Virtually everybody involved in baseball analysis believes in using
park factors in evaluating player performance. However, its a lot more
difficult to get anyone to agree on how to calculate park factors. The major
dispute revolves around how many years of data are needed to
calculate park factors. Some people argue that you need multiple years
of data to achieve a large enough sample size to have a useful adjustment,
while others maintain that there are so many year-to-year changes
– especially weather-related – that using multiple years of data wipes out
evidence of very real effects in individual seasons. Making things more
difficult are all the new and/or revamped ballparks.
Here we’ll take a look at what the 1998 data suggests about new
ballparks, changed ballparks, and a few other ballparks you might
be interested in. The park factors that have been calculated for
this article are pretty simple, calculated by dividing total runs by
both teams scored per home game by total runs scored by both teams per
road game, that are meant to provide additional information which you
can use to evaluate this season’s performances; more complex formulas,
which take into account things like the fact that home teams don’t get
road games in their home parks, are often used for more exact numerical
Let’s get to the ballparks. Statistics through Sunday, September 13th
Edison International Field – The revamped "Big A" in Anaheim underwent
massive renovations courtesy of Disney, and most of them seem to be
positive. The most significant change is the removal of the
outfield stands that were added for football. The Big A had
traditionally favored hitters by a small amount, but this year
it registers as a pitchers’ park, seemingly decreasing runs scored
by 7%. The major change seems to be that the Angels have been
able to homer far more often on the road than at home.
BOB – The Bank One Ballpark, home of the Arizona Diamondbacks.
Since the park is at a higher altitude than any other park in the
majors outside of Coors, it was widely expected that BOB would be
a good hitters’ park, though the dome factor made it harder to
predict. Anyway, BOB has shown up this year as a hitters’ park,
increasing runs scored by about 9%.
Coors Field – There’s been no change here, of course, from last
year, but most analysts are still trying to get a grip on how large
the effects actually are, since the difference between the 1996 and
1997 results were substantial. This year, the Coors park factor
settles in between the results of the last two years, increasing
runs scored by a still-whopping 50%.
Olympic Stadium – le Stade Olympique has a sorry history as a baseball
stadium, which has led to substantial changes several times
during its relatively short history. At first there was no dome.
Then the retractable dome arrived. Then the retractable dome
stopped being retractable. Then the dome began to fall apart, so,
for 1998, they took it off to fix it. The removal of the dome for
1998 seems to have temporaily altered Olympic Stadium back into
what it originally was, an extreme pitcher’s park, decreasing runs
scored by 19%. With the dome, Olympic had become a minor hitters park
in recent years, and, assuming the dome is put back on next season, it should
return to that state.
Oakland Coliseum – The word was that, in addition to destroying
the Coliseum’s virtue as a baseball venue, the reconstruction of
Oakland-Alameda had changed what had been the best pitchers’ park
in the AL into a hitters’ park. At least, that’s why they said
Mark McGwire was suddenly hitting a ton of home runs there again.
And the numbers seemed to confirm that. While the 1996 park factor
was probably useless since the park was physically under construction
during the season, the 1997 park data clearly suggested that Oakland had
turned away from its past as a pitcher’s haven. But hold on. In
1998, the Coliseum has reverted to its previous form, reducing
runs scored by about 16%. So was last year year a fluke, or what?
Let’s see in a year.
Turner Field – Presumably named after Frederick Jackson Turner
(yes, the owner’s name happens to be Turner, but who would be so egocentric to
name a stadium after himself?), its somewhat high altitude was expected to
produce a hitters’ park. Instead, last year it showed up as a very
slight pitchers park, and this year it has shown up as relatively neutral. Of
course, Coors makes it rather difficult to be a hitters’ park in the
National League by changing the context through which we evaluate ballparks for
the entire league.
Suncoast Dome – Oh, they renamed this hideous thing, didn’t they?
Its now Tropicana Field. Anyway, everyone was guessing before the
season that this would be a pitchers park, a la the Astrodome.
Not quite. Early on in the season, it looked like such a great
hitters park that some wits referred to its brand of inside baseball
as "arena baseball." The numbers, early on, seemed to indicate that
it was indeed a great hitters park in the making. But as the Devil
Rays limp towards the end of their first season, the park registers
as increasing runs by only 11% – the sign of a good hitters park, but nothing
spectacular. And the numbers look so different in the second half compared to
those of the first half that I definitely wouldn’t want to draw any strong
conclusions about the park factor here for another year.
County Stadium – The stadium didn’t change, but the league did. In
this year’s National League, County registers as a park that cuts down
runs by 3%. During Milwaukee’s last few years in the American League,
County appeared to be a moderate hitter’s park. Does this mean that
National League parks are better hitter’s parks than those in the AL?
It’s really too small a sample size to say that with any certainty, but
it’s quite possible that Coors is primarily to blame.