Making the Most of Fenway Park

How will J.D. Drew perform in Fenway (that is, if he ever actually
becomes a member of the Red Sox)? That is the question some Red Sox
fans attempted to answer recently over at the

Sons of Sam Horn Red Sox forum. It’s not an easy question to
answer, of course, but given the wealth of available information on
batted ball locations, we can at least start looking at these things.

First of all, let’s temper our expectations right from the start. Trying
to project a player’s performance into a different setting is
exceedingly complicated, and some might consider it a fool’s errand.
The depth and height of the outfield
fences is just one way a park might affect a particular hitter.
There are many other ballpark characteristics that can affect
different players in different ways: the hitting background, the
amount of foul territory and the length of the infield grass, to name
just three. And this doesn’t even address effects not associated with
the park itself: the aggressiveness of the fan base and press corps, for
example, or the quality of a city’s nightspots. I have often wondered
if Mickey Mantle might have been even greater than he was had
he played in some peaceful Midwestern town instead of under the
bright lights of the City that Never Sleeps.

Staying Within Our Game

So, let’s just take a baby step: Let’s try to figure out how Fenway
Park’s most characteristic feature, the short, high, fence in left
field, known to everybody as the Green Monster, affects different
players. Here’s a picture of the Monster, taken during the 1999 All-Star game:
note that there were no Green Monster seats back then, so you can clearly
see the screen that used to keep home run balls from crashing down
onto Landsdowne Street, just behind the wall. Many a famous home run
ball has come to rest in the screen — from Pudge’s immortal Game
6 shot-heard-round-New England to F’in Bucky’s devastating three-run
chip shot in October 1978.

As everybody knows, the distance down the left field line in Fenway is
short: 310 feet, to be precise. Not only that, but the wall cuts
straight across left field (the wall is parallel to the right field
foul line) without angling back as most outfield
fences do. That means that a fair amount of territory that is within
the field of play in most left fields is beyond the wall in Fenway.
Of course, the wall is much higher than most outfield fences, which
makes it harder to hit the ball over it.

The best way to get a handle on how the Wall influences play is to
look at hit charts of balls hit in the Wall’s vicinity.


Here is such a chart for batted balls hit near the Green Monster
during the 2005 season and, for comparison’s sake, balls hit to the same
area in Yankee Stadium. These charts have a lot going on, so
let’s take it one step at a time.

First, the black outlines
correspond to a generic ballpark that is 325 feet down the lines and
405 feet to dead center. The green line represents the position of
the Green Monster in Fenway Park. Each colored point represents a
ball hit in the air: either a fly ball or a line drive. The position
of each point shows where the ball was fielded.
Here, let me emphasize this important point:

• The points do not indicate where the ball landed, but rather where
  it was fielded.

It would be nice to know where each ball landed, but we have to work
with the data available to us. Home runs, of course, aren’t fielded
by anybody, so for home runs the location of the point shows where the
home run landed. I’ve only included hits with a location within 50
feet of the position of the Green Monster: either in front of it or
beyond it. That is why we see very few home runs in the Yankee Stadium chart:
balls that go over the left-field fence in New York go more than 50 feet beyond the green line
on the chart, hence they aren’t included.

The color of each point gives the outcome: either some kind of hit or
an out. For the Fenway chart, you can clearly see that the home runs
are going over the wall and that most of the balls fielded in front of
the wall are doubles and singles. Presumably, many of those balls hit
up on the wall somewhere and were fielded when they came back down.
Comparing the two hit charts, it’s very clear that quite a number of
Fenway home runs would have been either outs or doubles in Yankee
Stadium. Note that the black outlines have nothing to do with the
true dimensions of Yankee Stadium. I do not have the measurements
necessary to make an accurate drawing of the different ballparks, so I
just use the generic outlines for illustrative purposes.

If you look carefully at the hit locations in Fenway, you’ll probably spot a problem: as you move towards center
field there are fewer balls fielded near the base of the wall. At first I thought my positioning of the Green Monster might
be incorrect. I came up with the dimensions using the satellite photography available on Google Earth. I believe the MLB scorers
who record the hit locations use the drawing of Fenway Park that you see in their Gameday application or
in their hit charts. Here is a comparison of the two versions of Fenway:

Google Earth Image	MLB Drawing

I’ve oriented both images so the right field foul line is horizontal. Comparing the left field wall in both cases, you can see that they
differ substantially. The Google Earth photo shows that the Green Monster runs essentially parallel to the right-field line, as I mentioned above.
In the MLB drawing, however, the Monster is clearly sloping down to the right. I’m reluctant to say that I’m right and MLB is wrong, but the Google Earth
image is a photograph, after all. So, I’m forced to conclude that MLB scorers are likely using a faulty drawing of Fenway Park
to specify hit locations. This will skew the results, since (I assume) the hit locations are often determined with reference to
the position of the outfield walls.

