Today is a single-league card. MLB carries all five games; MLS and WNBA are both on scheduled breaks, so there is nothing to model there. That narrows the focus and, honestly, sharpens it — fewer games means each price gets closer attention. Our picks post about 30 minutes before each first pitch and settle afterward on a public, tamper-evident ledger, win or lose.
MLB
The headline is the Padres–Braves finale at Truist Park, 12:15 p.m. ET. Atlanta (59-42) hands the ball to Chris Sale (10-6, 2.06 ERA, 123 K) for his 19th start. Sale last threw seven scoreless innings against Texas on Friday, giving up two hits, and carries 16 starts of five-plus innings — he is chasing a third straight scoreless outing. San Diego (50-52) counters with Griffin Canning (1-8, 6.67 ERA), who allowed four earned runs in 3 2/3 innings Saturday against Kansas City. On paper this is a wide starter gap, and the market will price it that way. The model's job here is not to confirm the obvious but to test whether the number has overshot: a large favorite's price often bakes in more than the true edge, and Canning's line, while poor, is a small sample against a bullpen-dependent game. Atlanta took the prior meeting 7-6, a close margin that argues against assuming a blowout. San Diego is sub-.500 and just dropped a series to the Royals, so the storyline pressure is real, but pressure is not a probability.
The undercard is where value tends to hide. Twins at Guardians (1:10 p.m. ET) pits Taj Bradley against Gavin Williams — two arms whose outcomes swing heavily on command that night, which widens the honest uncertainty band. Rays at Blue Jays (3:07 p.m. ET) features Ian Seymour opposite Shane Bieber; the model weighs Bieber's track record against Seymour's shorter sample and the Toronto lineup's platoon shape. Diamondbacks at Cardinals (5:15 p.m. ET) is Brandon Pfaadt versus Michael McGreevy, a matchup where park factors and expected innings from each starter matter as much as raw ERA. Royals at Tigers (6:40 p.m. ET) closes the slate with Randy Dobnak against Troy Melton, a lower-profile pairing where thin public attention can leave a number slightly stale.
What a sharp model actually weighs across these games: projected starter innings and strikeout rate, the quality of each bullpen behind the starter, lineup handedness against the opposing arm, park and weather, and — critically — the gap between our model probability and the posted price. An edge only exists where our number and the market's number disagree by enough to survive the vig. On the Sale game, if the model probability and the market price converge, we pass rather than chase a name. On the quieter Central and NL matchups, small mispricings are more plausible, but so is variance; two starters with volatile control can produce a wide range of outcomes, and we state that plainly rather than dress it up.
We will publish the exact model probability, the market price, and the resulting edge for any game we release, about 30 minutes before first pitch, so the reasoning is visible before the result is known — not reconstructed after. If a game offers no edge, we will say so and stand down. Every posted pick settles on the ledger regardless of outcome.
MLS
No fixtures today. July 23 sits ahead of the 2026 MLS All-Star Game, scheduled for July 29 at Bank of America Stadium in Charlotte. With no league matches on the board, there is nothing to model and no pick to make. We would rather post nothing than manufacture a play against an empty schedule.
WNBA
The WNBA is in its All-Star Break, July 23–27, with the All-Star Game on July 25. No regular-season games are scheduled today, so there is no card to evaluate here either. When league play resumes, our process picks back up — probability, price, edge, and a ledger entry for each release.
---
A closing note on the day: a five-game MLB card with two leagues dark is a smaller sample than a full slate, which means fewer genuine edges, not more. We will only release where the number justifies it. Losses are part of any prediction record, and no single game is a certainty; size accordingly.
Predictions carry risk and results are never certain. For 21+ audiences; if the fun stops, call 1-800-GAMBLER.