Of all the personal testimonials honoring Jim Rice, my favorite is that of the much-beloved late commissioner of baseball, Bart Giamatti, who once wrote that Rice was “the Hammer of God sent to scourge the Yankees.”[1] That alone, in the minds of many baseball fans (outside of New York), should be enough to let Rice through the gates of the game’s Valhalla, Cooperstown’s Hall of Fame.

But, alas, Jim has stuck out 14 times with the Baseball Writers Association of America, and a debate rages over whether this final time will be the charm. Of course, even if he fails – or, rather, if the writers who vote on such matters fail him – his case will be shuffled off to the Veteran’s Committee where he may yet attain immortality. However, opinion across the land seems to be that there is something slightly dodgy and even undignified about entering the Hall in this manner, as though one has come through an inadvertently unlatched back door.

A lot the debate over Rice’s fate has been carried on at the level of “I saw him hit a home run against the [fill in a team name here] when I was [fill in an age under 10 here] and it was the most awesome sight I ever witnessed. [Therefore he should go to the Hall.]” We also see fierce, dramatic but intensely subjective judgments of the stature Rice had when he played. Pitchers, it is said by some, feared him, perhaps more than any other batter in baseball at that time.

SABR members and their intellectual brethren have debated Rice’s qualifications at a somewhat more sophisticated level (mostly), examining Rice’s statistics and awards while comparing his record to those of others who have (and haven’t) had their images inscribed on Cooperstown plaques. Consider, for instance, the claim that Rice was feared by opposing pitchers. Perhaps so, but then what are we to make of the fact that he never received more than 10 intentional base on balls in any one season? By this measure of “feared hitter,” Rice falls behind not only contemporaries Dale Murphy, Garry Templeton, Dave Winfield, and Dave Parker, but also Ted Simmons and Warren Cromartie (each of whom had two or three seasons with 20 or more IBB. With 77 career IBB, Rice is tied for 179th all time, along with players such as Jerry Grote, Ken Henderson, Claudell Washington, and Rice’s one-time teammate Fred Lynn.

In 16 seasons, Rice had a batting average of .298 with 2452 hits and 382 HR – each just a little short of the lifetime statistics that (used to?) assure one a ticket to the Hall. Still, Rice was an All-Star eight times and an MVP once (and he finished 3rd in MVP voting two other times, once in his rookie year, in which he lost to fellow rookie teammate Lynn). If the basic statistics fail to provide a clear answer, one can bring in second-generation statistics to help elucidate matters. For instance, Rice’s OBP was.352 and his SLG was .502, for an OPS of .854. This is just ahead of Hall-of-Famers Eddie Collins and Billy Williams, but behind non-Hall-of-Famers such as Reggie Smith and Jack Clark. So there is nothing decisive here for Rice’s case either. He remains precariously balanced on the cusp of greatness, like a star that is visible in the night sky only if you look slightly to one side of it.

The real statheads among us indulge in even more exotic stuff, like Bill James’ quantitative estimates of similarity among players.[2] Perhaps not surprisingly, Rice scores most similar to another legendary “tough case” for the Hall: Orlando Cepeda. In 17 seasons, Cepeda had 2351 hits (101 fewer than Rice), 379 HR (3 fewer), a .297 career BA (.001 lower), and a .849 career OPS (.005 lower). He was an All-Star 7 times (one fewer), a Rookie of the Year (one more), and an MVP once (tied). Rice fans will note that their man was just slightly better in nearly every case, and that Cepeda ultimately made it into the Hall. But Cepeda hit in an era of tougher pitching (lgOPS of .724 vs. .744 in Rice’s era) and, as a result, Cepeda has a slightly higher park-adjusted league-normalized *OPS+ (133 vs. Rice 128). Again, nothing decisive here. Let us move on. James has also developed some estimates of the likelihood of players entering the Hall. Naturally, Rice is low on one (HoF Standard = 44, where the avg. HoFer scores about 50) and high on the other (HoF Monitor = 144.5, where 100 represents a likely HoFer).

And so, finally, we come to James’ most recent, most influential, and perhaps most complicated estimate of player value: win shares. I won’t go into the calculations here (you can find it on the internet if you are interested), but win shares is supposed to tell us how many additional wins a given player was responsible for with his bat, his fielding, and (if applicable) his pitching. It is well-tested and well-known. It has its quirks, to be sure, but it is generally accepted to do a good job at measuring player performance.

How many win shares did Jim Rice have over the course of his career? 282. How good is that? It is tied with Boog Powell, the one-time MVP, mostly-Oriole LF-1B of the 1960s and 1970s. Powell is not, it should be noted, in the Hall. Fred Lynn is two win shares below Rice. He is not in the Hall. Minnie Minoso and Sal Bando are one win share ahead of Rice. They are not in the Hall. Amos Otis and Toby Harrah are a little further ahead (+4 and +5, respectively). George Sisler is 10 ahead and Dale Murphy (another notoriously tough HoF case) is 12 ahead, tied with Shoeless Joe Jackson. Then Cesar Cedeno (+14), Frank Howard (+15), Home Run Baker (+19), Ken Singleton (+20), Bobby Bonds (+20), Harold Baines (+25), and finally Orlando Cepeda at 310 win shares, a full 28 ahead of Rice.[3] At last, we have some solid evidence that Rice’s career contribution was, in cold reality, just a little below that usually needed to make it into the Hall; that perhaps his presence in Boston made him more visible nationally than Cepeda, who labored mostly in San Francisco and Atlanta (where he worked in the shadows of Willie Mays and Hank Aaron), but not actually quite as good a player.

