Copyright 2004 National Post, All Rights Reserved
National Post (Canada)

June 3, 2004 Thursday Toronto Edition

SECTION: Sports; Pg. S2

LENGTH: 808 words

HEADLINE: Survey says ... for best results, pitch to Bonds: Intentional
walks costly: Number-crunching by professor reveals flaw in strategy

SOURCE: CanWest News Service

BYLINE: Gary Mason

BODY:
Someone once said that smoking is the leading cause of statistics.

Wrong.

Baseball is. And the latest proof is on the line from his office in Chapel
Hill, N.C., with a message to major-league managers.

Pitch to Barry Bonds.

"A graduate student and I were talking one day about baseball and, in
particular, about Barry Bonds and the decision so many teams are taking to
walk him when he comes to bat," Jerry Reiter is saying.

"And it went from there."

Bonds holds the Major League record for most walks in a season -- 198 --
established in 2002. Through 2003, he held the National League record for
career walks -- 2,070 -- as well as the crown for most consecutive games
with one or more walks -- 18.

In 2002, Bonds also established a new Major League mark for most
intentional walks -- 68 --and through the 2003 season he also held the
league record for most intentional walks in a career -- 484.

Reiter, assistant professor at Duke University's Institute of Statistics
and Decision Sciences, decided to test the growing strategy to walk Bonds
by crunching the numbers.

The 34-year-old had his brainstorm near the end of the 2002 season and
waited until it was over before beginning to collect his data. He used
CBS's baseball Web site, which had pitch-by-pitch accounts of every game
Bonds had played in the last couple of years, as his source. The detailed
information allowed Reiter, and a small band of graduate students helping
him, to examine every possible situation.

For instance, they looked at what had happened right before the first pitch
to Bonds and what happened right after it. In the end, they were able to
record what occurred in each of Bonds' plate appearances and what happened
right after.

Because Reiter wanted three years worth of statistics to give his study
added validity, he waited until the conclusion of the 2003 season before
throwing all the numbers into the computer.

And what did he find?

"When Bonds walks with the bases empty and no outs -- which accounted for
roughly 80 plate appearances over the period we studied -- the Giants
scored at least one run 46% of the time and averaged about 0.9 runs per
inning," says Reiter, who figures the study took three months in total to
complete.

"When opposing teams pitched to Bonds in that situation -- which amounted
to roughly 300 at-bats over the three years -- the Giants scored at least
one run 36% of the time and averaged about 0.6 runs per inning."

In other words, avoiding Bonds's home-run power by giving him a free pass
to first base wasn't as effective as pitching to the brute.

Reiter analyzed other batting situations, too, and found that with at least
one runner on base, pitching to Bonds was still better, percentage wise,
than walking him no matter how many outs there were. It's only when there
was no one on base and at least one out that the odds changed.

"Here the numbers showed that walking Barry Bonds was more effective than
pitching to him," said Reiter. "In that situation the risk that Bonds hits
a home run outweighed the risk that he scored when put on first base."

Reiter knew that given his subject matter, his thesis and conclusion would
be subject to more scrutiny than the U.S. Army's conduct in Iraq. He
understood there'd be some pencil-necker somewhere saying: what about the
pitcher? Maybe better pitchers were pitching to Bonds and poorer pitchers
were walking him. And what about the guy coming up after? Maybe they were
walking the Giants slugger to get to the next player in the order.

Couldn't those factors confound the results?

"So I looked at that too," says Reiter.

Reiter examined the quality of every pitcher (based on his career
earned-run average) Bonds faced over the period of the study against
innings he was walked versus innings he was pitched to. His finding? It
didn't seem to make any difference who was staring down on him from the
mound.

The results were pretty much the same.

Neither did it seem to matter who was following Bonds in the order.

"Benito Santiago batted after Bonds 20% of the time he was walked and 21%
of the time he was pitched to so it wasn't like they were walking Bonds so
they could face a certain player after him," says Reiter.

Now Reiter, who wrote an article on this subject for the January 2004 issue
of the Baseball Research Journal, wants you to know he is a Red Sox fan.
Which is probably why he has gone public with this information. Were Boston
playing in the National League like the Giants, his findings would likely
been quietly leaked to members of the Red Sox coaching staff.

And it wouldn't have done Boston a bit of good.

Maybe Reiter's next project should be looking at the best way for his
beloved Beantowners to reach the World Series. Then again, that's an easy
one.

Get out of the American League.

(The Vancouver Sun)