BOSTON — Baseball fever is building this month with pitchers and catchers reporting for spring training and pitching great Roger Clemens dodging questions from members of Congress about performance-enhancing drugs.
Off the field, meanwhile, the art and science of prediction is coming of age. Super stat-masters are getting better and better at predicting the best baseball players and even the best managers for a team.
High performing offensive players, such as home run kings and RBI leaders, tend to garner the big contracts and fan ratings. But statisticians presenting here today at the annual meeting of the American Association for the Advancement of Science have figured out how to measure trickier and sometimes more important aspects of the game, such as fielding performance (the probability of catching and throwing well enough to make an out) and managerial decisions (such as taking out a tired starting pitcher before he gives up a game-losing grand slam).
Shane T. Jensen of the University of Pennsylvania has come up with a statistical way to compare the fielding performance of every Major League Baseball player with the average performance for all players in each position. Using stats from the 2002 to 2005 seasons and dividing the field up into smaller zones for analysis than previous modelers have done, he confirmed the opinions of many of fans and commentators about the abilities of well-known fielders.
Coco Crisp, Gary Matthews Jr., Reed Johnson and Andruw Jones are among the best outfielders, whereas Manny Ramirez, Bobby Higginson and Wily Mo Pena are among the worst.
"Things like hitting and pitching are a little easier to quantify because they are easier to tabulate, there are a finite number of outcomes," Jensen said. "Fielding is a harder endeavor" to evaluate, he said, because players are ranging toward a ball in play on a continuous surface.
Some revealing results came up when Jensen looked at shortstops. Alex Rodriguez and Clint Barnes came up as top fielders, while Derek Jeter and Michael Young ranked among the top 10 worst fielders.
For the New York Yankees then, the irony is that A-Rod has one of the best fielding performances among shortstops, Jensen's stats show, (based on Rodriguez's 2002-2003 season with the Texas Rangers), but he currently plays out of position at third base for the team in deference to one of the worst defensive shortstops (team captain Jeter).
But Jeter might not need to watch his back now that this research is out.
"I don't think we're on Derek Jeter's radar," Jensen said.
Steve C. Wang of Swarthmore College said he has come up with a way to use 2007 season statistics to give a detailed picture of managers' strategic tendencies. For instance, do they tend to leave pitchers in or yank them quickly? Do they tend to signal for bunts or stolen bases? Do they tend to leave relief pitchers in for longer than one inning?
Wang's statistical approach allows a comparison of managers to see which ones cluster together in terms of managerial style.
One irony he uncovered is a comparison of former Boston Red Sox Manager Grady Little and current manager Terry Francona, who succeeded Little in 2004, the first year the team won the World Series in 86 years. Little was widely criticized for leaving pitcher Pedro Martinez in too long in the 2003 American League Championship Series against the Yankees. But Francona measures about the same when it comes to tendency to leave in a starting pitcher, Wang said.
Wang's comparative managerial analysis could be useful to teams trying to assess which manager might best fit a team's needs, he said.
Some analysts are trying to evaluate the influence of performance-enhancing drugs on baseball players' stats, but it's nearly impossible to make the link stick statistically, Jensen said.
Nonetheless, statistics can suggest which players one might want to test more than others, said presenter David Pinto, a former ESPN analyst who now writes primarily for his blog BaseballMusings.com.
"I think we know that if we see someone in his 30s having a huge career surge, that should raise a red flag," Pinto said. "That can happen, but if it happens over two to three years, I might want to test him every month rather than twice a year."
Pinto, Wang and Jensen said baseball fans will see more focus in the future on fielding statistics.
Some observers say the Boston Red Sox (World Series Champions also in 2007) rely more heavily on statistics to guide strategy than other teams, however, baseball teams have always leaned heavily on statistics to analyze player performance, said New York Times columnist Alan Schwarz, a discussant during the presentations.
"They just don't make quite as much a fuss of it, fans don't make as much a fuss about it and their teams aren't the World Champions," he said.