Doctoring the Numbers: Defense in Colorado

by Rany Jazayerli


Continuing our discussion from last week on how to build a team at Coors Field, this time, from the run-prevention side.

In Baseball Prospectus 2002, Joe Sheehan wrote:

"If putting balls in play is the best approach for hitters in Denver, then it makes sense for the Rockies to favor pitchers who strike out a bunch of guys. This is difficult: the effect of Coors Field on breaking pitches makes it hard for pitchers to get the batter to swing and miss, much less do so three times. Coors is not a good strikeout park, reducing them by 15% a year (source: Stats, Inc.)."

Last week, I made the case that putting balls in play at all costs is not sound strategy at altitude, and therefore it does not make sense for the Rockies to put an absolute premium on strikeout pitchers. Joe points out that Coors Field decreases strikeouts by a significant margin. Knowing that, wouldn't you want to avoid pitchers whose primary strength is taken away by the ballpark? Think of it this way: Randy Johnson, who strikes out 12 batters per nine innings, is going to lose roughly 1.8 of those strikeouts to balls in play. Kirk Rueter, who strikes out four batters per nine innings, is going to lose roughly 0.6 strikeouts. Certainly, the Big Unit will not suffer from Coors Field's influence on batted balls as much as Rueter will, but does that make up for the disappearance of 1.2 additional outs?

This is the exact opposite situation that we discussed from the hitter's point of view last week. Comparing Tony Gwynn and Rob Deer, we found that they both benefited approximately equally from Coors Field. Without going into the math in too much detail, the same can be said about Rueter and Johnson. While Rueter will give up more additional hits on balls in play, that will almost exactly be cancelled out by the fact that Johnson loses more strikeouts. To sum up: there is no additional advantage to concentrating on power pitchers at altitude.

Back to Joe:

"What types of pitchers are we talking about? Think early Chuck Finley and David Cone or, even better, any version of Nolan Ryan. The Rockies will have to take some chances and be patient with the pitchers they acquire. Most importantly, they'll have to adjust their mind-set on walks, which will be the hardest part of this plan. The Rockies want a staff that will strike out 1300 batters, even if it walks 700, because it's that important to keep the opposition from putting the ball in play."

The Rockies don't have Nolan Ryan, or even an early Chuck Finley. They do have someone close: Mike Hampton. In the two years before he signed with the Rockies, Hampton struck out 328 batters in 457 innings, and refused to give in to hitters: he walked 200 batters while giving up just 22 homers. If ever there was a pitcher who understood the importance of keeping the opposition from putting the ball in play, it's Hampton.

For the first two months of his $151-million contract, it looked like he was born to be a Rockie. Now, he's just another data point.

Hampton aside, do you know how hard it is to build a staff that strikes out 1,300 batters in a ballpark that decreases strikeouts by 14%? That would be the equivalent of a staff that strikes out 1,512 batters at sea level, or more than one an inning. The Chicago Cubs set a major-league record by fanning 1,344 batters last season. Saying that the solution to the Rockies' pitching staff is to strike out 1,300 batters is like saying that the solution to the Tigers' woes in Comerica Park is to hit 250 homers.

The worst part about this plan, though, is the idea that walks don't matter, that giving the batter first base is somehow preferable to letting him put the ball in play. Of course walks matter, because the opposition is going to get their base hits no matter what you do, and it matters how many guys are on base when those hits come. Your opponents are still going to hit homers 60% more often than they will anywhere else, and of course it matters whether those are solo homers or three-run jacks. How can it not matter?

I also want to touch on defense for a moment. Here's what Joe has to write:

"One of the early ideas about winning at altitude was that the Rockies would need a fast outfield, able to cover the gaps and reach balls that would be plummeting towards earth much faster than they would be elsewhere. While this isn't a bad idea, it's also possible that the physics issues may preclude anyone from being a good defender at altitude."

I think that's a fair assessment of the issue. We understand that the physics of thin air allow balls in play to travel farther and faster, reaching their final destination quicker. What we don't know for certain is whether this makes defense more important—because a player (especially an outfielder) with exceptional range will reach even more balls that lesser players might miss—or whether the balls travel so far and so fast that more of them are simply uncatchable.

This is a question that I can't answer, and no one can answer, without the right data. This is the kind of question that a statistic like STATS' Zone Rating, or Sherri Nichols' old defensive measurements, would be extremely useful in answering. What we need to know is not whether fewer balls are caught in Colorado than elsewhere—we already know that's the case—but whether the variance among players increases or decreased at Coors Field. If the difference between Andruw Jones and Carl Everett is 100 points of Zone Rating at sea level, but 150 at Coors Field, than we know that outfield defense is even more important.

There is one other factor to consider: since we know that strikeouts are reduced in Colorado, this means that balls in play are increased. Since each batter is more likely to put the ball in play, this means that there are more potential balls for the defense to field. Other things being equal, this would make defense more, not less, important.

So, putting it all together: how do you win at Coors Field? The easy answer is, there are no easy answers. We know that Coors Field cuts strikeouts by a significant margin (incidentally, it appears to have no influence on walk totals), we know that it makes it more difficult for defenses to convert balls in play into outs, and we know that both of these influences pale in the face of the park's ungodly impact on homers.

We also know, from previous work that Keith Woolner has done, that there is a modest but real benefit in having batters that hit the ball in the air, and pitchers that keep the ball on the ground.

Given what we know, this is what I believe gives the Rockies their best chance for success:

  • Don't worry about batting average. All hitters are going to see their averages rise roughly equally anyway. Worry instead about two things: walks and power. (Yes, even more than usual.) Power is more important than ever, because a 20-homer player at sea level will hit 26 homers at altitude; a five-homer player will hit just six homers. Walks are more important in the same way that they're more important when the wind is blowing out at Wrigley: with all those homers flying over the fence, you want as many guys on base as possible when they do come.

