August 4, 2000

Doctoring the Numbers

Coors Field Revisited

by Rany Jazayerli and Keith Woolner

Printer-friendly

Just read your "Doctoring the Numbers" piece on the sucking Rockies. It would be good to see how non-Rockies teams fare on road trips. Perhaps fatigue, stinky underwear, the cumulative effects of restaurant food or some other aspect that builds over a road trip makes all teams hit significantly worse later in a road trip. And perhaps the "adjusting to hitting outside of Coors" effect does ameliorate this unspecified "long road trip" effect.

In other (simpler) words, you left out the control group.

--Mark Stevenson

The Baseball Prospectus readership ranks as one of the most intelligent demographics in the country. I was reminded of that fact once again last week, when within hours of posting last week's Doctoring the Numbers, I received literally dozens of e-mails almost indistinguishable from the one above: readers asking whether the Rockies' decline in production during road trips was all that significant, or whether it was simply a natural phenomenon that all teams endure the longer they are away from home. Some wondered, in fact, whether the average team declined more than the Rockies did during a road trip, which would then lend credence to Jeff Cirillo's theory that the Rockies' offense would improve as they got reacclimated to sea-level pitching.

Well, we've got the data, so let's take a look at it. Here is how the Rockies' offensive performance on road trips in 1998-99 (as showed last week) compares to the Houston Astros from the same two years. The Astros are an interesting comp in that they were the Rockies' diametric opposite: a good team playing in a great pitchers' park.

             Rockies                              Astros

#   Trips   AVG   OBP   SLG   OPS     #   Trips   AVG   OBP   SLG   OPS

1     24   .260  .315  .403   718     1     25   .280  .357  .451   808
2     24   .247  .306  .410   716     2     25   .257  .330  .404   734
3     23   .251  .308  .367   675     3     25   .279  .362  .452   814
4     19   .263  .321  .403   724     4     22   .244  .316  .365   681
5     19   .247  .300  .403   703     5     22   .271  .339  .399   738
6     16   .237  .289  .383   672     6     17   .273  .341  .448   789
7     10   .242  .288  .373   661     7     11   .303  .373  .515   888
8      8   .233  .297  .374   671     8      6   .267  .362  .436   798
9      6   .216  .259  .385   644     9      4   .314  .413  .453   866
10     4   .210  .280  .363   643     10     2   .329  .370  .382   752
11+    7   .312  .363  .435   798     11+    2   .337  .385  .650  1035

While the Rockies' offensive numbers appear to drop slightly as a road trip progresses, no clear pattern emerges for the Astros: while their offense dips during the middle games, they actually hit their best from the seventh game of a road trip on. Let's combine the numbers into three-game increments to get a better picture:

             Rockies                              Astros

#   Trips   AVG   OBP   SLG   OPS     #   Trips   AVG   OBP   SLG   OPS

1-3   71   .253  .310  .394   704     1-3   75   .272  .350  .435   785
4-6   54   .250  .304  .397   701     4-6   61   .262  .331  .400   731
7-9   24   .232  .284  .376   660     7-9   21   .295  .377  .482   859
10+   11   .275  .333  .410   743     10+    4   .333  .378  .517   895

Looked at this way, the Rockies' offense appears to decline steadily until the tenth game of a road trip--in a very small sample size--while the Astros are at their best in the late stages of a road trip.

It's an interesting one-on-one comparison, but making the Rockies go head-to-head with one of the best teams in baseball (remember, we're not looking at 2000 stats) may not be fair to them. Plus, it's just a one-team sample. Let's look at how the Rockies stack up against all other MLB teams combined:

             Rockies                             Rest of MLB

#   Trips   AVG   OBP   SLG   OPS     #     Trips   AVG   OBP   SLG   OPS

1-3   71   .253  .310  .394   704     1-3    2206  .265  .331  .421   752
4-6   54   .250  .304  .397   701     4-6    1647  .266  .333  .423   756
7-9   24   .232  .284  .376   660     7-9     645  .265  .330  .423   753
10+   11   .275  .333  .410   743     10+     170  .266  .328  .428   756

The consistency with which the rest of baseball hits during a road trip is astonishing. You would expect a swing of much greater than four points of OPS based on chance alone. There is simply no evidence that your typical team plays worse the longer they have to endure hotel mattresses and airplane food.

Frankly, there's no significant evidence that the Rockies are adversely affected by prolonged road trips either. But the hypothesis we're testing isn't whether the Rockies play worse the longer they're on the road; it's whether they play better.

And they don't. So we're back to our original conclusion: the Rockies don't struggle on the road because they have trouble adjusting to the way pitches move at sea level. The Rockies struggle on the road because they suck.

First, let me say that BP rocks, and I have particularly enjoyed your work since I found the site.

Now on to the first point: Jeff Cirillo does not suck.

Ignoring this season at altitude, he's got a career 835 OPS, not a superstar, perhaps, but somewhere above the suck-point for a competent defensive third baseman, I would think. Most teams not owned by Ted Turner would be pretty happy with a thrid baseman who'd posted an OBP over .390 in three of the last four seasons, scoring 95 runs or better in those three seasons, with Jeromy Burnitz and not much else to drive him in. If he'd stayed with the Brewers, and posted his career average for an OPS, he'd be fourth in the NL among qualifying third basemen, in Phil Nevin/Aaron Boone territory (though neither of those guys have career numbers that measure up to his).

