Our HR model uses BANG scores to make home run picks for today’s slate of MLB games, combining Statcast data with weather, park factors, previous batter vs. pitcher history and matchup quality.
🔥 / 🔥🔥 / 🔥🔥🔥 = heating up · 🥶 / 🥶🥶 / 🥶🥶🥶 = cooling off · 🚀 = last-game max EV · 🆚 = platoon advantage
This page surfaces today’s MLB home run picks: the home run props where the HeatRadar model thinks the price is mispriced versus what the books are charging. Every starting lineup gets scored, and only the players whose model-projected HR probability beats the implied probability by a meaningful margin show up as today’s MLB home run bets. The rest sit below the cut.
The model stacks four inputs for every batter. First, Statcast contact quality: a per-plate-appearance score called BANG, built from exit velocity, launch angle, hard-hit rate, and barrel rate. Hitters who already grade out as elite contact get a head start before any matchup math runs. Second, the pitcher he’s facing: BANG-allowed by handedness, actual HR-against rate, plus the bullpen behind him. Third, weather: wind direction, temperature, and humidity at game time. Coors at 88° with the wind blowing straight to right is a different game than a 55° marine layer in Oakland. Fourth, park factor, pulled from a rolling three-year HR sample and applied per-side (Yankee Stadium plays as a lefty park; Petco kills righties).
We only post picks above a +175 edge versus the best HR prop price across the major US sportsbooks. Below that line, the model’s edge isn’t reliably big enough to overcome juice, and we’d rather show nothing than show noise.
Every pick gets graded the next morning. The running ROI tracker lives in the page footer and updates daily. Under the strict edge filter (above +175), the model is currently running profitable across roughly 120 graded picks. The looser variants are shown alongside for transparency but aren’t recommended for actual bets.
Do you cover NO home run picks too? Yes. Alongside the YES picks, the model flags “fades”: props where the YES side is overpriced enough that the NO side is the actual +EV bet on a no-vig book like ProphetX or Novig. Fades show up tagged in the same table, color-coded by tier. They’ve been live since the start of the season and grade independently in the fade tracker.
Is this gambling advice? No. This is a model showing where we think the books are wrong. Bet responsibly, never more than you can lose, and check the responsible gambling resources if you need them.