Two plus-money run lines and a Nationals moneyline lean carry the board today, all graded A, all sitting at 80 confidence — here's the case for each.
Three A grades today, all parked at 80 confidence, which means the differentiator isn't how sure the model is — it's what the numbers actually say about each matchup. Two of these are run-line plays where a favorite needs to win by two, and the third is a straight moneyline where the model thinks the market has the win probability wrong. None of it is a total. Read the labels before you bet the number, because these are three different animals wearing the same grade.
Jared Jones takes the mound for Pittsburgh against Sean Manaea and the Mets, and the model likes the Pirates to win this one by more than a run. Jones carries a 3.83 FIP, Manaea sits at 3.96 — a real but modest gap, and it shows up in the projection: PIT 5.1 runs, NYM 3, a projected margin of 2.1. That's enough to clear -1.5.
The price is the story here. Plus-144 on a run line means the market isn't fully buying the Pirates covering, and the model puts PIT's cover probability at 53% — call it a coin flip that tilts Pittsburgh's way — for a stated edge of 6.0% over that price. That's not overwhelming, but at plus money you don't need overwhelming.
Worth saying plainly: the pitching component here grades out suppressed at 38, and offense, bullpen, and park/weather all sit basically neutral (49, 41, 48). Park score is 34, weather 40 — nothing loud from PNC Park today. This is a lean built on the run projection and the price, not on some environment stacked in Pittsburgh's favor.
Same shape, different city. Ryan Gusto starts for Miami against Grayson Rodriguez and the Angels, and the FIP gap is the widest of the day's three plays — Gusto at 3.68, Rodriguez at 5.04. The model has Miami covering -1.5 at 54%, with the same stated 6.0% edge over the +140 price, and projects MIA 5.7 to LAA 3.6, a 2.2-run margin.
The environment grades don't flatter this one either — offense is suppressed at 32, and pitching, bullpen, and park/weather all sit in the neutral band (47, 42, 43). Park score is 31, weather 35, both on the low end. So this isn't a park-inflated shootout pick. It's a bet that Gusto's edge over Rodriguez on the mound is real and large enough to produce a multi-run final margin, priced at plus money because the market is still just calling Miami a moneyline favorite rather than a two-run one.
Grade A, confidence 80, same as the Pirates play — but the FIP spread between these two starters is the widest gap on the slate, and that's what's carrying this number.
This is the one where the model and the market are further apart. Brad Lord starts for Washington against Brady Singer and the Reds — Lord's FIP is 3.95, Singer's is 4.83 — and the model's outright win probability for the Nationals is 71%. At a -104 price, that's a significant gap, though the model regresses its conviction down to a stated 6.0% edge after weighing the market. Treat that regression as the model admitting a 71% number is aggressive and it's not betting the full gap.
The run projection backs the lean: CIN 2.7, WSH 4.8, for a projected total of 7.5 against a market total of 9.5 — two full runs lower than what the board expects. Part of that comes from the bullpen component, which grades out elevated at 100, the loudest number on this entire slate. Park/weather also grades elevated at 60, with a park score of 44 and weather at 52. Offense (56) and pitching (50) are neutral, rest is neutral at 50.
A moneyline price of -104 isn't asking for much conviction to make sense, and a 71% model win probability clearing that number by a wide margin is exactly the kind of gap that's supposed to get action. Just know this is graded as a 6.0% edge after the model talked itself down from its own number — that's the honest version, not the flashy one.
Two plus-money run lines built on FIP gaps and multi-run projections, and a moneyline lean where the model's own outright number got regressed hard before it was offered up. All three are A grades at 80 confidence. None of them are locks — a 53% or 54% cover probability is still a coin that favors one side, not a guarantee, and even Washington's 71% got walked back before it hit the page. Bet the numbers as given, not the confidence in how the numbers are presented.