Furthermore, Cirillo never really displayed any substantial home/road split as a Brewer. He'd be 50 points up one way one year, and 50 points up the other way the next, but on average they were about the same. This year, of course, he's got a huge split, he's an outrageous 1153 at home, and an atrocious 620 on the road (70 points worse than his worst full season ever, a 126 at-bat tryout in his first year in the majors). That's a pretty huge difference for a guy playing in most of the same NL parks against most of the same NL pitchers he faced last year. It's not the biggest sample (163 road at-bats), but I think it merits an attempted explanation beyond, "He sucks."

--Roy Noah

Roy was not the only reader to take offense to the flippant suggestion that the Rockies suck. First off, a clarification: the Rockies as a whole suck. They have scored just 189 runs on the road, by far the fewest in baseball. But not every Rockies hitter sucks. In particular, I did not mean to imply that Jeff Cirillo sucks. Cirillo has been one of the best third basemen in baseball for five years; that his OPS this year (886) is just 24 points higher than it was last year (862) is a sign that he's having an off-year, not that he's a bad player.

The greater point here is that Cirillo is having the worst year of his career...in his first year with the Rockies. Is this an isolated phenomenon, or a sign of a more insidious effect on hitters by Coors Field and its predecessor, Mile High Stadium? Andres Galarraga, who we universally derided as a product of high altitude during his time with the Rockies, joined the Braves at age 37 and hit as well as ever. Maybe there is some sort of altitude-induced malaise at sea level that takes a lot longer than just a week or two to shake off.

Let's look at Cirillo's numbers before and after joining the Rockies and compare them to the other prominent hitters the Rockies have acquired. We'll look at players who played at Mile High Stadium as well as at Coors Field, since the theory we're testing here involves the altitude of their home park, which didn't change when they moved to Blake Street in 1995.

I combined the players' statistics for the two years prior to joining the Rockies, in order to get a more established level of performance on the road. In the cases of Andres Galarraga and Dante Bichette, I also compared their last season in Colorado with their first season with their new team (since Galarraga missed all of 1999 and Bichette is still in his first season with the Reds, it was impossible to generate a two-year average of their post-Rockie performance).

Cirillo:

Year     Team    Home         Away

1998-99  MIL     333/414/445  315/395/461
2000     COL     421/490/650  210/274/324

Larry Walker:

Year     Team    Home         Away

1993-94  MON     317/406/578  267/369/474
1995     COL     343/401/730  268/361/484

Andres Galarraga:

Year     Team    Home         Away

1991-92  MON/STL 244/278/363  215/271/336
1993     COL     402/430/647  328/368/544

1997     COL     342/406/611  295/372/560
1998     ATL     315/422/570  296/375/615

Dante Bichette:

Year     Team    Home         Away

1991-92  MIL     245/289/386  274/297/410
1993     COL     373/408/650  252/291/410

1999     COL     308/363/575  287/342/502
2000     CIN     309/367/539  271/324/401

Ellis Burks:

Year     Team    Home         Away

1992-93  BOS/CHW 277/360/460  260/330/410
1994-95  COL     316/398/636  258/343/480

(Because Burks was hurt and played only sparingly in his first year with the Rockies, I combined his numbers for his first two years with Colorado. I would have looked at his post-Rockie performance as well, but he was traded from the Rockies to the Giants in mid-season in 1998, and unfortunately I do not have his home/road splits broken down that year for before and after the trade.)

While Cirillo's performance on the road has been much worse than it was in previous seasons, none of the other major acquisitions by the Rockies had a similar split. Both Walker and Bichette hit essentially as well on the as before, though one could argue that since offense around the majors has gone up steadily over the last eight years, their relative performance has slipped slightly.

Ellis Burks's road numbers went up after joining Colorado, and Galarraga's improvement from the stiff who hit .219 and .243 in 1991 and 1992, not slugging .400 in either season, to the league leader in batting average in 1993, had very little to do with the thin air of Colorado: his road OPS jumped more than 300 points.

If Cirillo's poor performance on the road this season was a reflection of his new home environment, he would be the first Rockies hitter to be affected in this way. Similarly, if that were true then we would expect the road performance of a Rockie hitter to improve after he moved to a new team. Galarraga's road numbers with the Braves in 1998 were only marginally better than they were with the Rockies in 1997, and Bichette is actually hitting much worse on the road this year than last.

The enigma that is Coors Field is an enduring one, and there may not be any simple answers to explain its effects. Instead, we must examine what goes on there on a case-by-case basis. And in this case, it appears that however high altitude affects hitters, those effects occur only at high altitude. There is no evidence that suggests there is a hangover effect; there is no reason to think that playing half your games at high altitude affects how you play at low altitude.

Rany Jazayerli can be reached at ranyj@baseballprospectus.com.

Rany Jazayerli is an author of Baseball Prospectus. 
Click here to see Rany's other articles. You can contact Rany by clicking here
Keith Woolner is an author of Baseball Prospectus. 
Click here to see Keith's other articles. You can contact Keith by clicking here