Our model's win probability vs. the market's implied probability. The gap is the edge.
Every factor that moved the model. Every number sourced — no hallucinations.
Supreme Brain assigns Pittsburgh a 67.0% win probability against Miami at -158 odds, implying a market probability of just 50.0%. That 17-point gap translates to +5.0% expected value on the current number—enough to warrant a quarter-Kelly stake of 0.15 units. The thesis is straightforward: generational ace Paul Skenes, who has logged 16 strikeouts across his first two starts (1 inning pitched total), faces a Miami lineup that managed just 2 runs yesterday and enters with eight players on the injury report. Pittsburgh carries its own injury concerns—six players sidelined—but the pitching mismatch tilts this slate-opener toward the home side. Miami sits at +133, a number that reflects real upset risk if Skenes hangs a mistake pitch. Still, the model favors Pittsburgh by a meaningful margin, and the edge justifies a measured position.
Miami's offense managed just 2 runs yesterday, per Supreme Brain, and now draws a start against Paul Skenes—a generational ace who has punched out 16 batters across his first two starts (1 inning pitched total).
Supreme Brain assigns Pittsburgh a 67.0% win probability versus a 50.0% market-implied probability at -158, yielding +5.0% expected value and a quarter-Kelly stake of 0.15 units—a high-conviction play built on pitching mismatch and offensive drought.
If Skenes hangs one mistake pitch early and Miami's depleted lineup capitalizes, the +133 price on the underdog starts to look prescient. Eight injured players can mean opportunity for fresh bats without scouting history, and a single crooked number in the first three innings flips leverage. The model assumes Skenes pitches to his stuff; if command wavers or the umpire squeezes the zone, Miami's 2-run output yesterday becomes noise rather than signal.
Skenes versus a lineup that scraped together 2 runs a day ago is the kind of mismatch that justifies laying -158. The edge is there, and the model sees it clearly.