August 3, 2006
Something's Rotten in Den...verThanks to all of you who checked in with assistance on the Tejada/Tehama thing. I've added the name of the Orioles' shortstop to my Word dictionary, which has solved the problem. As always, writing for people smarter than yourself has its privileges.
The Associated Press reported a story yesterday that pushed me to do research I'd been wanting to do for a while. Tuesday afternoon, Brewers utility infielder Jeff Cirillo pointed out what should have been obvious for some time: that the Rockies' use of a humidor for storing game balls has gone past the point of a minor correction for atmospheric conditions and become a means to creating a pitchers' park. Cirillo cited little more than the way a ball felt in his hand and second-hand comments by his teammates, but he did add this:
"It's pretty dramatic, wouldn't you say? Most shutouts in Coors Field, in the National League, when you take into account Petco, Dodger Stadium, where Washington plays. Those are huge parks."Cirillo is no dummy. Long regarded as a heady player, he attended one of the nation's finest universities, and has been in the majors for 13 seasons, the last four as a part-time player. That stat quoted above is completely true and utterly mindboggling. If you'd told someone in June of 2004 that two years hence, one of every six or so games at Coors Field would end in a shutout, you would have been laughed out of the room.
How is this happening? Those close to the Rockies will cite a crop of young pitchers developing in concert, a younger team featuring an improved defense, higher grass and even a grand design before they'll mention the humidor.
The humidor was first used in 2002 in an attempt by the team to control scoring in one of the greatest run environments in MLB history. According to the linked story by the Denver Post's Mike Klis, the Rockies began storing their baseballs in a climate-controlled room at 40% humidity to keep the balls from drying out in Denver's thinner, drier air. The idea wasn't to make the balls dead, but to make them more like the ones used at lower altitudes. There was a great hue and cry over the issue, but Coors Field remained a great hitters' park, the best one in baseball from 2002-2005, albeit not quite as extreme as it had been before. The Rockies, set astray by The Great Change-up Experiment in 2001 (Mike Hampton and Denny Neagle were signed to $170 million worth of contracts and proceeded to implode), weren't any more successful in those years than they'd been in the previous four.
Two months ago, Dan Fox looked into the issue and relayed that the Rockies were leaving the baseballs in the humidor for a longer period of time, perhaps a factor in the early-season dropoff in runs. Following up, Fox concluded that the humidor and the longer grass at Coors Field were driving run scoring down. Mitchel Litchman used distance data to illustrate the impact of the humidor on fly balls. (If you're not reading the blogMGL writes with Tom Tango and Andy Dolphin, you should be.)
Whatever the Rockies did this season, it has had an astonishing effect. What follows are Coors Field's one-year park factors since 2001, per ESPN.com, and its rank in MLB:
Runs Rk HR Rk Hits Rk 2001 1458 2 1457 2 1270 4 2002 1440 1 1600 1 1232 3 2003 1243 4 1369 2 1146 2 2004 1412 1 1235 4 1240 1 2005 1285 1 1119 8 1254 1 2006 973 17 1114 9 1021 11In any given season, one park or another might have been better for some event than Coors Field was--81 games isn't a large enough sample to overcome variance, which is why most analysts use multi-year park factors for rigorous work-but no park was consistently in the top five of all three categories. You can see a drop in the home-run factor in 2005, a dip that may have foreshadowed things to come.
This year, the changed used of the humidor has made Coors Field, at more than 5,000 feet above sea level, a pitchers' park. There's no question, none at all, that Jeff Cirillo is onto something here. Coors Field has become a below-average place for run scoring, despite the thinner air at high altitude, and with no change in the park.
Wading in behind MGL and Dan Fox is a dangerous game, but I want to examine some data to see if it's possible to tease out whether the dive in run scoring at Coors Field really is about the young pitchers and the tall grass, or whether it has more to do with the baseballs. What I wanted to find out was how the distribution of balls hit by players at Coors Field has changed. I'm not talking about traditional park factors here: the story they tell is pretty clear and laid out above. I'm talking about the actual impact of hitting a baseball stored in a humidor long enough to, from a baseball standpoint, turn black into white.
BP staffer William Burke patiently assembled the data you're about to see. One of these days, the people who do the real work around here are going to go all Animals on our collective Mr. Jones, but until then, I'm going to keep leaning on them. Burke's research was a critical element in this piece. Keith Woolner was a considerable help as well. Burke assembled all of the statistics generated in each NL park for the last six seasons. Again, I point out that these are not park factors, and as such, may be corrupted to some extent by a team's personnel. I would caution you against misusing this data in the practice of evaluatng players.
