The first-inning scoring percentage in MLB is not a mystery number. It is a
count. Through September 3, 2026 the board’s nightly replay
(/api/record, pending 0) has graded
2,134 regular-season games across 160 slates. 1,059 of those games were
NRFI: 49.6%. The other 1,075 (50.4%) scored in the first inning.
That is the league average NRFI rate this site publishes. It is also the base rate every model on the board has to beat. A hit rate that cannot clear 49.6% on this window is a coin.
The competing claims
Guides and forum posts circulate three different NRFI base rates, usually with no window and no denominator:
- About 49%. This is the one that matches 2025 and 2026.
- 55 to 60% scoreless. That is a different season, or a selected month, treated as a law.
- 65 to 68% scoreless. We cannot find a recent full regular season that lands there.
The 2026 number on this page is 1,059/2,134 (49.6%), opening day through September 3. If someone quotes a different league rate, ask for the season and the count.
Year by year
The replay on the board covers 2026. For prior seasons we counted the same
way from the public MLB Stats API: regular-season games (gameType=R) with
a completed first-inning linescore. Spring training and the All-Star
exhibition are out.
| Season | Window | Scoreless 1st (NRFI) | First-inning scoring (YRFI) |
|---|---|---|---|
| 2024 | Full regular season | 1,295/2,430 (53.3%) | 1,135/2,430 (46.7%) |
| 2025 | Full regular season | 1,211/2,434 (49.8%) | 1,223/2,434 (50.2%) |
| 2026 | Opening day through September 3 | 1,059/2,134 (49.6%) | 1,075/2,134 (50.4%) |
2024 was an NRFI year. 2025 and 2026 are not. Is NRFI or YRFI more common? Over the last two seasons they are essentially even, with YRFI a hair ahead in the 2026 window. A page that says “NRFI hits about 60%” is describing a year we did not count, or a sample it did not show.
The 2026 replay count (2,134) and the Stats API linescore count for the same dates (2,109 games, 1,047 NRFI, 49.6%) differ by 25 games. The replay grades from the board’s own linescore cache; the year table’s 2024 and 2025 rows, and the park and weather posts, use schedule hydrate. Same source family, two tallies. We publish both rather than forcing them to match.
2026 by month
Monthly NRFI rate from the same Stats API linescores, 2,109 games with a completed first inning:
| Month | Scoreless 1st | Rate |
|---|---|---|
| March | 45/76 | 59.2% |
| April | 195/392 | 49.7% |
| May | 198/419 | 47.3% |
| June | 186/395 | 47.1% |
| July | 190/371 | 51.2% |
| August | 216/417 | 51.8% |
| September (through the 3rd) | 17/39 | 43.6% |
March is 76 games of cold openers. September so far is 39 games. Neither one is a season. May and June, the two largest mid-summer months, sat in the 47s: YRFI was the more common result.
What 49.6% is not
It is not a model hit rate. The Full Count’s daily top pick over this same replay is 98/160 (61.3%). That is a selected game per slate, not the league. The method and the three published measures (top pick, top 3, every game) are on the methodology page and in the 2,134-game backtest.
It is not a park rate. Coors Field and Camden Yards do not share a first inning. The venue table is in NRFI park factors.
It is not a weather law. Temperature and wind splits, including the bands that are too small to lean on, are in Does weather change the first inning.
Check it
The live tally updates overnight at
/api/record. Pick any past date on
the board and the first-inning result sits next to the lean.
Statistics and research, not betting advice. 21+. Gambling problem? Call 1-800-GAMBLER.