Duke Professor To Major League Managers: Pitch To Barry Bonds

This strategy seems sound when facing a man who is on track to break the career home-run record before he retires. But a Duke University statistician said he has run the numbers and discovered that in most cases the managers are wrong.

"I examined the number of runs scored by the Giants after each one of Bonds' plate appearances in 2001 through 2003," said Jerry Reiter, assistant professor of the practice at the

Institute of Statistics and Decision Sciences at Duke.

"When Bonds walks with the bases empty and no outs -- roughly 80 plate appearances over the three years -- the Giants score at least one run 46 percent of the time and average about 0.9 runsper inning. When he hits in these situations -- roughly 300 at-bats over the three years -- the Giants score at least one run 36 percent of the time and average about 0.6 runs per inning.

"These differences suggest that avoiding Bonds' home run power is outweighed by beginning an inning with a free pass. Pitching to Bonds is more effective than walking him in this situation."

Reiter analyzed other batting situations and found that with at least one runner on base, "pitching to Bonds still seems to be the better strategy no matter how many outs there are."

It's only when there is no one on base and at least one out that the situation changes.

"Here, statistics show that walking Bonds is more effective than pitching to him," said Reiter, who wrote an article on the subject in the January 2004 issue of the

Baseball Research Journal,

published by the

Society for American Baseball Research.

"The risk that Bonds hits a home run now appears to outweigh the risk that he scores when put on first base."

Reiter said these statistics were controlled for other factors, such as the quality of the pitcher and the player batting after Bonds.

So the next time Bonds starts an inning with a walk, Giants' fans should not boo the opposing manager. He's actually doing their team a favor.