The Wallbangers

How did the 2005 Red Sox exploit the Wall? (I restrict myself to
2005, since that is the only year for which I currently have data.)
One way to answer this question is to total up, for each player, the
number of times that a) a home run landed within 30 feet of the wall,
or b) a hit was fielded within 25 feet of the wall. The idea is that
these are balls that often would be outs in other parks, but are
either home runs or doubles (or singles) in Fenway Park.

An important point must be made here: when a ball hits off the Green
Monster it is typically fielded near the base of the wall, although
sometimes the rebound is substantial. That is why I consider any ball
fielded within 25 feet of the wall to be a possible wall-ball. Of
course, many balls are fielded near the base of the wall without
having bounced off it—plenty of balls go down the line, in the
left-center field gap or simply get by the left fielder and end
up rolling to the wall.

I do not know what fraction of balls fielded near the wall’s base
actually hit off the wall. To get some feel for this, I viewed video
of 27 wall-ball candidates hit by Kevin Millar, Bill Mueller, David Ortiz and
Doug Mirabelli. Of those 27 balls, 17 actually hit the wall, or 63%.

Ninety percent of Millar’s balls actually hit the
wall on the fly, while only 25% of Mueller’s did. The difference should not be
too surprising: Millar is a dead-pull, fly ball hitter with decent
power. Most of his wall-balls were high drives produced by an extreme
upper-cut swing. Mueller, in contrast, is more of a line-drive gap
hitter and most of his wall-ball candidates were either in the left-center
gap or else down the left-field line (especially when the
switching-hitting Mueller was batting left-handed).

Of course, this is a small sample, so I don’t really know
how many of the wall-ball candidates overall actually hit
the wall on the fly. I think 50% might be a reasonable guess: I believe my 63% number is
a bit high because three of the four players I looked at had
above-average power.

So here is the table of short home runs (landed within 30 feet of the
wall) and wall-ball candidates (doubles and singles that were fielded
near the base of the wall):

+------------------+----+------+------+------+------+
| Name             | H  | 1B   | 2B   | 3B   | HR   |
+------------------+----+------+------+------+------+
| Millar, Kevin    | 18 |    5 |    5 |    0 |    8 |
| Ramirez, Manny   | 14 |    5 |    1 |    0 |    8 |
| Ortiz, David     |  9 |    1 |    4 |    0 |    4 |
| Mueller, Bill    |  9 |    2 |    6 |    0 |    1 |
| Mirabelli, Doug  |  6 |    1 |    3 |    0 |    2 |
| Renteria, Edgar  |  6 |    0 |    5 |    0 |    1 |
| Nixon, Trot      |  5 |    1 |    4 |    0 |    0 |
| Varitek, Jason   |  5 |    0 |    4 |    0 |    1 |
| Graffanino, Tony |  5 |    0 |    3 |    0 |    2 |
| Damon, Johnny    |  4 |    0 |    4 |    0 |    0 |
| Bellhorn, Mark   |  3 |    0 |    2 |    0 |    1 |
| Olerud, John     |  3 |    2 |    1 |    0 |    0 |
| Kapler, Gabe     |  2 |    0 |    2 |    0 |    0 |
| Hyzdu, Adam      |  1 |    1 |    0 |    0 |    0 |
| Vazquez, Ramon   |  1 |    1 |    0 |    0 |    0 |
+------------------+----+------+------+------+------+

Kevin Millar really was made for Fenway, which is why I suppose the
Red Sox went to so much trouble to acquire him back in 2003. Millar
had extreme home/away splits in his three years in Boston and 2005 was no
exception. Those eight home runs that you see in the table were the
only ones he hit in Fenway in 2005; he only hit a single home run on the road
all season. He also hit 20 doubles at home, compared to eight on the
road. No wonder Millar didn’t want to leave Boston after the 2005
season.

Boston’s premier right-handed slugger, Manny Ramirez, also made good
use of the Green Monster, with eight short home runs over it and six
wall-ball candidates. Here’s something interesting: of Manny’s six
wall-balls only one went for a double. Good hustle, dude!
Oh, well, that’s just Manny being … lazy.

