It was ten years ago last month that baseball stats guru Bill James declared he was "breakin' the wand"--ending his annual Baseball Abstract series and abdicating his growing reputation as a wizard of the sport. I apply the overused term "guru"--a label James himself decries--because in his case it is an accurate metaphor. Working from his bank of computers on high--actually in Kansas, not quite as remote as the Himalayas, but close--James served as an oracle, a priest, a man who conveyed his convictions about baseball to whoever was open to them. His Baseball Abstract, which started as a newsletter in the mid-70s and grew until James's "retirement" in 1988, revolutionized the game, bringing terms like "ballpark differential" and "platoon advantage" and, perhaps most important of all, "on-base percentage" into the sport's mainstream. James never abandoned baseball--he has since churned out excellent books on the Hall of Fame and the history of baseball managers--but he did abandon the form of the Abstract as too limiting. He first moved to the sprawling annual Bill James Baseball Book, then to the more practical Bill James Player Ratings Book, a series he likewise ended a few years ago.
Since then, the sport has been lacking a well-researched, well-argued, well-written annual stats volume--in short, a dependable analysis of players and teams (though Total Baseball has done well in a more exhaustive encyclopedia format). The shortage is not of quantity but quality. Each spring, any number of books come out on baseball and its statistics. With "fantasy" and Rotisserie leagues still flourishing, it's a boom industry. Some of this material, like the antiquated Street & Smith annual or the newer but equally conventional Bill Mazeroski magazine, recycles the usual gossip and the same tired scouting reports about "the book" on various players. The James method, by contrast, is to remove the subjective analysis of scouts and so-called baseball experts from the equation and stick to statistics as an objective and accurate measure of worth. Yet even this approach has been taken to extremes by STATS, the group James passed his broken wand to a decade ago. STATS's annual baseball books, chock-full of raw data, are mind-numbing even to a former high-school calculus student and admitted stats hound like myself. Now, however, after a year of research, I am ready to anoint an heir apparent to the James mantle: Baseball Prospectus, a big, thick book now in its third year and for the first time receiving the support of a publisher, Brassey's. It's the best baseball annual since The Bill James Player Ratings Book, and it may turn out to be as invaluable as the old Abstracts.
To its credit, Baseball Prospectus is not timid about claiming the James birthright. The first paragraph, after the notes and dedications and such, cites James as an inspiration and bewails the continued lack of acceptance of many of his ideas. "The central idea of this book is that player performance can be objectively analyzed, without need for secret knowledge or mysticism; that empirical knowledge and the laws of statistics can be applied to baseball analysis." It's an attitude that continues to be dismissed by many baseball professionals and writers--or "mediots," as they're sometimes called in BP (as I'll refer to it from here on). The book mentions the humiliation Boston Red Sox general manager Dan Duquette suffered after admitting he relied on stats to make decisions--he was labeled a "Rotisserie" GM--and I could mention the skepticism that greeted the use of computers by managers Larry Dierker and Dave Johnson. Both, it's worth noting, made the playoffs last year with talented but also very demanding rosters.
This is really nothing new. Sparky Anderson once dismissed James as a little man in glasses who didn't know anything. (James does wear glasses, but he would tower over Anderson.) There is continued resistance to a stats-based scouting system throughout baseball, where players and managers and executives and scouts all demand that a player has to be seen--and studied with trained eyes--to be believed. There's a commonsense resistance to the idea that the eyes cannot be trusted, something BP makes note of in remarks about the Arizona Diamondbacks' pitcher Clint Sodowsky. After several mediocre years with the Detroit Tigers, he began putting up good numbers--and looking good--in Pittsburgh last year with the Pirates. "Very few players learn to hit or pitch that suddenly," says the writer, pointing out that the improved performance over just 54-plus innings is "probably pretty normal variation"--in other words a streak. (One of the problems with BP, it should be pointed out, is that it's produced by a staff of no fewer than eight authors, and it's usually impossible to tell who is writing what.) The writer gushes about how good Sodowsky looked, but goes on to conclude that "in the past, when I've paid more attention to my eyes than the numbers, I've been roughly as accurate as a Mike Ditka-coached quarterback."
