The Cardinals arrive at Dodger Stadium for Wednesday's middle game carrying a rare bright spot: a win to snap a run of frustration. St. Louis sits at 68-70, 3-7 over its last 10, five games back of Arizona for the final NL wild-card spot. Los Angeles, at 82-56 with a 40-27 home record, is looking to answer a lopsided Tuesday opener in which it surrendered 13 runs and five home runs while its own offense pushed across eight.
The matchup on paper favors the home side. Our model lists Tarik Skubal for Los Angeles against Quinn Mathews for St. Louis, though note the researched context points to the Cardinals leaning on a bullpen effort in this middle game rather than a single length starter. Either way, the projection frame is the same: a Los Angeles offense rated 52.1 against a St. Louis unit rated 27.7, in a park playing slightly below neutral at a 0.97 factor.
The number
Our pick is Los Angeles on the moneyline. The model puts the Dodgers' win probability at 80%, against a market-implied 74%. That is an edge of 9.4. Model confidence registers at 80.6.
The number reflects a projected low-scoring game: our model total is 7.0 against a market total of 8.5, an Under lean. The model projects Los Angeles to win by a margin of 2.38. None of that is a promise of outcome; it is a read on where the price and the probability diverge.
Why the model leans this way
Three inputs carry most of the weight here.
First, the offensive gap. A 52.1 rating for Los Angeles against a 27.7 for St. Louis is a wide separation, and it is the single largest driver of the projected margin. In a game where St. Louis is expected to piece innings together from its bullpen, that gap tends to compound across at-bats.
Second, the run environment. The 0.97 park factor nudges scoring down slightly, and our model total of 7.0 sits a full 1.5 runs beneath the market's 8.5. A lower-scoring projection generally favors the stronger, better-rested side on the moneyline, because fewer total runs reduce the variance that lets an underdog steal a game.
Third, the pitching and form. Skubal against a bullpen approach frames a starter-versus-committee edge for Los Angeles. On form, the Dodgers have lost one straight and St. Louis has won one straight, so this is not a case of riding a hot streak into a cold one; the projection rests on the underlying ratings rather than momentum. Tuesday's shootout does not change the model's read on run environment for Wednesday.
Taken together, those inputs produce the 80% figure and the 2.38 projected margin. The edge exists because the market's 74% is close to, but a step short of, where our model lands.
The case against
This pick is not free of real risk, and roughly one game in five is expected to go the other way even if the model is calibrated correctly.
Start with the opener. Los Angeles just gave up 13 runs and five home runs. Pitching staffs that were shelled the night before can carry that into the following game, and St. Louis, despite a modest offensive rating, showed on Tuesday that this ballpark can produce runs in bunches. A bullpen effort by the Cardinals is not automatically a disadvantage; a fresh committee with favorable matchups can suppress a strong lineup for a night, and it removes the risk of a single starter unraveling.
There is also lineup context. Will Smith is out of the Dodgers' lineup Wednesday as a planned day off while he eases back from a June absence. That subtracts from the offense our model rates so highly. On the other side, St. Louis is short JJ Wetherholt, on the injured list with wrist tendinitis, so both clubs are operating without a piece.
Finally, ordinary variance. A one-run projected edge in run environment and a 2.38 margin are meaningful but not large. Baseball's single-game noise is high; a well-timed home run, a bullpen misstep, or an early Cardinals lead can flip a game the model still calls correctly over a long sample. A 74% market price is not a mispricing anyone should treat as a formality — it already reflects a heavy favorite.
Our full record sits on a public, tamper-evident ledger, and this pick is one data point within it, not a standalone claim.
Predictions carry risk; only stake what you can afford to lose, and if it stops being fun, step away. 21+.