Going the Other Way

In the New Historical Baseball Abstract, Bill James wrote an
essay on the reasons for the big jump in offense that occurred
in the mid-90s and is still with us. The piece is classic James:
well-conceived and clearly presented, with plenty of interesting things
to chew on. One of the five reasons that James gives, surprisingly,
is the use of aluminum bats in amateur ball. James explains how the
aluminum bat allows players to drive an outside pitch to the opposite
field with authority, something that major leaguers generally did not
attempt with wooden bats. Here’s James:

It has always been considered a sucker’s game for a hitter to try to
drive an outside pitch. Up until 1990, young hitters were always
taught either to lay off the outside pitch or to go with it and guide
it into the opposite field. If you try to “drive” that ball, they
were told, you’re going to wind up with a ground ball to second base
(if you’re right-handed).

But players who had grown up swinging metal bats found out that that
they could benefit from the techniques they had always used. More from
James’ essay:

What hitters learned from using the aluminum bats was not that they
couldn’t hit the outside pitch hard, but that they
could.

James then recounts how he started thinking about this after a comment
by
Greg Maddux, who said that

…the biggest change in baseball since he
came into the league was in the number of hitters who stand right on
top of the plate and hit the outside pitch. He said that when he came
into the league (1986) he saw maybe a half-dozen opposite field home
runs all season. Now you see them all the time.

After checking his STATS database, James found out that Maddux was
right: the number of opposite field homers had tripled since 1987
(this was written in 2000). He also noted that the number of hit
batsmen had more than doubled, from 31 (in 1980) to 65 per 100 games
in 1999. THT’s own Steve Treder chronicles the increase
in hit batsmen in his excellent piece,
The HBP Explosion. James concludes that, “Batters have learned that they can stand right on the top of
the plate and blast away at the outside pitch.”

If James is correct, he is talking about a fundamental change in hitting technique. You
could (and James, in his essay, does) compare it to changes that came
about in the 1920s when Babe Ruth showed what you could do by
swinging hard, with a pronounced uppercut, every time.

Enter Retrosheet

This looks like something worth investigating and, thanks to
Retrosheet, we have the information we need to conduct a little
study. So, here’s what I’ve done: I’ve sifted through the Retrosheet data
from 1957 through the present, noting the number of
opposite field hits, of any kind, in each season. Then calculate the fraction of hits that went to the opposite field in each
year. If James’ hypothesis, that players in the 1990s
starting “blasting away at the outside pitch”, is correct, then we should see
an increase in the opposite field fraction starting somewhere around 1990.

First we need to define “opposite field”. Obviously, when a
right-handed batter hits the ball towards right field, that’s the
opposite field, but we need to be a little more precise. For home
runs, I rely on the hit location data provided by Retrosheet,
for which the definitions are shown here. If
a home run is in the center slice of that diagram, then I don’t
consider it. All other home runs are classified as “pulled” or
“opposite field”.

For hits other than home runs, I use the defensive position of the fielder to
tell me where the ball went. In other words, if a left-handed batter
hits a double that is fielded by the left-fielder, then that’s an
opposite field double. Any balls handled by the center fielder are not
considered.

One last caveat about the data: Hit location and the defensive position
of the fielder are not available for every single batted ball, although

for most years, there is enough data to perform this
study. The data for a few years in the mid-80s, though, is
quite incomplete, so I have eliminated them from the results that
follow.

All the way, the other way

Since Maddux pointed out the increase in opposite field home runs,
let’s look at these right off the bat.

This graphic shows the fraction
of home runs that are hit to the opposite field. I show the results
for both right- and left-handed batters to see if there was any
difference (for home runs there wasn’t).

A few things to note about this graph: A steady fraction of 10% of
home runs going to the opposite field is evident from 1957 to around
1980. It’s too bad the ’80s data has some holes, because it looks like
some stuff is going on there. The year 1987 was an anomalous
year for home runs overall and there seems to be a spike in the
opposite field variety as well. After 1990 the opposite field fraction grows for
half a decade, but has been in decline since around 1997.

We don’t see the Maddux/James factor of three increase in opposite field
homers, but we do see about a factor of two, which is a lot.
If the rise in opposite field homers reflects a real change in hitting
philosophy, though, it’s hard to explain the decline seen over the last
decade. I suspect something else must be going on.

Shorter hits

We can also look at what happened on hits other than home runs, of
course.

If batters are hitting to the opposite field with increased
authority, we should see an increase of doubles over there, as
well. Here’s the same graph as I showed above, but for doubles instead
of home runs.

