed, for the Negro League committee, that's all they have to go on.) The Hall of Fame, in choosing whom to honor, usually opts to reinforce whatever image of a player's quality was in force at the time of the player's career. To try and always keep current with historical opinion would necessitate throwing people out whom history no longer favors. ("You can't have people wondering if the honor was temporary," Bill James once noted.)

James argues in favor of using contemporary opinion as a factor in the historical record of a player's quality in the "Historical Abstract", notably on pp. 306-307 (the Johnny Bassler/Jimmy Brown segment). Regarding the Garvey/Clark/Downing example, how better to quantify the opinions of contemporaries than through the summation of MVP Award Shares?

MVP Award Shares, Career:
Steve Garvey 2.470
Jack Clark 1.155
Brian Downing 0.133

Clearly, the more that voters feel determined to defer to contemporary opinion, the better Garvey's chances are. Indeed, it appears that the Garvey-for-Hall-of-Fame question will be a very good test case for the power of sabermetrics to sway historical opinion. He seems to be right at the borderline of the area where contemporary opinion alone will get one into the Hall of Fame:

Garvey ranks 28th in the postwar ranking of MVP Shares. Of the 27 players ahead of him:

• 18 are in the Hall
• 3 are still active (Barry Bonds, Frank Thomas, Ken Griffey Jr.)
• 3 are retired but not yet eligible (George Brett, Eddie Murray, Kirby Puckett)
• 2 are eligible but have not attained induction (Jim Rice, Dave Parker)
• 1 is Pete Rose

Of the folks on the above list, Puckett and Parker will be interesting to watch, as they both had a higher perceived value among the sportswriters of their day than is borne out by sabermetric analysis of their stats. In addition, two other who are just behind Garvey, Andre Dawson (2.366 MVP Shares) and Dale Murphy (2.308) will also be important for vote-watchers.

Of course, historical judgement can work in favor of those who have been unfairly slighted in their own time:

Eddie Mathews: 1.696 MVP Shares
Willie McCovey: 1.645 MVP Shares
Billy Williams: 1.598 MVP Shares

All are in the Hall, with not much controversy. Williams, of course, may have hurt Ron Santo's chances with his own induction, but it's clear that voters of the time thought that Williams was a better player. Santo had 1.223 MVP Shares during his career.

Indeed, in my hazy crystal ball, Lou Whitaker, recipient of a puny 0.214 MVP Shares in his career, stands a better-than-even chance of induction.

Regarding the question of whether things are that bad today: well, they're certainly getting better. ESPN shows on-base percentage routinely as one of the stats for each batter. The BBWAA has made some perceptive choices lately, notably Pat Hentgen and Pedro Martinez's Cy Young votes. Indeed, I think that the sabermetric influence first hit the BBWAA in 1990, when they made the right pick with Barry Bonds for NL MVP, despite his having a teammate (Bonilla) with more R and RBI. (Bonds did have a higher batting average, though, so it may have just been old statistical prejudices inadvertently reinforcing a correct decision.)

Of course, since then there have been some howlers committed both the the BBWAA (Terry Pendleton? Juan Gonzalez?) and major league GMs (Joe Carter? Galarraga's contract??) but the climate is certainly improving.

In a Panglossian vein, I do believe that the sabermetric disciples do currently live in the best of all possible worlds. The fact that they're right means that the future of historical opinion does belong to them, while the present mixed state of acceptance means that they have many opportunities to vent their spleen at "the conventional wisdom", which is one of the truest delights of any sports fan.

Yours sincerely,
James Kushner

P.S. Rubio's reference to Bill James' comment that "the on-base percentage of a team's leadoff man was the most important factor in determining how many runs a team would score" is largely irrelevant to the debate. Of the three players he focusses on, only Downing was ever a leadoff man, and then for only a few seasons. If James' comment had referred to the on-base percentage of the entire team rather than that of the leadoff man (and he has made many comments like that over the years; couldn't Rubio find one to quote?) it would have had more relevance to his argument.

Steven Rubio replies:

James Kushner makes some excellent points in his response to my piece on sabermetric progress, and it's just the kind of exchange of ideas we welcome at the Prospectus. Having read James' reply here, and after further discussions with him, I can say we share many ideas about the proper use of sabermetric techniques to evaluate player production. Ultimately, the primary focus of our disagreement revolves around definitions of the "mainstream media" and/or the "average/casual fan." James is much more optimistic than I about the impact of sabermetric approaches in the mainstream, and I admit that I hope his optimism is warranted. As long as triple-crown hitting stats take precedence in the minds of most fans over even something as elementary as on-base percentage, though, I will remain pessimistic. Thanks to James for his thoughtful reply; nothing is more welcome to a writer than to know an audience has taken him seriously.