Lefty slugger David Ortiz and Bill Mueller each had nine balls where
the Wall potentially came into play. However, while Mueller actually
didn’t benefit much from the Wall (in 2005), as we saw above, Ortiz did: in
addition to the four home runs, Ortiz banged another three off
the wall.

Backup catcher Doug Mirabelli also did an admirable job of exploiting
the Monster in 2005, hitting the wall three times and clearing it twice in just
152 plate appearances. He also hit much better at Fenway than on the
road that year, to the tune of about 250 points of OPS.

Lefty or Righty? Conventional Wisdom is Right!

Conventional wisdom, that guy that I love to prove wrong, has it that
the Sox should stock up on right-handed power hitters to fully exploit
the Wall. But one of the things that I’ve learned since poking around with
baseball statistics is that fly balls are preferentially hit to the
opposite field, i.e., left-handed batters hit more fly balls to left
field than to right and vice versa for right-handed batters. So,
maybe you want a left-handed hitter with decent power who can loft
those fly balls over the Monster? Somebody like J.D. Drew, maybe?

Well, before we start looking at the data, there are a couple of things to keep
in mind: it’s true that a majority of fly balls hit by left-handed batters go to
left-of-center, but it’s a small majority, 54% to be precise. And if you restrict
yourself to home run fly balls, the vast majority of those are pulled, not hit the opposite
way. Furthermore, Fenway Park is quite deep in right field and overall,
Fenway significantly supresses home runs for left-handed
batters. Indeed, David Ortiz hits more home runs on the road than he
does at home.

Ok, with those ideas in mind, here are the Red Sox wallbangers, broken
out by handedness:

Batting Left-Handed
+----------------+---+------+------+------+------+
| Name           | N | 1B   | 2B   | 3B   | HR   |
+----------------+---+------+------+------+------+
| Ortiz, David   | 9 |    1 |    4 |    0 |    4 |
| Nixon, Trot    | 5 |    1 |    4 |    0 |    0 |
| Mueller, Bill  | 4 |    0 |    4 |    0 |    0 |
| Damon, Johnny  | 4 |    0 |    4 |    0 |    0 |
| Olerud, John   | 3 |    2 |    1 |    0 |    0 |
| Varitek, Jason | 3 |    0 |    3 |    0 |    0 |
| Bellhorn, Mark | 1 |    0 |    1 |    0 |    0 |
| Vazquez, Ramon | 1 |    1 |    0 |    0 |    0 |
+----------------+---+------+------+------+------+
Batting Right-Handed
+------------------+----+------+------+------+------+
| Name             | N  | 1B   | 2B   | 3B   | HR   |
+------------------+----+------+------+------+------+
| Millar, Kevin    | 18 |    5 |    5 |    0 |    8 |
| Ramirez, Manny   | 14 |    5 |    1 |    0 |    8 |
| Renteria, Edgar  |  6 |    0 |    5 |    0 |    1 |
| Mirabelli, Doug  |  6 |    1 |    3 |    0 |    2 |
| Mueller, Bill    |  5 |    2 |    2 |    0 |    1 |
| Graffanino, Tony |  5 |    0 |    3 |    0 |    2 |
| Kapler, Gabe     |  2 |    0 |    2 |    0 |    0 |
| Bellhorn, Mark   |  2 |    0 |    1 |    0 |    1 |
| Varitek, Jason   |  2 |    0 |    1 |    0 |    1 |
| Hyzdu, Adam      |  1 |    1 |    0 |    0 |    0 |
+------------------+----+------+------+------+------+

David Ortiz was the only left-handed batter to hit a short home run
over the Green Monster in 2005.
Of course, this is a tiny sample, but it looks to me like a
left-hander needs pretty prodigious power, like Ortiz, to really
exploit the Wall. Medium-power lefties like Trot Nixon, Mueller, and
Jason Varitek (when the latter two switch-hitters are swinging lefty) just can’t generate the power to
to the opposite field to consistently
bang ’em off the Wall. Remember, though, that this is a small
sample and these results may not hold up when we analyze more data.

Next Up — Out of Towners

I suppose many Red Sox fans will read this article and say, “Well,
duh!” They know that Ortiz and Ramirez and, especially, Millar have put
the Green Monster to good use. Of course, it’s nice when the data
support our perceptions, plus I like to make those plots with the
colored dots and stuff.

Next time I hope to uncover something interesting about players on
other teams and how they might be expected to perform at Fenway, at
least as regards the Green Monster. Stay tuned.

References & Resources
The hit location data that I used here was obtained at mlb.com.
I merged that data with the 2005 play-by-play data provided by
Retrosheet.