Every October the same sentence shows up: cold weather favors the pitcher, take the NRFI. It is half right, and the half that is right is smaller than the way it gets said. This is what cold air actually does to a batted ball, to the man throwing it, to the man swinging at it, and what the 2026 first inning shows when you sort it by temperature.
Sample for everything below: every 2026 regular-season game through September 21 with a completed first inning and a weather reading, from the MLB Stats API game feed. 2,370 games. 450 were dome or closed roof and are held out of the temperature rows. 1,920 open-air games carried a parsed temperature. League NRFI rate across all 2,370: 1,166 (49.2%).
What cold does to the ball
Cold air is dense air. Alan Nathan, the physicist whose work most ballistics claims in baseball trace back to, puts the temperature effect at about 3 feet of fly-ball distance per 10°F. A 400-foot drive at 90°F is a 388-foot drive at 50°F. That is the difference between a home run and a warning-track out on a meaningful share of well-struck balls.
The effect is large enough to show up at league scale. A 2023 Dartmouth study in the Bulletin of the American Meteorological Society attributed more than 500 home runs since 2010 to warmer-than-normal game temperatures, about 1% of all home runs in the window, by the same air-density mechanism run in the other direction.
Chris Constancio’s Hardball Times study of roughly 2,000 games found the home-run rate on batted balls at over 4% in warm weather and about 3.2% in the cold, with balls in play also less likely to become hits when it is cold. The ball does not carry, and it does not get through the infield as fast.
What cold does to the pitcher
Here the story turns. Cold favors the pitcher on contact quality and hurts him on command.
Pitchers describe a cold ball as hard and slick. Jeff Samardzija told FanGraphs the ball is “a pool cue, it’s hard as a rock” in cold weather. Eno Sarris’s look at pitching in the cold found whiff rates on every off-speed and breaking pitch slightly lower under 60°F than above it (sliders 14.8% against 16.1%, changeups 15.1% against 15.6%), four-seamers and cutters losing about a third of an inch of vertical movement, and usage drifting toward sinkers and two-seamers, the pitches that need the least feel.
Constancio’s numbers agree on the cost: pitchers walk more batters in cold weather and hit more of them, which he attributes to control rather than to hitters failing to get out of the way. They also strike out a higher share of hitters in the cold. Fewer whiffs per pitch, more strikeouts per plate appearance sounds contradictory until you remember the hitter is cold too.
What cold does to the hitter
FanGraphs measured bat speed against game temperature and found about 0.2 mph of bat speed per 10°F, a little over half a mile per hour between the coldest and warmest games. That is small. The bigger cost is feel: a mishit on a cold night stings the hands, hitters know it, and the two-strike swing gets shorter. Add the dense air taking distance off the balls that are squared up, and the hitter’s ceiling drops more than the pitcher’s floor does.
The largest study of the question, Koch and Panorska’s 22,215 games from 2000 through 2011, found runs, batting average, slugging, on-base and home runs all significantly higher in warm games than cold ones, with walks the one statistic moving the other way. Runs rose from cold to warm at 18 of the 23 parks in the study. Net, cold suppresses scoring. It does not suppress baserunners.
The 2026 first inning by temperature
The first inning is where a weather claim should be easiest to see. Both starters are fresh, both lineups send their top three, and nothing about the bullpen has happened yet. Open-air games only.
| Temperature | Games | Scoreless 1st | NRFI rate | 1st innings of 2+ runs | Runs per 1st |
|---|---|---|---|---|---|
| Under 50°F | 62 | 36 | 58.1% | 25.8% | 0.87 |
| 50 to 59°F | 140 | 76 | 54.3% | 27.1% | 0.97 |
| 60 to 69°F | 327 | 160 | 48.9% | 28.4% | 1.06 |
| 70 to 79°F | 703 | 355 | 50.5% | 27.0% | 1.01 |
| 80 to 89°F | 554 | 262 | 47.3% | 29.8% | 1.13 |
| 90°F and up | 134 | 58 | 43.3% | 35.8% | 1.28 |
Collapsed to the line everyone quotes: under 60°F, 112/202 (55.4%) scoreless. At 60°F and up, 835/1,718 (48.6%). A 6.8-point gap, and the runs-per-inning column moves the same way, 0.94 against 1.08.
The multi-run column is the physics showing up. A first inning of two or more runs usually needs the ball to leave the yard or reach the wall. In the coldest band that happened in 25.8% of first innings. In the hottest, 35.8%. The scoreless rate moves 15 points across the table. The big-inning rate moves 10.
Cold is not one thing
202 cold games is enough to see a direction and not enough to price it, because cold arrives with company.
Calendar. 118 of the 202 cold games were in April, 59 in May, 18 in the March opening series. Seven were in June through September combined. April cold went 67/118 (56.8%). May cold went 30/59 (50.8%). The April figure carries Opening Day lineups, starters still building pitch counts, and whichever northern parks drew the early home dates. The 60°F to 79°F control for April went 83/166 (50.0%), so cold April did run above comfortable April. It is one month of evidence.
Wind. Inside the cold band, wind in from the outfield went 32/54 (59.3%) scoreless and wind out went 43/87 (49.4%). Cold with the wind blowing out scored at the league rate. Cold with the wind in, or across (32/51, 62.7%), is where the scoreless innings were. Temperature alone is the weaker half of the forecast.
Day and night. Cold day games went 60/104 (57.7%), cold night games 52/98 (53.1%). Both above the league rate, four points apart, fifty games each. Not a split to act on.
Reading the NRFI and YRFI price when it is cold
An NRFI at -130 needs 56.5% to break even. At -120, 54.5%. At -110, 52.4%. The 2026 cold band hit 55.4%, and the under-50°F band hit 58.1% on 62 games. Cold alone, with no read on the two starters, clears -110 and does not clear -130. That is the whole finding, and it is smaller than the sentence it usually travels in.
The YRFI side is the mirror. YRFI hit 44.6% of cold first innings and 51.4% of the rest. A YRFI at +110 needs 47.6%. A cold-weather YRFI is paying plus money for a side that, in this sample, lost more often than the price implies. When the book moves the YRFI to +125 or +130 on a cold night, that movement is the market pricing what this table shows. Where the number lands relative to the starters is the question the board exists to answer, not the weather line by itself.
None of this is a pick. It is the size of the effect, next to its sample, so a price can be read honestly.
What The Full Count does with it
The board applies a weather multiplier to expected first-inning runs: 0.004 per degree from 70°F, clamped between 0.85 and 1.15, with a wind term on top. At 50°F that is 0.92. At 40°F, 0.88.
Across the six bands above, The Full Count’s average stated probability moved 10 points from coldest to hottest (0.387 to 0.287). The observed scoreless rate moved 15 (58.1% to 43.3%). The model leans the right way and may be leaning too gently. We are not retuning a multiplier on 62 games under 50°F; that is the kind of overfit we wrote a methodology page to avoid. It goes on the watch list with a number attached, and the park table and the full weather split sit beside it.
The September that has not been cold
The reason to publish this now is the calendar, and the calendar has not cooperated. Through September 21, 2026 open-air games were 43.2% scoreless in September (92/213). 188 of those 213 games were played at 70°F or warmer. Exactly one September first inning has been played under 60°F.
So the cold-weather first inning of 2026 is still an April fact. The postseason is where it gets tested again, in different parks, with different starters, on a sample of a few dozen games. When it does, the result will land on the board graded beside the model’s number, the same as every other night.
Statistics and research, not betting advice. 21+. Gambling problem? Call 1-800-GAMBLER.