The Numbers Game


Believe it or not, there was once a time when baseball statistics meant 61 homers, 4192 hits, 56 consecutive games, and 17 beers consumed. But that was before sabremetrics—the science of baseball statistics named after the Society for American Baseball Research, or SABR. Now we can’t open up a newspaper or flip on a game without hearing about pitch counts, on-base percentage, and situational hitting. It’s changed the way we understand the game, but which numbers really mean something, and what do they mean anyway? With hot-stove league baseball in full swing, and a copy of the recently published The New Bill James Historical Baseball Abstract (hopefully) under your tree, here’s a primer on these new-school stats and the people behind them:

Abstract, The Baseball: This was the book that started it all. Originally written by Bill James in the late 1970s, while he was working as a night watchman, the Xeroxed research caught the attention of baseball fans hungry for something smarter than Street & Smith. The Baseball Abstract became an annual bestseller and gave way to even more ambitious projects like the New Historical Baseball Abstract, which rates more than 900 players from all eras. Sabrefact: Norman Mailer was among the Abstract‘s earliest fans.

Barra, Allen: The Wall Street Journal‘s By the Numbers columnist and contributor who coined SLOB, which marries the two most important hitting stats—slugging percentage is multiplied by on-base percentage. Sabrefact: In 2001, Barry Bonds posted the second-best SLOB of all time (.444), trailing only Babe Ruth’s .451 in 1920.

Caught Stealing: The classic hidden stat. Back in the day, a player would get credit for stealing a base. And when he got thrown out, he would get credit for hustling. Problem is, when a player gets nailed at second, it costs the team both a base runner and an out. Which is why, sabremetricians point out, any base stealer who isn’t successful at least two-thirds of the time is costing his team runs. Sabrefact: Over the last nine seasons of his career, Ty Cobb stole 127 bases—and was caught 99 times—for a well-below-break-even 56.2 percent.

Denver: All runs are not created equal. At, say, Shea Stadium, three runs can get you a win, but at Denver’s Coors Field, three runs is a moderately productive inning. Park Factor attempts to quantify this difference, by comparing the number of runs scored in a stadium to the league average. Sabrefact: After moving to the hitter-friendly Ballpark in Arlington—its park factor of 105 is third best in the AL—Texas Ranger hitters won three MVP awards in four years.

ERA: The most important pitching stat, at least when you’re comparing pitchers within an era— say, Mike Mussina (3.49) and Kevin Appier (3.62), as opposed to across eras (Roger Clemens (3.09) and Walter Johnson (2.16). Sabrefact: Lefty Grove led the league in ERA a phenomenal nine times.

Favorite Toy: A way to estimate a player’s chances of reaching a goal, by factoring in his distance from the goal, his progress toward the goal (for example, the weighted average of home runs over the past three seasons), and the number of years he’s likely to play. Sabrefact: Alex Rodriguez’s chance of hitting 756 home runs is roughly 26 percent.

Game Score: James’s method of determining a starting pitcher’s dominance, by adding points for innings pitched and strikeouts and subtracting them for walks, hits, and errors. Sabrefact: Showing the limitations of this kind of largely descriptive stat, the record for a nine-inning game score is held by Kerry Wood, whose 20-strikeout one-hitter eclipsed every perfect game ever thrown.

Henderson, Rickey: The ultimate sabremetric baseball player. His career batting average is only .280, but he combines the secondary attributes—walks, base running, and power—that don’t show up in batting average. And not coincidentally, he owns the career record for runs scored. Sabrefact: Henderson has more secondary bases (4886) than any player except Babe Ruth (5099).

Intentional Walk: Writers like Pete Palmer and John Thorn analyzed the unwritten managerial “book,” arguing that intentional walks often blow up in your face, and that sacrifices, while increasing your chances of scoring one run, all but kill your chances of scoring more than one. Sabrefact: On May 29, 1998, Buck Showalter issued an intentional walk to Barry Bonds with the bases loaded and two outs in the ninth, while leading 8-6. The gamble paid off when the next batter, Brent Mayne, lined out to right on a 3-2 count.

James, Bill: Arguably, along with Marvin Miller, the most influential nonplayer of the second half of the last century, James combines cheeky wit (he once described Don Zimmer thusly: “He looks like he was stuffed into his double knits by a taxidermist’s apprentice”) with iconoclastic arguments (he maintains that the DH actually increases strategy in a game, by removing knee-jerk moves). Sabrefact: The Chicago Tribune once called the Lawrence, Kansas, resident “the Mozart of baseball statisticians.”

K: Joe D. never struck out like these guys. So what? Sabremetric research shows that players with high strikeout totals really aren’t costing their team very much—what they may give up in “productive outs” they gain back by running up pitch counts and not hitting into double plays. Sabrefact: Bobby Bonds set the single-season strikeout record of 189 in 1970, but in the past two years, five players have fanned 178 times or more.

Late and Close: Clutch hitting used to be baseball’s equivalent of the Loch Ness monster—often talked about but rarely seen. “And if you look, really look at the ‘evidence’ of clutch hitting as a true ability rather than happenstance,” writes’s Rob Neyer, “you find out that, at best, it’s a bunch of blurry photos, in the form of poorly constructed studies presented by people who desperately want to believe.” More convincing evidence came in the form of large-sample stats like “late and close,” which shows a player’s batting average after the seventh inning in a situation in which the potential tying run is at least on deck. Sabrefact: A 1989 Elias Sports Bureau study pegged Tim Raines as the best clutch hitter in baseball. He was followed by Jeff Newman, Garth Iorg, Glenn Hoffman, Thad Bosley, and Larry Milbourne.

