Best MLB plays today — Sunday, July 19, 2026
Every board's highest-conviction edges in one place — where our model diverges far enough from the book to have been profitable in the season backtest. We found 3 qualifying edges, but plays post only once lineups are final — until the official lineup is in, the projection can still move, so we don't surface a play that could flip. They'll appear below as each lineup locks. Each play is graded in the open with our running hit rate and ROI. Research signals, not betting advice — 21+.
◆ Team totals
3 team-total leans are waiting on lineups. Our edge is the gap between our projection and the book's line — but a projection rests on who's actually playing: the batting order, and the matchup the starter faces. Until the official lineup posts (~1–2 hours before first pitch) that projection can still move, so a play could flip. To keep our calls honest we hold team-total plays until lineups are final, then post the ones that still clear our edge threshold. This season team-total leans are hitting 62% — a +12% return · +9.9u net over 85 graded. We're posting them as they prove out, not as locks.
★ Strikeouts — Plays of the Day
No identified Plays of the Day on this slate — our model didn't find a strikeout disagreement with the book big enough to clear our edge threshold. The model still ran every matchup; it just didn't flag a play worth posting today. This season strikeout PODs are hitting 60% — a +9% return · +7.7u net.
Why we only surface these markets
We post a play only where a full-season backtest shows a durable, repeatable edge — today that's strikeout props (big projection-vs-line gaps) and team-total leans (our strongest run-total disagreements). Everything else we publish as graded transparency, not plays: our number set beside the market and scored in the open, but with no lean attached.
Markets like the moneyline, game totals, home runs, hits, RBI, runs, and total bases are priced sharply and managed tightly by the books. Across a full season of backtesting, even a well-tuned model can't find a repeatable edge in them — beating these markets consistently is strikingly hard, so we won't manufacture a play we can't stand behind. If one of them ever earns a sustained edge, it'll surface here automatically.
Δ shows how far our projection diverges from the book's line — the bigger the disagreement, the stronger the historical edge. Hit rates and ROI are season-to-date, graded against the posted line. Not betting advice.