This article was published in Scientific American’s former blog network and reflects the views of the author, not necessarily those of Scientific American
The National League Championship Series gets under way this evening when the Los Angeles Dodgers host the Philadelphia Phillies, and tomorrow the American League follows suit as the Los Angeles Angels of Anaheim visit the New York Yankees.
The outcome of these all-important seven-game series, which feed into the World Series, is anyone's guess, but one mathematician who has made educated guesses on baseball games for years predicts that the teams vying for baseball's biggest prize will be wearing Dodger blue and Yankee pinstripes.
On supporting science journalism
If you're enjoying this article, consider supporting our award-winning journalism by subscribing. By purchasing a subscription you are helping to ensure the future of impactful stories about the discoveries and ideas shaping our world today.
Bruce Bukiet, a professor of mathematics at the New Jersey Institute of Technology, simulates the outcome of individual games using a method first described in a paper (pdf) he co-authored in Operations Research in 1997. Since then, Bukiet says he has made minor refinements to the system—approximating the effects of in-game substitutions, for instance—but has kept it largely intact.
The core of the model is that a batter steps to the plate in one of 24 unique circumstances—eight different possible arrangements of base runners and, for each of those arrangements, three possible numbers of outs (zero, one or two). By determining the probability of the batter taking any of those 24 states to any other (or to the 25th state, which is recording the inning-ending third out), the game can be simulated step-by-step.
According to the algorithm, the Yankees have a 67.5 percent chance of nabbing the American League crown this year, whereas the Dodgers enter their play-off series with a 63.2 percent chance of winning. Bukiet's predictions relied on incomplete early information on which pitchers will start which games—for instance, he had the Dodgers' Clayton Kershaw facing off against Philadelphia's Cole Hamels in game two, but in fact both left-handers have been tapped to take the mound tonight in game one.
Although Dodger fans and Yankee lovers may enjoy seeing one expert's analysis tilt in their favor, past years have not always played out according to Bukiet's predictions. "A couple years ago, all four winners of the division series were the teams I said should be the underdogs," he says, noting that he took a great deal of ribbing that fall. And last year's eventual World Series champs, the Phillies, were heavy underdogs (33.6 percent) in their championship series against the Dodgers and were not projected to beat (40.8 percent) the Tampa Bay Rays in the World Series.
Bukiet says his primary missions are "to promote the idea that math is useful all around us and interest students in studying math," so even when the statistical predictions don't align with the actual outcomes, he considers the effort a success. Besides, the probabilities are just that—indications of probable outcomes—and are hardly guarantees of a particular result in a notoriously unpredictable enterprise such as baseball. "I like to joke that it is not my 'predictions' that are incorrect," Bukiet says. "It is just that the players haven't performed the way they should have."
Photo of the Dodgers and Yankees playing in the first game at Ebbets Field in Brooklyn in 1913, when the two teams were crosstown rivals: Library of Congress