how
many
for
mcgwire?
orie glen, ph.d.
contributing editor
june 2 1998
80 four-baggers ?
Through 55 Cardinals games, Mark McGwire has launched 27
homers. Predicting whether he will pass Ruth and Maris has become a national pastime in
and of itself. Let's unpack Big Mac's year and see what his chances really are.
First off, we could assume that Mark will continue at his
current pace of about one homer every other game. He'd finish with 80 homers.
This would be a feat similar to Bob Beamon's having jumped
34 feet rather than 29 feet in Mexico City. Here at the Jones, while we enjoy the thought
of 80 four-baggers, we don't believe the hype.
Such prognostications rely on linear extrapolation. Taken
to its logical extreme, linear extrapolation implies that someone who hit a home run in
his last at bat should hit 500 for the season all he has to do is keep hitting 'em
at the same pace!
57 or 62 ?
Let's instead take a classical statistical approach to analyzing Big Macs odds.
In 4797 career at bats, going back to 1986, McGwire has deposited the ball in, off, or
over the outfield seats 414 times. That's a homer every 11.58 at bats. The media has made
much over the years of Mark's at bats per tater (ABPT) rate, especially the fact that it's
Ruthian and beyond.
That's a long track record. But is it his real rate as of June 3, 1998? This season, he
has 27 homers in 175 at bats, or an ABPT of 6.48. That's quite a rate jump. So, what's his
real rate? Hmmm ... more on that in a minute.
Let's assume for the moment that he's going to continue to knock one out every 11.6 at
bats, his career rate. Will he get the record?
Mark has played in 52 games, but his team has played in 55. And, all those walks count
as plate appearances, but not at bats, so he's had only 175 at bats, or 3.2 per game.
Despite his well-documented injury history, McGwire has been in the neighborhood of 500
at bats in seven of his eleven full seasons. So, let's assume he stays in the lineup, even
with his bad back to reach this record, he's going to have to play day after day,
not sit "day-to-day."
And let's assume he continues to average 3.2 at bats in the remaining 107 games. That
makes around 342 more at bats. With an ABPT of 11.6 at bats, he should hit about 30 more
homers. Adding this to the 27 that he has currently hit, that works out to 57 for the
year.
That doesn't mean 62 is out of reach, even if McGwire's still a 11.6 ABPT guy. Let's roll
the die and see if he can get there.
Think of a McGwire at bat as the toss of an 11-sided die with one face painted red, the
others painted blue. If the red side comes up, then we give Mark credit for a home run.
This is much less fun than watching gopher balls bounce off Waveland Avenue, but it's
how statisticians actuaries, accountants et al. think of the world.
If we toss the die 342 times (at bats), then the average number of times that Big Red
lands facing up is 29.53. However, it is still possible that it could appear 35 times, to
give him the magic number, 62.
We can easily calculate the likelihood of the die coming up red 35 times that
is, the probability of McGwire hitting 35 more. (This next step requires a little bit of
math, the central limit theorem. Just keep reading to cut to the chase, or jump to McGwire and the central limit theorem.)
And the answer is: The chances of this occurring are about 16 percent. If McGwire's
real rate of homers is 11.6 ABPT, and if he stays healthy, he has a one-in-six shot at
reaching 62 homers.
27 + 38
But his rate this year is 6.48 ABPT. Wow. But we have to be careful here, because the fact
is that someone who hits homers once per 11.6 at bats will have short periods where he
hits them once per six and a half.
When you maintain an ABPT of 6.5 for two months, however, that's pretty significant to
a statistician. In fact, we know there is only about a 33 percent chance that someone who
normally has 11.6 ABPT would have 6.5 ABPT for two solid months. (Jump to McGwire and ABPT #1 to see how we figure this.)
Any ol' weekend sportscaster can tell you McGwire is hitting a homer every six times
up, but what is the rate that a statistician can rely on?
Well, it ain't 11.6, and that's news, statistically speaking. (Jump to McGwire and ABPT #2 to see the math which says that,
despite his career rate, McGwire's no longer just an 11.6 ABPT guy.)
Remarkably, McGwire has taken his incredible tater rate to a new level. So what is it?
While it's theoretically possible that the new rate is 6.5 ABPT, it's extremely unlikely.
Increases in walks, mounting pressure as he gets closer to Maris's record, and nagging
injuries make it more likely is that he will continue to slug at about the rate he's
maintained since 1991: a homer every 9.07 at bats.
Given our estimate of 342 at bats the rest of the way, that works out to another 38
dingers. And 27 + 38 = 65.
Congrats, Mark!
But, just for fun, let's play the games anyway.
Orie Glen is the
pen name of Scott Page, associate professor of Economics at the University of Iowa. In
addition to maintaining an oral history of University of Michigan football ("AC was
way better than Desmond!"), Scott runs a summer workshop in computational economics
at the Santa Fe Institute and is part of a MacArthur Foundation working group on income
inequality. He has done analyses of the economic impact of the Rose Bowl and the World
Cup. |