The eyes sometimes lie, as any baseball fan knows and as most baseball experts should be willing to admit. The common phrase for a player who is deceptive in such a manner is "someone who looks good in a uniform," someone with athleticism but no real sense for the fine points of the game. Where BP and the conventional baseball scouting report--"the book"--diverge is that BP believes there are many players who look good but actually hurt a team, players like Brian McRae and Shawon Dunston and Ozzie Guillen (to name just three familiar former Chicagoans). They may look good, they may be skilled, they may even be exuberant, but there are holes in their games that hurt a team, and they show up over time in statistics--and in runs. The main drawback with all three of those players is an inability to take a walk--that is, a low on-base percentage. Like James, BP hammers away that OBP is still not granted its proper importance in the baseball world. It seems obvious: get more men on base, score more runs. Yet much of baseball still believes the macho idea that a walk isn't as good as a hit. USA Today and Baseball Weekly--both deeply influenced by Jamesian statistical research or Sabermetrics (derived from SABR, the stats-oriented Society for American Baseball Research)--both include OBP in their statistical tables, but neither the Cubs nor the White Sox included OBP in their 1997 media guide--for good reason, where the Cubs are concerned.
On-base percentage has been the biggest single problem with the Cubs for years. Both Dunston and Ryne Sandberg were problems on that front last year, as was McRae, a barely serviceable leadoff man at best who went south early in the season, acting as a catalyst for the team's atrocious start. From that standpoint, BP analysis would seem to suggest that the Cubs improved themselves offensively through the addition of Jeff Blauser and Mickey Morandini--both proven OBP performers--though Blauser will have to be watched, as will prospective leadoff man Lance Johnson, whose OBP flourished when he played with the Mets but who was never quite good enough to bat leadoff for the White Sox. There were also warning signs. Henry Rodriguez, like Sammy Sosa, refuses to walk, so don't expect any increase in runs batted in from Kevin Orie hitting behind them.
Of course, it is best for all concerned when the numbers endorse what the eyes see. BP projects big things for the Sox' Mike Cameron, just as I did last summer. That's because in addition to looking like an athlete, with Popeye forearms and a bull neck, as well as hitting for power, fielding like a demon, and possessing excellent speed, Cameron knows how to take a walk. That reflects something James called "athletic intelligence," and it indicates a player more likely to develop into something--or someone--exceptional.
That's the thing about baseball's detailed system of statistics. It isn't just numbers, it's a portrait in numbers that reflects development, injuries, the role of a player on a team, and, yes, even changes in personality that go with personal changes such as a move from one city to another. As James wrote in the 1985 Abstract: "I don't pay any attention to statistics on the stock market, the weather, the crime rate, the gross national product, the circulation of magazines, the ebb and flow of literacy among football fans: just baseball....Baseball statistics, unlike the statistics in any other area, have acquired the powers of language."
On that note, a warning about BP: its statistical language, though simple enough on the surface, is an incredibly complicated dialect in structure. Each of the 1,800 player profiles comes with the usual statistical table, but the figures in those tables aren't "real"; rather, they're too real, having already been adjusted for ballpark differentials, players' ages, and levels of competition (minor leaguers as well as those in the majors), to give the best possible measure of each against the norm. The conclusions to be drawn are sometimes quite amazing, such as BP's declaration that "Andres Galarraga certainly isn't the hitter [career minor leaguer Roberto] Petagine is." Yet I'm here to testify that those conclusions are oftentimes quite astute and on target (if weighed against a given manager's ability to recognize the same things the stats do, because without playing time nobody is going to put up good numbers). Using last year's BP in my own Rotisserie League, I found it more productive than anything since James. That's the single best testimonial I can give.
Yet James and BP share qualities that are perhaps even more important. Both James and the BP authors write well and conversationally, undercutting their authoritative data with a disarming sense of humor. Get the South Park reference in the Hal Morris item, or this remark on the weak-armed reliever Greg McMichael: "I'm pretty sure one of the changeups he threw in July just hit a catcher's mitt somewhere." Or there's this on the Pirates: "A middle infield of [Tony] Womack and [Abraham] Nunez in 1998 sounds about as appealing as the 'all-Musburger, all-the-time' network." BP acknowledges the limitations of statistics. Hitters tend to be relatively consistent and predictable from year to year, but pitchers are much more erratic, so projections for this year's hurlers are labeled WFG, for "wild fucking guess." BP marvels at the amazing things statistics can do in analyzing and even predicting performance, but it marvels more at the players who defy expectations and break through in unexpected ways. That, after all, is why they play the games on the field and not in a computer.