A plus-money value spot in Denver-East, a Sale-Yamamoto pitchers' duel with a twist, and a Citi Field under nobody's talking about.
Three Grade A calls today, all sitting at 80 confidence, and none of them are the chalk. Two moneylines on plus-money dogs and an under that the market has priced like it doesn't believe its own starting pitching matchup. That's the through-line: places where the model's number and the listed price aren't talking to each other.
Colorado at Washington, and the model wants the Rockies outright. Win probability sits at 57%, good for a 6.0% edge over that +126 price. That's real disagreement with the market, not a coin-flip nudge.
Gabriel Hughes takes the ball for COL with a 3.74 FIP; Jake Irvin counters for Washington at 4.77. Advantage Colorado on the mound, though the pitching component for this game grades out neutral at 47 — the model isn't leaning on Hughes dominating, it's leaning on the whole picture. That picture includes an offense score of 67 (elevated) and a bullpen score of 95, which is about as loud a number as you'll see in this column. Park and weather come in neutral at 59 and 49. Add it up and the model projects an 11.4-run game against a market total of 10 — COL 6.2, WSH 5.2. High-scoring environment, Colorado favored to be on the right side of it, and you're getting paid plus money to take them straight up. Grade A, and the number backs it.
This is the one with the interesting wrinkle. Yoshinobu Yamamoto starts for LA at a 3.17 FIP. Chris Sale starts for Atlanta at 2.03 — better number, home park, and still the model has the Dodgers winning 56% of the time, a 4.9% edge at +111.
Look at the components and you'll see why this is a lean, not a lock. Pitching grades suppressed at 5. Bullpen suppressed at 21. This is projected as a low-scoring, pitching-controlled game — LAD 3.7, ATL 3, total of 6.7 against a market total of 6.5, barely above the line. Offense is elevated at 60 despite all that, park is dead neutral at 50, weather 41. The edge here is smaller than the Rockies play for a reason: the model likes LA to escape a tight, defensively-inclined game, not to blow it open. Sale is the better pitcher by the raw number. The model just thinks the Dodgers win the game anyway, and it's willing to take plus money to say so. That's a bet, not a certainty — treat it accordingly.
Jacob Misiorowski for Milwaukee carries a 1.97 FIP into Citi Field. Sean Manaea for the Mets sits at 4.21. That gap alone tells you where the model's head is at, but the surrounding numbers back it up hard.
Projected total: 5.6 runs against a market line of 7. That's a 6.0% edge, and the model's probability split makes it explicit — 71% under, 29% over. Split projection has MIL at 3.8, NYM at 1.8, so this isn't a 3-3 grind projection, it's lopsided toward the pitcher who owns the better FIP. The components support the shape: pitching graded suppressed at 19, bullpen suppressed at 25, offense a flat neutral 48. Park score comes in at 34 — on the lower end of what you'll see in this column, which tracks with a game the model expects to stay south of the total. Weather's neutral at 41. Everything here points the same direction, and the price is basically a coin flip at -105 for a lean this lopsided in probability.
Three plays, three different shapes — a plus-money favorite in a track-meet environment, a plus-money favorite in a pitchers' game, and an under built on one start's worth of FIP gap doing a lot of work. Bet them according to the numbers in front of you, not the vibes.