A number of people have made exactly this case in the debate currently swirling around the vote for the 2009 Hall of Fame induction class. Of course, the win shares numbers are just evidence. They do not constitute definitive “proof.” One can continue to debate, among other things, the relative weaknesses of the various measures used, the importance of “peak” years, and a variety of “intangibles” that are not captured by any of the numbers. Fair enough. But this is how this sort of debate productively proceeds – from impressions, to statistics, to comparative statistics, to better comparative statistics, and so on. For instance, on 14 December 2008, David Kaiser posted an analysis of this kind to the SABR-L list, using win shares (among various other measures) to answer a number of questions about whether Rice should be in the Hall of Fame. Kaiser concluded:

The answers to this quiz are interesting because they show Rice as an almost classic case of a player writers tend to overrate: coming up with the Red Sox in one of their glory eras, he put up some spectacular home run and RBI numbers in his first few years and had one truly fantastic season. As a result he did quite well in MVP voting and was picked for a lot of All-Star games but his actual value was only once (1978) as large as it seemed, his secondary numbers were very poor, and he faded out quickly.

But then comes along Gabriel Schechter, a Research Associate at the National Baseball Hall of Fame & Museum, who wrote in a posting to the SABR-L list on 15 December 2008:

I simply want to register a strong protest over David [Kaiser]'s use of win shares as the primary tool of his analysis…. Rice played in the 1970s and 1980s, so how is it fitting to apply a sabermetrical measure that wasn't even created until 2001? Aren't those questions supposed to reflect how the player was regarded AT THE TIME he was playing? To say that Fred Lynn or Carlton Fisk had more win shares than Rice in a given season and equate that with considering Lynn or Fisk as more highly regarded than Rice is ridiculous.

And so we come to the real point of this column, which was not, it may surprise you to learn, to contribute to the Jim Rice HoF debate but, rather, to discuss the justice of using modern statistical tools (like win shares) to decide historical questions (like whether Jim Rice was so great a ballplayer that he belongs in the Hall of Fame).

I do not know Mr. Schechter’s views of statistical analysis generally. There are some fans (and players and managers) who believe they see plainly with their eyes (and with their memories), and that statistics, with all their fussy formulas, only confuse the issue. Without further ado, I commend to them the cognitive psychological work of people such as Paul Meehl, Gerd Gigerenzer, and Nobel prize-winner Daniel Kahneman to disabuse them of their misapprehension. I will assume that Mr. Schechter, instead, is only objecting to the casting back of modern statistics into historical eras. I suspect, however, that he has confused two superficially similar, though, in point of fact, quite distinct complaints. The one, to which many object, has to do with creating leader boards and records for statistics that did not exist when a particular season was played. So, for instance, claiming that Three-Finger Brown led the NL in saves four years running, from 1908-1911 (5, 7, 7, 13), seems a little silly not just because there was no such statistic for Brown to lead the league in then, but also because the conception of the relief pitcher as a kind of “specialist” with a particular “function” (such as “saving”) was not yet in place in Brown’s time. It is a little like claiming that Hannibal had more “tanks” than the Romans on account of his use of elephants. I have some sympathy with this objection.

However, that is not what is going on when Mr. Kaiser (and others) use win shares to analyze the performance of players past. First of all, there is nothing that goes into computing win shares that would have been foreign to Rice or his cohort: hits, at bats, bases on balls, total bases, outs, etc. Mr. James has just stirred a little differently a pot of wholly familiar ingredients. Second, the point of doing this kind of analysis is not (only) to create a retrospective leader board, but rather to use quantitative methods to analyze Rice’s performance relative to his peers (and to others throughout the history of major league baseball). With a modicum of judiciousness, there is nothing in the least ridiculous about this process. Indeed, we do it all the time.

To wit, which of these historical questions are ridiculous? How many people lived in the city of Rome in 44 bc? What proportion of them were slaves? What was the average life expectancy? What were the leading causes of death? Among the land owners? Among slaves? Across genders? All of them require quantitative answers. All of them were questions that went unasked (and unanswered) by the Romans themselves. That does not make them historically illegitimate. Consider more questions of the same type: What proportion of the US population spoke English as a mother tongue in 1776? What proportion of the American population approved of Abraham’s Lincoln’s actions in 1863? Would Woodrow Wilson have won the 1912 presidential election if either William Howard Taft or Theodore Roosevelt had dropped out of the race?

The people of these eras did not have either the data or the methods (or both) to answer such questions definitively, but certainly there is nothing to prevent us from using the methods we have since developed on the data that we still have from those times to develop answers that are in some ways better than the ones people of the time in question could have generated (for instance, computers make it possible for us to manipulate huge masses of data that would have been impracticable, if not strictly impossible, prior to their invention).

Far from being illegitimate, a statistic like win shares is precisely the kind of evidence to which members of the BBWAA should attend more fully when deciding questions like whether Jim Rice was as good a player as the others who are now in the Hall. It allows us to separate dispassionate consideration of the merits of the case from contentious but ultimately irrelevant stories of who thrilled us when we were young. Isn’t that exactly why the BBWAA waits five years after a player retires before considering his case for entering the Hall – to let passions cool and allow the facts to rise to the surface?

-----

Notes:

[1] Giamatti, A. Bartlett (1998). "The Green Fields of the Mind." In A Great and Glorious Game: Baseball Writings of A. Bartlett Giamatti. Algonquin. Available on-line at: http://mason.gmu.edu/~rmatz/giamatti.html.

[2] The source I used was baseball-reference.com.

[3] I have only picked a few familiar names between Rice and Cepeda. In fact there were 56 players separating the two on the all-time win shares list, as of 2002. (Players like Frank Thomas have since passed Cepeda. Others have, no doubt, crept between them from below Rice in the intervening years.)

Christopher D. Green teaches statistics in the Department of Psychology at York University in Toronto. His academic research is mostly concerned with the history of psychology.