Keeping it dry: What would happen if all MLB parks installed humidors?

You may have heard: baseballs are stored in a humidor at Coors Field in Denver. Although the desire to suppress offense probably came from the (likely larger) effect of altitude on the flight of baseballs there, game-time humidity percentages of around 35% also lent a helping hand to hitters, assuming that the baseballs themselves matched that level of humidity. Rawlings stores its balls at 50% humidity and 70 degrees, and now the Rockies do, too.

Dr. Alan Nathan has studied the probable effect of a humidor on home runs at Coors Field. There appears to have been a change in how the humidor balls were handled in 2005; when Dr. Nathan compared the home runs per game at Coors before the humidor (1995-2001) to home runs per game from 2005 through 2010, he found a reduction of 32%, closely matching his calculations (30 ±9%, later revised to 27.5 ±4.3%). Please refer to Dr. Nathan's article at Baseball Prospectus from four years ago, which I won't reconstitute here.

The most pressing humidor issue in baseball is that of the Diamondbacks; the discrepancy between Phoenix humidity and humidor humidity is far and away the greatest of any MLB city, more than double that of Denver. It seems that on his arrival, former GM Kevin Towers quashed a humidor proposal that the D-backs' front office had considered (after all, the interim GM that preceded him knew what it was like to pitch at Coors). With Towers gone and in search of solutions for a chronically under-performing pitching staff, I expect the new Arizona front-office-by-committee to take a close look at the possibility of using a humidor.

Having addressed the idea that it's time to install a humidor at Chase Field and the probable consequences of doing so, I turn now to a different question: what if the other 28 teams installed humidors of their own?

Calling shenanigans

Depending on which measurement you use (more on that later), it appears that all MLB cities other than Denver and Phoenix hover mostly above the humidity percentage of a humidor. When the Rockies made a change in how the baseballs were handled, players on other teams noticed. But it shouldn't have stuck out that much (unless BP and pre-game activities were done with dry balls), because all the humidor did was better simulate conditions in the player's home city. In 2006, Rockie-turned-Brewer Jeff Cirillo thought the humidor balls were "bigger, heavier, and visibly oblong".. But his manager was having none of the story:

"Both teams are playing with the same balls," Brewers manager Ned Yost said. "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."

Now that we know that Yost probably disapproves of me trying to discuss humidors here, I might as well push ahead and take issue with his contention. Even if a team is not able to manage giving separate sets of balls to their pitchers and visiting pitchers, they might be able to switch during the game; depending on whether a team was behind or ahead, it might be in the team's interest to "run out" of humidor balls in hopes of spurring a rally.

Thanks in part to the eminently sportsmanlike Tim Lincecum, those kinds of shenanigans are no longer possible at Coors, where the humidor and its balls are now overseen by MLB, directly. But think about it: They are oh, so possible at the other 29 parks right now.

Where are baseballs currently stored? Are they all stored in the same place at every park? If the Red Sox staggered back home from a road trip with a heavily taxed pitching staff, could they get a benefit from pulling the next day's baseballs from "auxiliary" storage outside by the swampy Fens, instead of from inside, where the HVAC system is working at full non-steam with the possible aid of a de-humidifier or two in the air vents? Why not? If a catcher would think to one-hop practice throws to second just to add a little scuff to the ball, would you put it past a team to manipulate the environment of baseballs a bit to gain an edge, with so much on the line? Is it so different from watering down an infield when playing a speedy team? One way to cut down on those possibilities could be to make ball humidity standard throughout the majors.

Promoting offense with humidors

I'm not as smart as Dr. Nathan and I don't have his expertise, but in the analysis in the main text of the BP article, Dr. Nathan says that the "experimental data show that the dependence of both weight and [coefficient of restitution] on relative humidity is also linear." If all factors are either constant or linear, than the effect of humidity on the distance of home run balls could be literally a question of degrees.

In fact, that's just what Dr. Nathan relied on in briefly sketching out the possible effect of a humidor at Chase Field. Just as he cautioned there that the calculations were rough, let's acknowledge here that the following figures are just estimates. I also need to include the separate caveat that when Dr. Nathan updated his HR-reduction numbers in the comments of that article, I don't know what he did differently to arrive at them; the Coors Field and Chase Field estimates were not reduced in lockstep (the Chase number has a 1.35 multiplier on Coors, instead of 1.5), but overall they didn't change dramatically.

