Jerry Daniels knows well the heartbreak of the Red Sox fan. Last October the engineering professor watched in agony as the Yankees knocked Boston out of the playoffs with a dramatic game-seven win. It was the last straw. “I decided to see what I could do to help them win,” he says.

Daniels, an associate professor of engineering, took an algorithm he’d been using to analyze human decision making and applied it to baseball, creating what he called the Batting Order Optimization Method (BOOM). Working with John Raiti, a graduate student in engineering, Daniels examined the way American League managers chose batting orders last season.

Selecting a nine-man lineup, it turns out, presents a potential 362,880 choices. BOOM rips through these in about forty minutes, analyzing 150 lineups per second as it computes which combination should generate the most runs in a game. BOOM then weighs the lineups against the stats of the opposing pitcher before churning out results that are then compared with the results of the lineup that actually played.

It’s baseball without the hunches, moneyball on steroids. BOOM sifts through more than a thousand Web pages a day, collecting nine different statistics for each player. The program also analyzes pitchers, compiling an array of numbers to calculate its own Earned Run Average. BOOM, for example, calculated that Boston could have im­proved its offense by 13 percent during the first half of this season by tweaking its lineup.

The technology behind BOOM came from years of studying unbiased optimization—computations made without guidelines, expert knowledge, or gut instinct. In the area of neural networks, for example, Daniels, a biomedical and electrical engineer, studies how nerve cells connect to carry out their tasks—when he’s not trying to save his beloved BoSox from its latest swoon, that is.

As a real-world test, Daniels and his colleagues configured the program to select the best American and National leagues starters for this year’s All-Star Game. BOOM picked just half the position players actually elected to the teams, selecting Boston’s Nomar Garciaparra over New York’s Derek Jeter, for example, even though Garciaparra had missed much of the first half of the season. Shaking off any suggestion that the program might be biased in favor of Boston, Daniels explains that BOOM only considers the at-bats a player has had, not the ones he’s missed.

The program also predicted the National League would win the game, 4–3. And things worked out that way. Sort of. While the American League actually won the game 9–4, BOOM correctly named the National League score. And, notes Daniels, if you subtract the six runs given up in the first inning by former BoSox and Yankees star Roger Clemens (whom the program, interestingly, did not pick for the team), you come up with BOOM’s forecast score.

“We picked the wrong team and the wrong score, but we got a few things right,” Daniels says, sounding more like Don Zimmer than Bill James. “That’s baseball, I guess.”