Coors Field Is a Different Sport
5,280 feet of altitude. 27.2% more runs. And somehow, adjusting for it makes predictions worse.
The Physics
Coors Field sits one mile above sea level in Denver, Colorado. At that altitude, the air is about 15% thinner than at sea level. This has three compounding effects on baseball:
🚀 Balls fly farther
Less air resistance means fly balls carry 5-9% farther. Warning track flies become home runs. Deep flies become doubles off the wall.
🌀 Breaking balls don't break
Curveballs and sliders rely on air resistance to generate movement. Thinner air means less movement, which means pitchers are less effective.
📐 Outfield is massive
To compensate, Coors has the largest outfield in MLB. This means more ground to cover and more balls falling for hits.
The result: Coors Field produces roughly 27% more runs than an average MLB park.
The Numbers
We calculated park factors from 7,292 games across 2023-2025. Here's where Coors lands relative to every other park:
🔥 Top 5 (Hitter-Friendly)
🧊 Bottom 5 (Pitcher-Friendly)
Look at the gap. Coors at 1.272 vs. the second-highest (Chase Field) at 1.097. That's not a gentle slope — it's a cliff. Coors is 2.5× more extreme than the next-most-hitter-friendly park.
Coors vs. Average
+27.2%
T-Mobile vs. Average
-13.4%
Coors vs. T-Mobile
+40.6%
A game at Coors produces 40% more runs than a game at T-Mobile Park. Same sport. Same rules. Completely different game.
The Prediction Paradox
Here's where it gets weird. You'd think that knowing Coors is extreme would help you predict games there. Just adjust up, right?
We tried. Three different park factor adjustment methods. And for Coors specifically:
Coors accuracy WITHOUT park factors: baseline
Coors accuracy WITH park factors: -6.2% worse
Adding park factor adjustments made Coors predictions less accurate. How?
Variance scales with scoring. More runs means more variance. A 10-8 game has far more ways to unfold than a 2-1 game. Park factor adjustments increase your scoring estimates but can't model the increased chaos.
The "Coors hangover" effect. Rockies hitters' road stats are deflated because they adjust to thin-air conditions, then struggle at sea level. This messes with player-level models that use career averages.
Elo already knows. Our Elo ratings already capture home-field advantage implicitly from historical results. Layering explicit park factors on top double-counts part of the effect.
The Per-Park Accuracy Map
When we tested park factor adjustments across all 30 parks, the variation was dramatic. Some parks loved the adjustment. Others were destroyed by it.
📈 Biggest Improvements
📉 Biggest Declines
Notice something? The most extreme parks — the ones where you'd most expect park factors to help — are the ones where they hurt. Meanwhile, mid-range parks like Guaranteed Rate Field in Chicago saw the biggest gains. The lesson: static adjustments work best where the signal is moderate, not where it's overwhelming.
Coors by the Fun Numbers
The same two teams playing the same game would score 3.4 more total runs in Denver than in Seattle. That's not a rounding error — that's an entirely different game.
Imagine if NBA courts varied so much that some gyms produced 130-point games and others topped out at 90. That's basically what's happening across MLB parks, and Coors is the extreme end of the spectrum.
So What Do We Do?
Currently, our production model does not apply static park factor adjustments to win predictions. The backtests showed inconsistent results — gains one season, losses the next — and the Elo system already captures home environment effects through historical team performance.
We do use park factors for score estimates and total run projections, where they're genuinely useful. If you're looking at a Coors game, you should expect more runs. That's physics, and the data confirms it.
The real unlock for Coors (and every park) is probably game-day weather data — temperature, humidity, wind speed and direction. A 95°F day at Coors is a different beast than a 55°F night. Static park factors can't capture that, but real-time environmental data could.
Until then, when you see a game at Coors Field on our predictions page, just know: the model is doing its best with the hardest park in baseball. Some things are just genuinely hard to predict, and a mile-high stadium where baseballs fly like they're on the moon is one of them.