Coors-and-Effect: How Altitude Reshapes College Arsenals
It’s no secret that elevation leads to runs. Coors Field has created exceptional amounts of offense each year since 1995, and there’s been no shortage of coverage on how usually-steady pitchers become unraveled by the park’s huge gaps, thin air, and strange effect on pitch movement.
In professional baseball, altitude is generally a solved problem for orgs with strong R&D and scouting departments. Opposing front offices spend hours preparing their roster for their 1–2 series a year in Denver. However, the college game presents a lot of questions about our version of baseball at altitude, and there aren’t any multi-million dollar departments around to answer them. How does the increased drag of the college baseball respond at elevation? How much does the lack of humidor use increase fly ball carry? What pitches remain steady when I run a pitcher out at a 7,000-foot high ballpark, and what pitch types should I avoid calling? At Paradigm, we see altitude as a unique challenge to weaponize data and create altitude resilience in our partner programs when they go on the road.
First, to measure the effects of elevation in college baseball, we pulled four seasons worth of data at five of the highest ballparks in the league.
The difficult part of a study like this is separating the park from the pitcher, since strong and weak staffs visit different teams in different years. We chose to look only at pitching staffs that had traveled to parks at elevation and compared their performance up high to their outings on the road near sea level. We then used RPI to control for opponent quality in both environments.
By setting the study up this way, we were able to remove the effect of home-field advantage and standardize opponent quality to isolate exclusively for elevation’s effect on a visiting pitching staff. After cleaning the data, we had a sample of more than 53,000 pitches thrown by visitors up in the mountains against roughly 200,000 of those same pitchers’ offerings at sea level.
1. Air It Out
Two things happen to a baseball at altitude, and they compound. The first is that the baseball moves less on the way to the plate. Because a pitch’s movement is created from the seams of the baseball “paddling” the ball through the air, the thinner air up high cuts into pitch movement because there’s simply less air for the seams to paddle against.
4,500 feet of elevation decreases overall pitch movement by about 16%, meaning that a riding fastball with 20 inches of induced vertical break at sea level loses over three inches of carry at Utah Valley. Air Force’s Erdle Field is roughly 7,000 feet above sea level, which cuts total pitch movement by roughly 27% and regresses that same 20 IVB fastball even further to a 15 IVB meatball.
At the same time, fly balls carry five to six percent farther than they would at sea level (and 7 to 8 percent at Air Force), meaning that a well-struck, 400-foot flyout to deep center at a sea-level park becomes a 420+ foot bomb at elevation. Another factor in this is that programs that play at elevation generally don’t store their balls in a humidor like some MLB teams do, with BYU being a possible exception.
Interestingly, fly balls at BYU’s Larry H. Miller Field carry only about 2.5% further than sea level, half as much gain as Utah Valley sees just four miles down the road. This effect is likely due to BYU’s use of a humidor to climate-control their baseballs, though that can’t be confirmed based on data alone.
It’s intuitive that less movement and more fly ball carry create an offensive explosion similar to what MLB offenses see at Coors Field, and that played out in our analysis.
When pitching near sea level, our visiting arms had a solid wOBA against of .373, but that figure climbed to .420 in the mountains. This forty-seven point jump represents the gap in skill between an average Saturday starter and a fringe midweek reliever. This degradation of ability meant that staffs in our sample got torched for 10.2 runs per game pitching at altitude against 7.8 runs per game at sea level.
2. Pick Your Poison
Interestingly, this elevation-based damage is not spread evenly across an arsenal, and the split between pitch groups is the most useful finding in our study. Sorting each pitch type by how much its wOBA rose at altitude splits the repertoire cleanly in two. Cutters and breaking balls were the least affected pitch types at elevation, while fastball shapes and changeups were heavily taxed.
There are two methods that combat elevation in these pitch types. First, pitches with gyroscopic (not spin-induced) movement like tight sliders and cutters hold their shape because they don’t need to paddle through the air at all. A traditional gyro slider with around 0 inches of induced vertical break and limited horizontal movement would move about the same at 30,000 feet as it would in Death Valley. Gyros don’t need the air to move because they effectively don’t move, they just drop according to the time they spend in the air.
But what about depthier curveballs or big sweepers? Larger breaking ball shapes that do rely on spin-induced movement have their shapes taxed the same as any other pitch. In fact, breaking balls in our sample actually lost a higher percentage of their movement than fastballs did. So why are big spinners still a decent option at elevation?
Get ParadigmPDS’s stories in your inbox
Join Medium for free to get updates from this writer.