What happens when a bat hits a ball at Coors Field? In Baseball Prospectus 2002, I found that not striking out was more valuable at Coors Field--where in 2001 batters hit .378 with a .649 slugging average when making contact--than anywhere else. Rany Jazayerli notably disagreed, and I'm four years and a month late with my rebuttal. It's on the list; suffice to say I remain certain that the way to win at altitude is to make a lot of hard contact and have pitchers who keep the ball out of play.
Unless, of course, they replaces the horsehide baseballs with Nerf. Here are the ranks of Coors Field among all 16 NL parks--not a constant cast--in an assortment of metrics that illustrate the impact of deadening the baseballs:
XBH/CT HR/CT XBH/FB HR/FB BABIP 2001 1 1 1 1 1 2002 1 2 1 1 1 2003 1 2 1 1 1 2004 1 5 1 1 1 2005 4 9 3 7 1 2006 14 15 8 13 10 XBH/CT: extra-base hits per ball contacted (XBH/(AB-SO)) HR/CT: home runs per ball contacted (HR/(AB/SO) XBH/FB: extra-base hits per flyball HR/FB: home runs per flyball BABIP: batting average on balls in playFor years, Coors Field was the best place in baseball to make contact. You hit the ball, good things happen, more than they did anywhere. There was a wobble last season, which might well have been written off as a personnel matter--the Rockies had improved pitching and a wretched offense--if it hadn't been followed up by this year's line. Coors Field is now average in turning batted balls into extra-base hits and homers, and slightly above when it comes to doing the same with flyballs. For the first time I can recall, it's also not the safest place for a ball in play. Even conceding that the above are park stats and not park factors, the extremes listed make a compelling argument that the natural order of things has been badly upset.
Think about why Coors Field is such a good hitters' park. The air in Denver is thinner than it is at sea level, so it provides less resistance, less drag, on struck balls. Robert Adair, author of The Physics of Baseball, estimated this effect at 9% for an average fly ball. This is the biggest reason why Denver is such a great place to hit. Another reason is that the same thin air makes it harder to throw good breaking stuff, because breaking balls need that same resistance to do their dancing. Finally, a common complaint by pitchers was the dry Denver air made it hard to grip the baseball, adding to problems with command, especially of breaking balls, as the baseballs become slicker, harder and, according to some, smaller.
Now, soaking the baseballs in milk, or whatever it is they're doing, addresses the first of those problems and the last. Heavier baseballs mean more resistance, which is one reason why there's a cap--5 ¼ ounces--on the weight of a baseball. But they shouldn't have much effect on the second other that some improvement in pitchers' ability to grip the ball. The thin air is still not going to mix real well with breaking pitches.
There aren't a lot of great objective ways to measure this. I would speculate that if pitchers were having an easier time with the baseballs, that should show up in their command numbers, their strikeouts and walks. So as above, let's looks at the ranks of Coors Field in command indicators over the past six years. Again, this are park stats, and subject to skewing by personnel.
K/PA BB/PA K/BB 2001 16 6 16 2002 16 9 16 2003 14 8 14 2004 15 1 15 2005 14 6 14 2006 14 10 12As you can see, Coors Field has been a terrible place for pitchers to get results on their own. If the humidor is tuned to cancel out the effect of dryness--a problem that was only raised to crisis levels once the humidor came to be--it shouldn't be enough to turn Coors Field into a pitchers' park. Adair's 9% should still be affecting things. The humidor is overcorrecting for the dryness, having a minor effect on command, but the real driver of run prevention is that the humidor is producing baseballs that simply don't travel.
Now, the Rockies don't want any part of this. They want to point to Jeff Francis and Aaron Cook and Jason Jennings and say that the lowered run scoring is the end result of having good pitchers. If this is about personnel, then the Rockies' pitchers should be showing improvement at home and on the road, and in their peripherals as well as their ERAs.
