Tracking on-base percentage has become an obsession for baseball managers
and fans ever since Michael Lewis’ best-selling book, “Moneyball,” and the film
of the same name promoted this stat as the magic number for run production. But
does this formula live up to its fame?
Daniel N. Deli, assistant professor of finance
at DePaul University, tested Lewis’s hypothesis by examining 27 years of Major
League Baseball data in a study published in the Journal of Sports
Economics this spring. Here, Deli talks about the surprising results:
What was the most significant result of your research?
Since Michael Lewis’s book, “Moneyball: The Art of Winning an Unfair Game,” was
published, the conventional wisdom has been that getting on base is more
important for scoring runs than hitting for power. Various estimates of the
relative importance of getting on base put it at somewhere between one and
one-half, to three times more important than hitting for power.
My research shows that is not the case. At best, getting on base is equally
important to run production as hitting for power. And there is evidence that
hitting for power may actually be more important for run than getting on base.
Do your findings refute Moneyball’s hypothesis, which is based on how
the Oakland A’s won 103 games and a division championship in 2002 despite
having one of the lowest team payrolls?
According to the hypothesis, the Oakland Athletics’ success was the result of
two things. First, they understood that getting on base was more important than
hitting for power while other teams did not (or they understood that it was
more important than other teams realized). Second, they realized that
increasing on base percentage could be done more cheaply than increasing power.
My research only focuses on the first part of the hypothesis and concludes that
it is incorrect. That does not mean that the hypothesis is entirely wrong,
For example, let’s assume that getting on base and hitting for power are
equally important to scoring runs, but that it is cheaper, in terms of player
salaries, to increase getting on base than to increase hitting for power. A low
budget team like the Athletics, then, might be able to compete with
better-funded teams by taking advantage of the relative under-valuation of
getting on base versus hitting for power. My research does not address player
compensation, so it cannot really say anything about that part of the
hypothesis. The question does, however, suggest a potentially interesting
avenue for future research.
Based on your findings, what could the Cubs or Sox do to improve their
prospects this season (and beyond)?
One of the problems with how people have come at the question of getting on
base versus hitting for power is that they treat it as an either-or situation.
They frame the question along the lines of “Which would you rather improve,
your team’s on-base percentage or its slugging percentage?” That framing of the
question is overly narrow, and likely a bit misguided.
What earlier researchers have ignored, but that also comes out in the results
of my study, is that the value of each of the two, in terms of run production,
depends on the other. Getting on base is more valuable if there are people
behind you in the batting order who can drive you home. Likewise, hitting for
power is more valuable when there are people in the batting order ahead of you
who can get on base.
I suspect that most teams understand the relationship between getting on base
and hitting for power, but to the extent that the do not, my recommendation
would be that they should not focus on improving one while ignoring the other.
As a finance professor, what interested you in doing a study about
Over the past 20 years or so there has been a tremendous increase in the amount
of statistical analysis in baseball. That research, broadly referred to as
Sabermetrics, has done a lot to advance our understanding of the game. For me,
though, I thought there might be a problem with the conclusion that getting on
base was more important than hitting for power.
My concerns stemmed from research that I had done (with three other colleagues)
looking at the relationship between mutual fund performance and flows of money
into those funds. In that study we were comparing how investors in different
types of mutual funds respond to “good” performance. What we soon realized,
however, was that “good” performance for one type of mutual fund is not
necessarily the same as “good” performance for another type of fund. For
example, outperforming a benchmark by one percent is a lot easier for a foreign
equity fund than it is for a domestic government bond fund. Because we were
interested in examining how investors responded to comparable amounts of
performance across fund types, we had to develop tools to make performance
comparable across fund types.
There is a similar intuition that holds with respect getting on base versus
hitting for power. Increasing a team’s on-base percentage is a lot harder than
increasing its slugging percentage. Once you make an adjustment for the
relative difficulty of increasing both measures you see that getting on base is
not more important to scoring runs than hitting for power. If anything, as I
said earlier, hitting for power is more important than getting on base.
Deli’s study, “Assessing the Relative Importance of Inputs to a Production
Function: Getting on Base versus Hitting for Power,” appears in the April issue of the Journal of Sports Economics.