Out of Thin Air—the New Coors

You know how when you lose something important, in your desperation
to find it, you end up looking for it in the same place more than
once, even though you know it’s not there? This happened to
me recently when looking at the team pitching stats for the National
League. The Rockies were not in their usual place at the bottom of the
list. My eyes instinctively moved upward, slowly, figuring I’d see
them somewhere near the bottom. I checked the bottom of the list again
(no, still not there, idiot), after which I started randomly looking for them up and down
the ERA list. Finally, I found them: near the middle of the pack, 7th
in the NL. That may not seem so incredible, but consider this: In the
13 years of the Rockies’ existence, the team has finished dead last in
the NL in team ERA 10 times. The other three years, they finished next-to-last.

Of course, as everyone knows, Rockies pitchers are at a big
disadvantage: they have to pitch at one mile above sea level. The high
altitude leads to increased offensive levels, which leads to those
high ERA’s for Rockies pitchers. So, what’s going on this year?

Humidor (noun): a case or enclosure in which the air is kept properly humidified

If you’ve been paying attention to this sort of thing, then you’ve
probably heard about the “humidor.” The Rockies have been storing
their baseballs in a humidity-controlled environment, causing the
balls to absorb some moisture and hence lose some “bounce.” They also
presumably become slightly heavier. The result
has been that Coors has been much more pitcher-friendly than in the
past, not just for Rockies pitchers, but for opposing hurlers, as
well.

There was a lot of talk and a fair amount of complaining about the
Rockies’ humidor in the press earlier in the season. One of my favorite
articles was href="http://mlb.mlb.com/NASApp/mlb/news/article.jsp?ymd=20060801&content_id=1587911&vkey=news_mil&fext=.jsp&c_id=mil"
target="new">this one, where Jeff Cirillo claims that the Rockies
are cheating by deadening their baseballs. Now, Jeff Cirillo went to
USC, so I suppose he must be reasonably intelligent, but he sure says
some, well, silly things in the article. Talking about the baseballs
at Coors, he said, “It’s all spongy and it’s big and it’s
waterlogged. They’re illegal baseballs. They are non-flying
baseballs.” And then, “You wouldn’t think that they’d be
cheating. They are.”

Um, Jeff, isn’t there one point that you are overlooking? The fact
that … wait, Cirillo’s own manager Ned Yost put it nicely:

Both teams are playing with the same balls. It doesn’t matter if
they’re mushy, if they’re square, if they’re triangular … That’s
nuts, man. That’s just a waste of time, discussing it.

Cirillo had an answer for that: “They’re a mile up, every day,” he
said. “That’s a distinct advantage in the fact they’re used to playing
in altitude and we’re not.” That’s a good point, Jeff, but it has
nothing to do with soggy baseballs, does it? What else?
“Or, say they get behind by a bunch of runs in a game. Who’s to say
that can’t break out the non-humidor balls?” I suppose that’s true as
well, but any team could do this, couldn’t they?

Actually, back in the
day, Connie Mack, owner-manager of the Philadelphia Athletics (for a mere 50 years), was
occasionally accused of freezing baseballs to deaden them. Apparently,
if you take a ball out of the freezer a few hours before game time,
the outside of the ball returns to normal temperature, although the cold center leads
to a deader ball. In the late 1960s, manager Eddie Stanky and the grounds crew
of the Chicago White Sox were accused of deadening their baseballs by
storing them in a humidor. Curiously, everyone referred to them as
“frozen baseballs”, even though they were not frozen, merely
humidified.

Getting back to today’s subject, as you might imagine, pitchers do not necessarily agree with Cirillo’s
viewpoint. Chris Capuano, who’d given up two home runs in a loss the
previous day, remarked, “It didn’t look like like Garrett Atkins hit a
sponge ball.”

It’s amusing that Cirillo is the one complaining. His years in
Colorado marked the beginning of the
falling-off-a-cliff decline phase of his
career. After posting OPS+’s of 122 and 118 in Milwaukee the previous
two seasons, his two years in Denver (2000-2001) saw him slip to 96
and 98. Of course, his raw numbers looked pretty good, because at the
time the Coors effect was still going strong, as we’ll see below.
More recently, Cirillo went 0-for-4 in his one appearance at Coors
this year. Non-flying baseballs, no doubt.

Run Scoring in (and out of) Colorado

So, what’s going on in Colorado? Is the ball really deader or are
Rockies pitchers simply better this year, as many have stated in the
press?
The way to answer the
question is to compare the number of runs scored in Rockies home games
to the number scored in Rockies away games. This comparison ensures
that you are comparing Coors Field with the “average” NL park (it’s
not quite average, because of the unbalanced schedule, but it’s close
enough). The quality of play is the same for both sides of the
comparison, since, to a good approximation, the same players are
playing in Rockies home games and their away games. This ratio of
(total runs at home) divided by (total runs away) is the simplest form
of park factor.

rockies_pf.png
So, if
the ball is much deader in Colorado this year, we should see a
significant drop in the park factor. The graph on the right shows the
Coors Field (and Mile High Stadium) one-year park factors, calculated
with a single season’s worth of data. There are several things to
note from this graph. First, there’s quite a bit of variation in the the
park factors, despite the fact that they are all measuring the same
thing (starting in 1995, when the Rockies moved into Coors Field). One-year factors are susceptible to large
statistical fluctuations, which is why most analysts use multi-year
park factors.

