MLB ballpark data
MLB retractable roofs: the rules, and what open vs closed is actually worth
By OverUnderWeather Research Desk
Seven MLB parks can open or close their roof on game day. Here is who makes that call, when a roof can close mid-game, and what every game since 2021 says about whether the roof actually changes scoring.
Quick answer
Runs, open vs closed
+0.13
per game · t = 0.87
Home runs, open vs closed
+0.07
per game · t = 1.26
Games measured
3,281
7 parks · 2021–2026
Once each game is compared with its own park and season, an open roof is worth about a tenth of a run and seven-hundredths of a home run — statistically indistinguishable from zero. Against the closing total the gap is -0.10 runs (t = -0.66): the market already prices the roof state, and there is nothing left to chase.
The rules: who opens and closes the roof
- Before the game, the home club decides. Each team sets its own policy based on the forecast, temperature and fan comfort. There is no single league-wide temperature rule.
- After lineups are exchanged, the umpire crew chief takes over. From that point the club cannot ask for the roof to be opened or closed for competitive reasons.
- A roof can close mid-game if weather moves in — normally between innings, or immediately if rain would otherwise stop play. Re-opening mid-game is rare and needs the crew chief's approval.
- Heat closes more roofs than rain. In Phoenix and Arlington the roof is primarily air conditioning; in Seattle and Toronto it is mostly an umbrella.
Park by park, 2021 to 2026
| Park | Open / closed games | Runs open / closed | HR open / closed | Temp open / closed |
|---|---|---|---|---|
| American Family Field | 258 / 218 | 8.60 / 8.51 | 2.38 / 2.25 | 75.8 / 58.6 |
| Chase Field | 118 / 356 | 9.74 / 9.44 | 2.14 / 2.05 | 88.0 / 99.1 |
| Globe Life Field | 59 / 415 | 9.36 / 8.53 | 2.51 / 2.39 | 76.8 / 86.9 |
| Marlins Park | 42 / 432 | 8.69 / 8.58 | 1.79 / 1.99 | 78.8 / 83.7 |
| Minute Maid ParkToo few games to read | 11 / 466 | 9.27 / 8.71 | 2.73 / 2.46 | 76.6 / 84.2 |
| Rogers Centre | 269 / 163 | 8.86 / 8.66 | 2.48 / 2.25 | 72.9 / 58.1 |
| T-Mobile Park | 402 / 72 | 7.83 / 7.75 | 2.35 / 2.18 | 67.3 / 53.9 |
Per-park splits are descriptive only. Parks with fewer than 40 games in one roof state are flagged — a split built on a handful of games is noise, not a signal.
Why the raw numbers mislead
Roofs close for heat, cold and rain, so closed-roof games are not a random sample of weather. Comparing raw averages mixes in which park hosted the game and which season it was. Every headline figure here compares each game with its own park-season average first — which is why the honest answer is "no measurable difference" rather than a scary-looking raw gap.
How we measured this
Every regular-season game at the seven retractable-roof parks from 2021 to 2026, with the roof state taken from the official MLB game feed. Runs and home runs are measured against each park's own season average and compared with a Welch t-test; we only call a gap real at |t| ≥ 1.96 with at least 40 games per group. Closing totals are the last consensus line from ESPN. The full academic write-up is in our open vs closed research study.
Related
Where this is applied
These findings feed the game-time read on every MLB venue page, using the same sample-size and significance rules published here.
OverUnderWeather Research Desk
Independent data analysis on public sports and weather records. Studies match historical game outcomes to archived game-window weather, apply pre-set sample and significance rules, and publish null results when that is what the data show. Editorial policy and corrections log.
Methodology reviewed by Kerry G. KauffmanB.S. in Statistics, Lehigh University
“After examining the complete methodology and structure of the study, I am in agreement that the Welch’s t-test is the appropriate method when comparing two means.”— Kerry G. Kauffman, 2026-09-26
Cite this research
Free to republish with a visible credit and a link to this page. Ready-made citations below — click to copy.
Frequently asked
Who decides if an MLB roof is open or closed?
The home club decides before the game, based on weather forecasts and fan comfort. Once the lineup cards are exchanged, the decision passes to the umpire crew chief for the rest of the game.
Can an MLB roof close in the middle of a game?
Yes. If conditions deteriorate — usually rain starting or a storm arriving — the crew chief can order the roof closed. It happens between innings where possible, or immediately if play would otherwise have to stop.
Does an open roof mean more runs and home runs?
Not measurably. Across every game at the seven retractable-roof parks since 2021, open-roof games scored about 0.13 more runs and 0.07 more home runs than closed-roof games at the same park in the same season — both within statistical noise.
Which MLB stadiums have retractable roofs?
Seven: American Family Field (Brewers), Chase Field (Diamondbacks), Daikin Park (Astros, formerly Minute Maid Park), Globe Life Field (Rangers), loanDepot park (Marlins), Rogers Centre (Blue Jays) and T-Mobile Park (Mariners). Tropicana Field has a fixed dome and is not retractable.
Why do teams close the roof?
Heat is the most common reason — Chase Field and Globe Life Field close for extreme summer temperatures — followed by rain and cold. Each club has its own comfort policy, which is why the same forecast can produce different roof decisions at different parks.