  • Get guys who hit fastballs. Remember Pedro Serrano from "Major League"? The guy who hit batting-practice fastballs halfway to Kissimmee, then swung three feet over every curveball they tossed him? You want a couple of guys like that, guys who can turn on any fastball, but have trouble with breaking stuff. They won't have to worry about pitches that bend as much at high altitude. On the road, of course, the only solution is to bench them; you can always use them as pinch-hitters against the likes of Armando Benitez. It may be wise for the Rockies to have one or two positions at which they platoon, only instead of the traditional left/right platoon they use a home/road platoon, using a power hitter with holes in his swing at home, and a more "traditional" hitter on the road.

  • Dump conventional wisdom. What the Rockies can't do is play a bunch of contact singles hitters at traditional non-power positions. They have to have eight power threats in the lineup, and putting Walt Weiss or Neifi Perez or Juan Pierre in the lineup surrenders the awesome power advantage that the environment provides. This may be the most difficult challenge the Rockies face: finding a good defensive shortstop who also hits for power.

  • Conserve outs. This should go without saying, but managerial strategy must be focused on protecting outs. The Rockies should bunt and steal less than any team in the league. The break-even rate on steal attempts at Coors Field is somewhere in the area of 80%.

Defensively, we don't know whether the difference in range among different players is accentuated or dampened by the ballpark. We do know that there will be more balls in play, and a higher percentage of hits on balls hit to the outfield will be doubles or triples. I also suspect—and this is another question that needs to be researched—that a higher percentage of balls in play are hit in the air, as players try to get some loft on their swing. Therefore, I think the early conventional wisdom, that range in the outfield is paramount, still holds. Andruw Jones might take away 50 hits a year in this ballpark. If he's not available, get Mike Cameron.

Finally, the pitching staff. There are two ways to look at Coors Field's influence on pitchers: in terms of statistics, and in terms of repertoire.

In terms of statistics, all pitchers are going to get fewer strikeouts than they will elsewhere, and they're going to give up more hits—a lot more hits—on balls in play. That leaves two things still under their control: the number of walks they give up, and the number of homers they surrender. Nolan Ryan isn't the ideal Coors Field pitcher, Tommy John is. Pitchers who get ground balls, who don't give up the homer, and who get double-play balls to eliminate all those extra baserunners are ideal; guys who never beat themselves by walking the leadoff hitter.

The Rockies' current closer, Jose Jimenez, is one such pitcher. He's an extreme groundball pitcher (3.39 G/F ratio this year) who has surrendered just three home runs in 45 innings. He's walked just nine batters all year. He has just 29 strikeouts, and he's allowed 51 hits, more than one an inning--but the lack of walks and homers has kept his ERA at 3.80. Most telling, he's pitched even better at home (3.12 ERA) than on the road. Jason Jennings is another pitcher like this; he's 9-4 with a solid 4.59 ERA.

Avoid flyball pitchers with control issues. Denny Neagle is a flyball pitcher who gives up too many walks, and his $10 million salary is now parked in the bullpen.

The primary influence of altitude on pitching is that breaking balls don't break. Pitches simply don't move the way they do at sea level, so you don't want pitchers who rely on movement to get hitters out. There are other ways to skin that cat: with velocity, with change of speed, and with deception. The first of those three is coveted by all teams, which means the Rockies should concentrate on guys who throw their pitches at a variety of speeds, and guys who throw in an unconventional manner.

The best reliever the Rockies have ever had, Steve Reed, almost never touches 90 mph, but he throws from a near-sidearm angle that makes it tough for right-handed hitters to pick him up. He also throws strikes, and when he misses, he misses low. Coors may take away your curve, but it won't take away a funky delivery that makes it difficult to pick up the ball, so sidearmers, submariners, and guys whose fastball seems to come out of their shirt are the guys to chase. Mike Myers had a 2.62 ERA in Coors Field the past three years, because there's nothing altitude or thin air can do about a guy who throws left-handed from four o'clock.

They don't want Nolan Ryan. They want Kent Tekulve. Since he's not available, take Chad Bradford instead.

After spending all this time talking about how to use their ballpark to their advantage, though, it must be said that perhaps the most important thing for the Rockies to do if they ever want to win at Coors Field is to find a way to lessen the impact of their ballpark. Most ballparks with unique characteristics present an opportunity for the home team to build an immense home-field advantage, and certainly it seems that way with the Rockies, who have had very good home/road differentials throughout their history. But as long as Coors Field exists in its current form, there are two structural handicaps that the Rockies must deal with no matter what they do:

  • Their hitters seem to do much worse on the road than would be expected from their pre-Coors performance. Many people think this is because, after seeing a steady diet of pitches that don't break very much at home, their hitters simply aren't prepared to see full-breaking curve balls when they hit the road. A previous study we did failed to show that Rockies hitters got their groove back as a road trip progressed; still, it's something to consider (and another reason to have different hitters starting at home and on the road).

  • Their pitchers, by virtue of pitching in a run-friendly environment, are forced to throw more pitches than any other staff in baseball. This means that either the Rockies' hurlers must throw more pitches—risking injury—or that fewer of their innings are thrown by their front-line talent.

These structural handicaps can only be muted by altering the physics that make Coors Field the hitters' park that it is, which is why the humidor, while of only limited benefit, is nevertheless an excellent idea. It's not Dan O'Dowd's job to give baseball fans a unique experience, it's his job to win. If the humidor makes Coors Field a little more like every other ballpark, it's going to help him do his job that much better.

Rany Jazayerli is an author of Baseball Prospectus. You can contact him by clicking here.


Baseball Prospectus Home