Well, that’s not what I was expecting, not at all. The fraction of
doubles hit to the opposite field seems fairly constant over the last
50 years, and perhaps it is actually declining in recent years. The striking
difference between lefty and righty batters puzzled
me for a moment, but I now believe it is due to the fact that right fielders have stronger
arms then left fielders. In other words, righty batters are reluctant to take an
extra base against the strong-armed right fielders, while lefties run wild on those noodle-armed left fielders.

I checked if the difference between righties and lefties was influenced by the
Green Monster at Fenway Park. The answer is “yes”, but the effect is
very small, much too small to explain the difference seen in the graph.

Well, we’ve already looked at homers and doubles, we might as well
complete the job and look at triples and singles. Those graphs are shown
below. The fraction of triples hit to the opposite field, like the
doubles fraction, doesn’t appear to be increasing over time.
Note that for triples, in direct contrast to the doubles case, it’s easier for
righties to go to the opposite field, since the right fielder has a
much longer throw to third base than does the left fielder, who is
handling would be opposite field triples from left-handed batters. The stronger arm of right fielders
cannot make up for the longer throw.

The graph for singles shows yet a different behavior. A slowly
increasing opposite field fraction, which began back in the mid-60s,
however. And now there is no difference for left-handed and
right-handed batters, which is what we should expect, I guess. By the
way, I’ve excluded infield hits from this study.

I’ve also looked at what happened to the fraction of opposite field
outs (hit to the outfield) over time. The data for outs is complete,
we always know who fielded a caught ball. You’re probably tired of
looking at graphs by now, this is the last one, I promise.
The fraction of outs to the opposite field shows a slow rise similar
to the case for singles, although here there is a small difference for right-
and left-handed batters.

What it all means

I wish I knew! Let’s review what we’ve learned about the fraction
of balls being hit to the opposite field over time:

• Home runs: The fraction seemed to
  double from 1990 to 1996, roughly. This supports James’
  hypothesis of hitters learning to hit with power to the opposite
  field, although the decline in the opposite field fraction since 1997 doesn’t fit this scenario.

• Doubles/Triples: The fraction of opposite field doubles and triples
  has been essentially constant for the last 50 years. In other words, the data for
  doubles and triples do not support the idea that batters have learned to hit to the opposite
  field in the last 15 years.

• Singles/OF outs: There has been an increase over the last 50
  years of singles and fly outs to the opposite field. The changes have
  been gradual, though, and don’t seem to fit with the idea that hitters
  have learned to hit to the opposite field since the ’90s.

I don’t know, there doesn’t seem to be a large body of evidence here that
supports James’ idea. I have my own theory, which I admit is largely
speculation.

Walsh’s theory of opposite field hitting

Professor Robert Adair, who for a time held the position of “Official
Physicist to the National League,” studied the conditions under which a
“typical” swing results in a ball pulled down the line or hit to the
opposite field. Adair’s findings, which appear in the classic The
Physics of Baseball, explain why, on average, when a batter swings
too soon, the ball is pulled (obviously), but it also tends to be a
grounder. Late swings, on the other hand, tend to result in opposite field
fly balls. This pull/ground, opposite field/fly correlation is a
consequence of most batters having an upper-cut swing nowadays. The
data bear this out: balls hit to the opposite field tend to be
fly balls and the majority of ground balls are pulled.

So, here’s my own hypothesis: of course, batters stand closer to the
plate than they previously did. That clearly gives them better
coverage, but perhaps they generally still pull the ball as much as
they ever did. Actually, if you think about it, if you stand closer
to the plate, you’ll pull more outside pitches than you otherwise
would.

In my scenario, the increase in opposite field home runs goes
hand-in-hand with the overall increase in all home runs. Look, if
something is causing the overall home run rate to rise, as it did
fairly steadily from 1990 to 2000, e.g. smaller parks or a juicier
ball, bigger muscles and harder swings, this could easily result in
more “ordinary” fly balls going over the fence. Since balls hit to the opposite field
tend to be ordinary fly balls, I would expect the fraction of opposite
field home runs to increase as overall home runs increase. In other
words, warning track opposite field fly ball outs get turned into home
runs. Shorter fly balls are still outs, they don’t become doubles or
triples.

Or maybe not

Sounds good, I know, but there is a problem. The opposite field home
run fraction increased until 1996 or so, after which it has steadily
declined. Overall, home runs have also declined in recent years, but
that didn’t start until 2001, five years later. So maybe my theory
doesn’t hold water, but hey, at least I have something in common with
Bill James.

References & Resources
Many thanks to Retrosheet for providing the play-by-play data that I regularly use to perform studies like these.

The Physics of Baseball, by Yale physics Professor Robert K. Adair, is an excellent resource for those wanting to understand the origin of the platoon split or why a ball hit straight at an outfielder is the hardest one to judge. It’s not for everybody, but if you enjoyed your high school physics class, then this book is for you.