Pulling a Rockies Pitching Solution Out of Thin Air

The success of the Colorado Rockies franchise has historically been impeded by air: the thin air of Coors Field and the hot air blown by higher-ups in the front office.

Due in large part to playing their home games in a comically extreme hitters’ park, the Rockies have finished 14th or worse in the National League in runs allowed per game in 21 of their 28 seasons in franchise history. Colorado has finished with a winning record five times in the past 20 seasons, and in four of those they ranked in the top 10 in the NL in RA/G. No, their run prevention as a whole has never been what you would call “good” or even “well above average,” but their only brushes with success have come at times when their pitching ventured beyond putrid.

The adverse effect of the thin atmosphere on pitching is twofold. The more apparent aspect is that it imparts less drag on a batted ball, allowing for fly balls to carry further, resulting in increased slugging at Coors. Perhaps less obviously, movement of pitches due to the Magnus effect is diminished. At the risk of triggering memories of my undergraduate fluid dynamics course, the lift on a baseball (or any spinning sphere) is proportional to the density of the fluid it moves through. Thus, when a fastball is thrown at Coors Field, it has less “rise” (or more accurately, is less affected by gravity) than it would at other major league parks.

Does this mean that every pitcher will perform demonstrably worse if he takes up in-season residence in Denver? Well, yes, but actually no.

Not all pitchers are created equal. More specifically, not all fastballs are created equal, which brings us to 26-year-old Rockies ace Germán Márquez. On the surface, it may seem like he falls victim to the Coors Effect as much as anyone else, but some of the peripherals paint a different picture. In his major league career (all with the Rockies), Márquez’s home/road splits have produced nearly identical FIP values (3.89 vs 3.82), with an xFIP markedly lower at home. In 2019, his FIP (which is not park-adjusted) was actually lower when pitching in the mountains.

So what is it about Márquez that makes him less susceptible to the deleterious effects of Coors than the average hurler? It goes back to Dr. Magnus. Márquez’s heater, while comfortably sitting in the mid-to-upper 90s, is consistently below league average in spin rate. This is not inherently a good trait, but it means that pitching at altitude is a less disorienting experience.

Márquez is not the only one to have relative success in Colorado with such a stark disparity between his fastball velocity percentile and fastball spin rate percentile. Jon Gray, another Rockies rotation mainstay, exhibits these traits to an even greater extent than his teammate. Particularly in his excellent 2017 and 2019 seasons (both with an ERA- in the mid-70s), Gray’s fastball flew with a velocity near the 90th percentile while spinning at a 10th percentile rate. Despite the altitude, he has actually posted a lower FIP for his career at home than on the road.

The same can be said to a lesser extent about fellow Rockies pitcher Antonio Senzatela. While he has not pitched to the levels of Márquez and Gray at their peaks, he has proven to be relatively immune to the park effects of Coors while imparting subpar spin levels on his heater.

These examples beg the question of whether Colorado’s front office is aware of this and is actively trying to acquire high-velo, low-spin pitchers on the cheap, or whether these pitchers are just the ones that happened to survive the ascent to Denver.

In a recent piece outlining the shortcomings of the Rockies front office, Nick Groke and Ken Rosenthal laid out all the free agent signings made by general manager Jeff Bridich in Colorado. The pitchers from that list are shown in the following table along with each player’s percentile in fastball velocity and spin rate in the season preceding the signing.

Preceding Seasons by Pitchers Signed Under Bridich
Player Year FB Velo Percentile FB Spin Percentile fWAR for COL
Kyle Kendrick* 2015 N/A (8) N/A (47) -0.9
Jason Motte 2016 79 73 -0.3
Chad Qualls 2016 40 9 -0.1
Mike Dunn 2017 74 69 -0.1
Greg Holland 2017* 74 79 1.1
Wade Davis 2018 51 90 0.2
Jake McGee 2018 80 57 -0.6
Bryan Shaw 2018 75 90 -0.5
José Mujica** 2020 N/A N/A -0.3
SOURCE: Baseball Savant
* – Baseball Savant only provides percentiles back to 2014, so Kendrick’s parenthetical values are from the 2015 season.
** – Mujica only pitched 4 innings in 2020 and did not meet the minimum PA requirement.