Moreno, Omar: The sabremetric Antichrist. The Pirate outfielder looked good in a uniform, but he had no power, drew no walks, and was overrated as fielder and base runner—imagine Rey Ordoñez playing the outfield. “There can be no excuse for writing his name on a lineup card,” James wrote. Sabrefact: His .343 slugging percentage is only 20 points higher than that of Pirate pitcher Rick Rhoden.

Negro Leagues: Arguably the baseball analyst’s toughest call. Just how good were Josh Gibson, Satchel Paige, or Oscar Charleston? Without reliable stats, analysts must weigh anecdotal opinion, like Johnny Mize’s assessment of Martin Dihigo: “I thought I was having a pretty good year, down there, but they were walking him to pitch to me.” Sabrefact: Bill James puts Negro Leaguers Charleston (4), Gibson (9) and Paige (17) in the top 20 in the New Historical Baseball Abstract.

On-Base Percentage: Originally created by Branch Rickey and Allan Roth in the 1950s, and declared an official stat in 1984, on-base percentage (hits, walks, and hit-by-pitches, divided by plate appearances) is the most important offensive statistic because the benchmark of an efficient offense is the ability to get players on base. Sabrefact: The career leader in on-base percentage is Ted Williams, with an astonishing .481 mark.

Pitch Count: You would have thought it would have been intuitive—that as they throw more pitches, pitchers are less effective and more prone to injuries. But until pitch counts became public knowledge, pitching coaches would leave young starters like Dwight Gooden and Fernando Valenzuela out there to throw 140 pitches on chilly April afternoons and wonder why they blew out their arms. Sabrefact: According to Rany Jazayerli of Baseball Prospectus, Oakland’s Barry Zito, Mark Mulder, and Tim Hudson each had only one first-half start last season in which he threw more than 120 pitches.

Quisenberry, Dan: The Royal reliever ranks third in the Index of Self Destructive Acts, a “garbage” stat in which Bill James combines wild pitches, hit batters, balks, and errors, behind only Robin Roberts and Pete Alexander. The highest ISDA among top pitchers? David Cone. Sabrefact: Quiz threw only four wild pitches in his career.

Range Factor: Chuck Knoblauch notwithstanding, most major leaguers can hang on to a routine pop-up or avoid throwing the ball into the stands, so fielding percentage doesn’t give a true picture of a player’s defensive abilities. A much better yardstick is “range factor,” which shows how many chances a player makes per game. Sabrefact: Cal Ripken’s career fielding percentage is better than Ozzie Smith’s (.979 to .978), but Smith’s range factor (5.02, compared with Ripken’s 4.62) means that over the course of a season, Smith would make 65 additional plays.

Secondary Average: Whereas players were once rated almost solely on the basis of their batting averages, this James-created stat—total bases plus walks and stolen bases, minus hits and times caught stealing, the whole thing divided by at-bats—summarizes how well a player does those things that don’t show up in batting average. Sabrefact: In 2001, Barry Bonds led the majors with a record .929 secondary average, while Cal Ripken (.175) and Rey Ordoñez (.165) were among the trailers.

Total Baseball: Originally the sabremetric alternative to the Macmillan Baseball Encyclopedia, adding stats like on-base percentage and caught stealing, it was one of the bellwethers of the acceptance of new-school stats when TB superceded the Big Mac as baseball’s official record book. Sabrefact: Total Baseball adjusted Ty Cobb’s hit total from 4192 to 4189, which means that neither Pete Rose nor his real record-breaking balls are enshrined in Cooperstown.

Urbina, Ugueth: The Montreal Expos’ closer is an example of the way relief pitching has changed in the past 20 years. Now a closer like Urbina is brought in only for a “save” situation, and almost never for more than six outs, a clear case in which a statistical definition is driving a managerial decision, and, not coincidentally, protecting the manager from a second-guess. Sabrefact: In Game One of the 1974 World Series, Rollie Fingers was brought in the fifth inning and pitched four and a third innings in an Oakland loss.

Vaughn, Arky: One of the by-products of new-school baseball analysts is that overlooked old-timers get their due. According to James’s new Abstract, Vaughn is the second-best shortstop of all time, behind Honus Wagner. Sabrefact: Between 1936 and 1936, Vaughn hit .351 and led the National League in walks all three years.

Whirlpool theory: James theorizes that human nature pulls teams toward mediocrity: Bad (or disappointing) teams are aggressive in addressing their problems, while good teams tend to get complacent and often fail to address their needs. Sabrefact: Despite Bud Selig’s protestations to the contrary, there’s more parity. In 2000, for the first time ever, not one team finished with a winning percentage below .400 or above .600.

Yount, Robin: What’s the most important thing you want to know about a rookie hitter? His age. Players like Yount, who break into the majors as teenagers, have a propensity to become very good major leaguers. Older rookies—say, Lou Piniella—have almost no chance to develop into Hall of Famers. Sabrefact: The top five in batting average among players through age 20 are four Hall of Famers and a shoo-in: Jimmie Foxx, Ted Williams, Alex Rodriguez, Ty Cobb, and Mel Ott.

Zminda, Don: The majordomo of STATs Inc., the Chicago-based company that enabled more detailed analysis by scoring games in a far more detailed way, including, for example, where every ball was hit. STATs Inc. also made sabremetrics a business by publishing the ubiquitous Stats Handbooks and the annual Baseball Scorecard.

Archive Highlights