To look at how the use of a humidor might affect offense in each MLB city, I'm using humidity data from the National Oceanic and Atmospheric Administration that seems to match up with what Dr. Nathan used. It's annualized (good), using averages of all measurements (every three hours) for each calendar month; the numbers below are the average of the April through September monthly averages. The batted ball distance is only for home runs, and is based on the fact that Dr. Nathan came up with a 13.2 foot change (at Coors) for a fifteen percentage point change in humidity. I'm including Buffalo as a maybe-not-terrible replacement for Toronto, and throwing Oakland in with San Francisco.

City	Humidity (%)	HR distance (feet)
Atlanta	53	2.8
Baltimore	52	1.9
Boston	59	7.9
Buffalo	56	5.6
Chicago	53	2.9
Cincinnati	53	2.6
Cleveland	57	6.2
Dallas	47	-2.6
Denver	35	-13.1
Detroit	53	2.5
Houston	56	5.0
Kansas City	53	2.3
Los Angeles	67	14.7
Miami	65	12.8
Milwaukee	60	8.9
Minneapolis	50	0.1
New York	53	2.2
Philadelphia	53	2.5
Phoenix	18	-28.5
Pittsburgh	52	1.6
San Diego	64	12.6
San Francisco	59	8.1
Seattle	53	2.9
St. Louis	52	1.3
Tampa	59	8.2
Washington DC	52	1.6

Phoenix is the clearest outlier, but, again, I've written about the implication of that elsewhere. The only other cities where it appears a humidor might suppress offense are Denver and Dallas. Most MLB cities are all within three percentage points of humidor humidity, with Astros, Indians and possibly Blue Jays on the wait-maybe-we-should-care-about-this bubble.

It's the Red Sox, Giants, Rays, Brewers, Padres, Marlins and Dodgers who are best positioned to give offense a boost with a humidor. The latter three might even see an effect as positive as the negative one we've seen in Denver.

With apologies to Ned Yost, there are two other humidor issues I'd like to discuss — the latter of which may suggest that the benefits to offense might be even more extreme.

The trouble with ramping up home runs

As Dr. Nathan observed, drawing back home run distance is possible because the height of the wall that was cleared is typically built in already. But we have a problem in estimating home run increases that we don't necessarily have with reductions: even if a ball fell one foot short of a wall, we can't infer just from that fact that the wall would have been cleared with additional feet of what would be distance for home runs.

This is a problem for which I do not yet have a solution. And if there is one, it seems to me that it'll be on a park-by-park basis, which is beyond the scope of this article. When Dr. Nathan says that the number of home runs at Coors Field is decreased by 27.5% for a 15-degree change in humidity, I wish I could turn around and say that there would be an increase in home runs of 27.5% at Miami, and a pickup even greater than that in LA. But I really don't know, and the effect might be greater when it's positive distance.

So, especially in light of the possibility that the solution may need to be discovered on a park-by-park basis, I pass that baton to our wonderful community of saberists: how would you solve this problem? (And drop me a line!)

How quickly do baseballs change their humidity?

In using a 35% humidity figure for Denver and a 20% figure for Phoenix, Dr. Nathan appears to be using afternoon measurements of humidity. That makes sense; that measurement is taken at 5pm local time, a good proxy for the mostly-night-but-occasionally-day start times for games. But humidity appears to cycle everywhere, every day. I see three published humidity percentages from NOAA (all annualized): morning, afternoon, and average. Morning is measured at 8 am local time. The average is a daily average of all measurements, which are taken three hours apart.

The cycle appears to be that morning humidity is higher, with the level dropping during the day and eventually picking up steam (see what I did there?) overnight. That means that the afternoon figures used by Dr. Nathan probably are as low as they could be. I think the afternoon percentage is clearly the best one to use in measuring air density and the flight of the ball during games, but I'm not convinced it's the best one for this exercise.

Saying that a ball left around at Fenway Park will be at afternoon humidity at the afternoon time is saying that the ball gains or loses humidity at the same rate as air, right? I'm not sure I buy that. Ultimately, this is a thermodynamics question that probably cannot be adequately answered without an experiment in laboratory conditions. I'm not in a position to do that.