No, it appears the Rockies have not been making a point of targeting pitchers that may be less affected by the move to Coors. This is accentuated by Wade Davis, who received the largest contract of this group while having fastball attributes in stark contrast to the ones in question. The players acquired recently via trade reiterate the same point.

During the 2020 season, Colorado acquired Mychal Givens via trade, and his fastball is consistently spun in the 80th percentile. In the 2021 offseason, the Rockies traded for yet another pitcher with a high-spin fastball, Robert “Bob Steve” Stephenson, whose stuff didn’t hold up in Cincinnati and isn’t particularly likely to find success further from sea level.

Last and certainly not least is the man so unfortunate as to be forever tied to a disastrous trade, Austin Gomber. He certainly has a low-spin fastball, so he fits half the mold of a “potentially decent Rockies pitcher,” but he wasn’t exactly cheap, running up a cost of fifty million American dollars and one Nolan Arenado.

In the following table are the pitchers who faced at least 200 batters from 2019-2020 with the greatest difference between their fastball velocity percentile and fastball spin rate percentile. Some of these players may fly under the radar due to their less impressive fastball rise but could be useful to the Rockies because their stuff is in theory less affected by the mountain air relative to their peers.

Velo-Spin Percentile Difference, 200+ PA, 2019-2020
Player PA FB Velo Percentile FB Spin Percentile Difference
Luis Garcia 323 96 13 83
Diego Castillo 379 94 12 82
Hector Rondón 354 90 10 80
Jimmy Cordero 270 94 16 78
Joe Kelly 268 99 22 77
Noah Syndergaard 825 98 25 73
Luis Castillo 1073 95 23 72
Zack Britton 321 78 7 71
Anthony Bass 289 79 9 70
Tayron Guerrero 216 100 31 69
Jonathan Hernández 203 97 31 67
Jairo Díaz 351 89 23 66
Luis Perdomo 370 67 3 65
Josh Hader 367 79 15 64
Clay Holmes 246 72 8 64

Gray ranks 21st on this list, Márquez 30th, and Senzatela 41st out of 450 pitchers. If the PA minimum is raised to 450, all three rank in the top 10 of 164 pitchers.

Obviously many of these pitchers (namely Syndergaard, Luis Castillo, Britton, and Hader) don’t fall into the cheap and/or available category. However, five of the guys on this list (Garcia, Rondon, Bass, Perdomo, and Holmes) were free agents just this offseason.

This of course isn’t to say that any or all of these pitchers is going to excel in 2021 or even that they would prevent the Rockies from losing 100 games this year. But for a franchise that is even more perpetually in need of pitching than most teams, finding these types of guys at the margins that can help assuage the geographical disadvantage is imperative to competing in a hellacious division, or at least shedding the role of a laughingstock. Hopefully Bridich and his increasingly sparse front office realize this before jettisoning German Márquez off to greener pastures.





4 Comments
Oldest
Newest Most Voted
Inline Feedbacks
View all comments
feddy
3 years ago

good stuff

Jimmember
3 years ago

VERY interesting. Jeff?

Lanidrac
3 years ago

Nice analysis, but don’t forget that pitching at Coors Field is about more than just fastballs. Pitchers more likely to succeed there are those who rely more on fastballs (and their variants except maybe splitters) and change-ups in general over breaking pitches. When they do throw breaking pitches, a slider tends to work at least better than a curveball.

Although, I’m not sure how effective a knuckleball would be. It would have less movement like any other breaking pitch; but I think its greatest asset, the unpredictability of its movement, should remain mostly unaffected. Does anyone have any data on this?