Building sample: only 0 graded inning picks so far.
Every prediction and result is shown from day one — treat early records as noise
until the sample grows.
🎯 Top 5 No-Run Innings on the Slate
Every Game, All Nine Innings
Loading…
How NRI No-Run Innings Works
What it predicts
For every MLB game, a calibrated probability that no run scores in each
inning 1 through 9 (both halves, as played). When the model says 70%, it aims
to be right about 70% of the time. The starred cell is each game's best (highest)
no-run inning; the ribbon marks slate-wide top picks.
Top 5 picks
The 5 highest-probability (game, inning) combinations across
the whole slate are the featured picks, ranked purely by model probability —
market edge is context and never changes the ranking.
Market lines & the 1st inning
Sportsbooks only post per-inning run markets for the 1st inning. Inning-1
cells show the de-vigged market probability and edge where a line exists, plus the
sibling NRFI model's probability as a cross-check (two different models,
both labeled). Innings 2–9 always show "no line".
Locking
All nine inning predictions for a game lock together
10 minutes before first pitch. At lock the probabilities,
slate ranks, top-5 membership, and best-inning flag freeze for
grading; before lock they drift as lineups and starters firm up.
Grading (book settlement rules)
An inning wins if it finishes 0–0 as played: a skipped bottom of the 9th
settles on the top half alone, and a walk-off run counts as scored. Innings that
never start (postponement, rain-shortened game) are void. Units assume a flat
$25 stake — inning-1 picks at the locked market
price, innings 2–9 at a −110 notional (bookkeeping, not a bettable claim).
Explorer
The Inning Explorer shows every team's
zero-run rates by inning — offense and pitching lenses, season/last-30, home/away,
and starter-vs-bullpen splits.