This likely comes down to the suboptimal contact quality they induce as well as their inherent ability to miss barrels entirely. At elevation, the backspun fly ball is the ultimate evil. Breaking balls (that aren’t completely hung over the middle of the plate) usually avoid becoming backspun fly balls, often inducing hard-hit but topspun contact to the pullside or harmless sliced fly balls backside.
On top of that, you’d be hard-pressed to find a hitter that wants to hit spin more than he wants to hit a heater, even when accounting for less movement on the breaking ball. Whiffs are the single most valuable outcome at elevation because of the heightened risk of any ball in play, and breaking balls will always induce whiffs.
3. Chump Change
The changeup is the ugly surprise in the original chart, with a whopping 69 points of wOBA added at elevation against the pitch. In theory, changeups should be fairly well insulated from elevation’s effects. Although changeups might induce more backspun fly balls than breaking balls, they create similar levels of whiff and chase as spin. In fact, on the surface level, changeup/splitter shapes should degrade less than breaking balls since they usually aren’t as reliant on raw movement as they are on deception. So why did the changeups in our sample get hammered?
Our first idea was location-based; visiting pitchers who aren’t used to pitching at elevation leave the changeup up just a little more. This was proven false in the data: the average changeup location didn’t change in our sample, with better than sixty percent of them thrown below the zone in both settings. More granular location testing revealed no difference in effective location (tested by our Command+ model) or shadow-zone rates.
Diving further in, we checked changeup performance split by batter handedness, and there it was: in our sample, altitude created a massive gulf between same-handed and opposite-handed changeups.
Surprisingly, our sample’s changeups degraded more than twice as much against opposite-handed hitters at altitude as they did against same-handed hitters. Changeups are often cited as a uniquely platoon-resistant pitch that hold their value versus opposite-handed hitters. So why are they so bad at elevation?
If we reference our 17% movement loss rule-of-thumb, we know that changeups thrown at elevation are losing about three inches of run on average. In a same-handed matchup, that’s not significant due to bat path. Armside run isn’t really doing much for a right-on-right changeup because the right-handed hitter’s swing path follows a runny, fading changeup’s pitch movement and makes the pitch dependent on depth rather than fade. It then follows that same-handed changeups only added .039 points of wOBA at elevation, right in line with breaking balls and altitude-resistant shapes. However, the traditional opposite-handed changeup that relies on fade and run to make the hitter “run out of barrel” got absolutely annihilated when altitude took those last three inches of run away.
This is best visualized below, where a Washington State arm pitching at UCLA induces a whiff on 2026 #1 Overall Pick Roch Cholowsky with his runny, 9 IVB/-24 HB changeup. Beat by the deception AND the run on the pitch, Cholowsky just runs out of barrel.
This same arm pitching up at Nevada later in the year gives up a long home run on a similarly-located changeup. This changeup only ran 13 inches, making it a significantly worse offering against the righty and keeping the pitch on his barrel a tick longer. Losing that run killed this fade-dependent changeup’s effectiveness.
Unfortunately, statistical rigor stops us short of being 100% sure that the opposite-handed changeup is significantly worse than the same-handed changeup when thrown at altitude. Because same-handed changeups are rarer, the arms throwing them often have better feel for the pitch, though standardizing for Stuff+ and Command+ between the two groups showed the same results. Between that doubt and our relatively small sample (only 665 BIP for R/R and L/L changeups), there’s enough noise built in to doubt that the gap really is that large. However, this trend is certainly something that our team will be keeping an eye on.
4. Cover Your Bases
Elevation is just one variable in a much longer list of park-specific and situational factors that can quietly swing outcomes on a given night. Park Factor and altitude effects are examples of the kind of nuance that’s easy to talk about but hard to actually quantify and act on. That’s the gap Paradigm is built to close.
Take our Catch Probability Model inside SHIFT: it doesn’t just tell you whether a ball was caught, it factors in park-specific carry and elevation the same way this study did. This way, a coaching staff evaluating outfield range or positioning can work off numbers adjusted to the parks their guys actually play in instead of basing their analysis on a generic spray chart. The same logic extends across all of Paradigm’s client services: custom reports, advanced scouting, self-service data tools, and more.
Each program’s schedule, roster, and home park create a different set of these variables, which is why our tools are designed to be tailored rather than one-size-fits-all. A Mountain West staff that plays a quarter of its road games at altitude has a different problem to solve than one that never leaves sea level, and the data infrastructure around a program should reflect that. That’s the standard we hold ourselves to at Paradigm: giving programs personalized insights that impact winning.