I am in a position, however, to at least speculate that the stickiness of a baseball is sticky, and doesn't change as quickly as the air around it. If a baseball takes more than a day to acclimate to its environment, then it seems to me that the average humidity measurement would be more useful than the afternoon one. Chances are it is in between the average and the afternoon. Here's how the table above looks if we use the average:

City	Aft. Humidity (%)	HR distance (feet)	Avg. Humidity (%)	HR distance (feet)	Humidity delta	distance delta
Atlanta	53	2.8	70	17.7	17	14.9
Baltimore	52	1.9	67	15.3	15	13.4
Boston	59	7.9	68	16.1	9	8.2
Buffalo	56	5.6	70	17.3	14	11.7
Chicago	53	2.9	68	15.5	15	12.6
Cincinnati	53	2.6	69	17.0	16	14.4
Cleveland	57	6.2	70	17.5	13	11.3
Dallas	47	-2.6	65	12.9	18	15.5
Denver	35	-13.1	50	-0.3	15	12.8
Detroit	53	2.5	69	16.4	16	13.9
Houston	56	5.0	74	21.3	18	16.3
Kansas City	53	2.3	68	15.7	15	13.4
Los Angeles	67	14.7	75	21.9	8	7.2
Miami	65	12.8	74	21.1	9	8.3
Milwaukee	60	8.9	72	19.1	12	10.3
Minneapolis	50	0.1	65	12.9	15	12.8
New York	53	2.2	64	11.9	11	9.7
Philadelphia	53	2.5	68	15.4	15	12.9
Phoenix	18	-28.5	29	-18.6	11	9.9
Pittsburgh	52	1.6	67	14.8	15	13.2
San Diego	64	12.6	72	19.7	8	7.1
San Francisco	59	8.1	72	19.5	13	11.4
Seattle	53	2.9	69	16.6	16	13.7
St. Louis	52	1.3	68	15.8	16	14.5
Tampa	59	8.2	74	21.4	15	13.2
Washington DC	52	1.6	66	13.9	14	12.3

Two things from this table. One is that the effect of a humidor could be significant, really, in any MLB city. The other is that daily changes in humidity aren't consistent across the league; if it is true that the afternoon humidity measurements don't fully capture the actual state of a non-humidor baseball, then the discrepancies won't be uniform.

I do think the true state of a non-humidor baseball is likely to be between the average and afternoon measurements. Whatever it was, it does appear that the Rockies changed their humidor procedure in 2005. But they didn't change the humidor's settings; therefore, the only explanation for the difference must have to do with how quickly a baseball acclimates to its environment. It could be that removal closer to game time is the main reason, or it could be a longer length of storage time. Either way, it's a question of the baseball retaining dryness or humidity.

And at the end of the day, all we need to know to know that there's some kind of lag is that the Coors Field humidor works. If baseballs immediately acclimated to air, then humidors would be completely ineffectual.

Where do we go from here?

I have a ton of further questions, but unfortunately, the answers depend mostly on details about how baseballs are currently stored throughout MLB and on the solution to that question about how quickly a baseball gains or loses wetness.

1. In many MLB cities, humidity is a lot different in the first half of the season than the second. Does that matter?

2. If humidity of baseballs plays a role in batted ball distance, then it's already baked into park factor. But how much of park factor is humidity? If we knew that, then we might be able to make park factor more accurate for a given day, simply by noting humidity levels.

3. This really gets back to the point about needing to know how baseballs are currently stored at every park, but are some teams already insulated from the possibility that their balls are more humid than they would be if stored in a humidor? I'm thinking about the Blue Jays, and especially the Rays.

4. If it's raining, then humidity is damn near 100%, which means that the air is holding as much water vapor as it possibly can. In rainy games, are baseballs — even the ones that never touch the ground — just completely different from dry balls?

5. How does the humidity of a baseball affect pitching? Anecdotal reports suggest a drier ball is harder to grip, and dryness could affect the effect of spin, as well. It's possible that this affects the game more than a rise or fall in home run rate.

Tons of questions.

The humidor at Coors Field was precipitated not necessarily by a huge humidity problem, but by the effect of elevation; it's just that the humidity issue was the solvable one of the two. There's a "something extra" reason for the Diamondbacks to install one at Chase Field: high heat can have about half as big an effect on air density as the elevation in Denver. But this may not be the last you hear about humidors at other venues. At the very least, there is reason to think that humidors could promote offense throughout the league, and with the recent change in run environment, that may start to look very attractive.

. . .

Ryan P. Morrison is a writer and editor at Beyond The Box Score. He writes about the Arizona Diamondbacks at Inside the 'Zona, and talks D-backs and sabermetrics with co-author Jeff Wiser on The Pool Shot. Follow him on Twitter: @InsidetheZona.