Also, there has been a dip in the park factor this year, and
its current value of 1.08 is the lowest in Rockies history. A more
general downward trend seems to have started with the 2001 season. In
fact, the article linked above states that the Rockies starting using
the humidor in 2001. So, even if things are jumping around due to
statistical variation, it appears that the humidor is having a
significant effect on the offensive levels at Coors Field.

One can investigate further to see if the downward trend in the park
factor is compatible with a heavier, soggier ball. You can look at
component park factors, or how a park favors (or disfavors) different
parts of offensive output: home runs, non-home run hits and strikeouts, for
example.

rockies_component_pf.png
We expect that both home runs and non-home run hits will decrease
with the soggy ball, a point that I think needs no explanation. What
about strikeouts? Coors Field reduces strikeouts because breaking
balls don’t break as much in the thin air. And since most strikeouts
are obtained on breaking balls, there are relatively fewer strikeouts
at altitude. However, we wouldn’t expect the soggy ball to have any
effect on the number of strikeouts at Coors Field, since a bouncy dry ball and a dead wet ball
behave just the same if you miss them entirely.

The next graph shows
the component park factors for home runs, non-home run hits and strikeouts.
Looking at the trends in the “humidor era” (2001 to the present), we see pretty much what we
expected. Home runs seem to be affected the most, although non-home run hits
are also down since 2000.

This year, the home run park factor has shown only a modest
decrease, while the non-HR hit factor has dipped more strongly.
Earlier in the season, THT’s
Dave Studeman
and Baseball Prospectus’ Dan Fox
noted that fewer ground balls were getting through for base hits at Coors Field this year. This effect would reduce the non-HR hit factor,
which is what we see in the graph. I’ve also calculated the park factor for batting average on ground balls, and my results
are in agreement with the previous findings:

Coors Park Factor - Batting Average on Ground Balls
2003: 0.937
2004: 1.016
2005: 1.117
2006: 1.011

From the table we see that ground balls are about 10% less likely to go for hits at Coors this year compared to 2005. Although, the 2006 value is in line with the period 2003-2004. These may be real differences although it’s also possible that statistical fluctuations are contributing, as well. We’ll know more (one would hope) after next season.

Lastly, you can see from the graph that we
observe little change in the strikeout park factor, which agrees with
our reasoning, above.

Home/Road Splits

Another way to look at this issue is to consider the home/road splits
of the Rockies pitchers and hitters. First, let’s look at the team
splits:

              Home R/G       Road R/G
Pitchers       4.74           4.87
Hitters        5.53           4.31

Rockies pitchers have actually been a little better in the rarefied
air of Denver, while their hitters have been better at home. Either
they’ve learned how to hit that soggy sphere or maybe Cirillo is right
about the Rockies cheating. Just kidding there, I expect this
difference between pitchers and hitters is mostly due to small sample size.

Let’s have a look at some individual players to see who likes to play
in Coors and who fares better on the road. First, the starting pitchers:

              Home ERA     Road ERA    Home - Road
Francis        4.02         3.77         +0.25
Jennings       3.28         3.97         -0.69
Cook           3.73         4.62         -0.89
Fogg           6.45         4.99         +1.46
Kim            4.03         6.78         -2.75

Jason Jennings, Aaron Cook and Byung-Hyung Kim all pitched much better on the road, Josh Fogg was
happier in Coors, as was Jeff Francis, but to a lesser degree.

On the other side of the ball, we see a more familiar picture. Most
Rockies hitters still perform better at Coors Field. Here are a few
Rockies regulars:

              Home OPS     Road OPS    Home - Road
Atkins         1.011        .904       +.107
Holliday       1.130        .828       +.302
Helton          .983        .795       +.188
Hawpe           .800        .929       -.129
Carroll         .927        .591       +.336
Sullivan        .657        .796       -.239
Torrealba       .683        .787       -.104

It’s interesting, though, that Brad Hawpe, Cory Sullivan and Yorvit Torrealba hit
substantially better on the road this year.

Who Cares?

Well, you should, if you play fantasy baseball. Until this year, most
fantasy players avoided Colorado pitchers
like the plague. In my keeper league, only closer Brian Fuentes and starter Jeff Francis were owned at
the start of the season. That changed as some sharp-eyed folks started
noticing the success of Jennings and Cook, and even Kim at home.

On the flip side, if current trends continue, you might not see
mediocre offensive players putting up huge numbers in Colorado in the
future. I know that several Rockies hitters still performed much better at home
this year, but remember that the humidor has a big dial on it and that
gives Rockies management a knob for turning down the offense as much
as they please. Who knows what their plans are?

But it’s not just about going after Rockies pitchers and perhaps
devaluing their hitters somewhat. All fantasy players have
to make lineup decisions and you might want to re-think benching all
your pitchers’ starts in Coors Field next year. Likewise, you might
not want to promote Joe Average to the starting lineup, simply because
he’ll be playing four games in Colorado.

Print Friendly
 Share on Facebook0Tweet about this on Twitter0Share on Google+0Share on Reddit0Email this to someone
« Previous: Five Cups of Coffee
Next: THT Daily: Mariners snap losing streak to A’s »

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>

Current day month ye@r *