IP ERA BFP/K BFP/BB K/BB BFP/H BFP/HR BFP/XBH 2006 494.0 4.03 6.44 12.95 2.01 4.23 37.04 12.87 2005 735.0 5.18 6.56 12.61 1.92 3.86 39.48 11.55 2004 733.0 6.27 7.02 10.32 1.47 3.78 31.15 10.57 2003 737.0 5.07 6.92 14.52 2.10 3.86 27.93 10.61ROAD
IP ERA BFP/K BFP/BB K/BB BFP/H BFP/HR BFP/XBH 2006 451.2 4.32 6.52 14.12 2.17 4.20 56.05 12.58 2005 683.2 5.07 6.45 10.69 1.66 4.15 33.73 11.08 2004 702.1 4.77 6.77 11.06 1.63 4.27 35.33 12.29 2003 683.0 5.35 7.86 11.22 1.43 3.96 37.30 11.86We're on to something here. This is an improved pitching staff. Look at the road numbers: Rockies' pitchers have cut their ERAs, improved their command and dramatically improved their home-run rate from one season to the next. You can see the effects of the humidor at home in the vastly decreased hit rate and extra-base hit rate, but it's clear that this Rockies' staff is not entirely the product of wet baseballs.
I wanted to delve into some groundball/flyball stuff here, as much to compare to 2004 as 2005. However, the data we have in those categories doesn't compare well back to '04, the difference being tracking all balls versus just outs. I thought the Rockies had become more of a groundball staff this year, but there's virtually no difference between their 2005 and 2006 GB/FB ratios. I'm comfortable saying that this is a better staff, but I'm not convinced that it's a ton better, in part because the peripherals of the key guys aren't great.
Francis, for example, has a K/BB of 88/54 in 138 1/3 innings, nothing special. Is his ERA 3.58 because of that, or because he has an insane .202 BABIP allowed at home? His extra-base-hit rates are about the same home and away as well, which is the humidor at play. In 2005, he allowed a .467 SLG and .159 isolated power at home; in 2006 those numbers are .345 and .157. Aaron Cook has a .278 BABIP at Coors. Jennings' slugging allowed at home has gone from .437 to .369.
Over two years, the difference is even more stark. The Rockies allowed a .501 slugging and .194 isolated power at home in 2004. This year, those numbers are .420 and .155. Some percentage of that is the pitchers, and how you assign credit among the hurlers and the humidor depends largely on your proximity to Blake Street.
Earlier this year, I wrote about how the use of the humidor was a mistake because it homogenized the baseball world. While conceding that playing at altitude presented challenges for the Rockies, I lamented the loss of a unique environment. At the time, I didn't think that Coors would play neutral all season long, and I was right; it's become a better pitchers' park as the year has worn on.
I don't blame the Rockies at all. They're convinced that their best chance to win is to deaden the baseballs and create a nine-run environment. That they had their most success as a franchise when they hit the snot out of the ball--while getting adequate starting pitching and great relief work--hasn't registered. There's nothing wrong with them asking for permission to alter the baseballs in a manner that fits their needs. However, they shouldn't be able to turn a ballpark one mile high into what it's become today, a slight pitchers' park. MLB, which approved the humidor plan to begin with, has to step in and restore order. There's no way--no way--that Coors Field should play neutral. You want to bring it with a certain distance of the other parks, that's fine, but you can't mess with the baseballs so much that it cancels out all the run-enhancing effects of the altitude.
Of course, we're edging close to some of the biases I've written about in the past. Lower run environments are seen as "better" than higher ones, and winning with "pitching and defense" is held in highest regard. It's acceptable for a team to attempt to lower its run environment; it would be considered insane for a team to try and raise the same. If the Mariners or Padres petitioned to keep their baseballs extra dry to make up for the difficulty in hitting balls out of their yards, they would almost certainly be rejected, but they'd also be vilified.
My question is, how acceptable is this kind of thing? The closest MLB environment to Colorado is Phoenix, a dry region well above sea level. Chase Field has played as a very good hitters' park over the years, and fairly consistently so. Can the D'backs go to MLB and request a humidifier to bring their baseballs in line with everyone else's? For that matter, the Rangers have a hitters' park and three terrific pitching prospects coming up. Can they point to their run environment and ask for relief?. How far from the norm do you have to be to warrant a change, and how much change is acceptable?
In the case of the Rockies, they've implemented a solution that has radically changed the way the games are played at home. If the massive effects of the humidor are acceptable to MLB, then I believe it opens the door for teams looking to make less-radical changes. What I'd prefer, though, is for the league to step in and take a stand on how wet is too wet. Well, what I'd prefer is for no tampering with the baseballs at all, but I don't think that will happen.
What I do know is that Jeff Cirillo is right. The baseballs in use at Coors Field have been modified in a way that completely changes the game played there. Where batted balls used to go for hits, extra-base hits and home runs, they now become outs. If you want to give some credit to the Rockies' pitchers, you can, but the primary reason is the humidor. Cirillo's conspiracy theories are wrong--there's no swapping of dry and wet baseballs depending on the Rockies' situation--but he's right to point out what should have been obvious all along.