Away - LA Angels
56%
Lean
"Angels +113 ML: Bayesian posterior says 50.7% true win prob vs market 42.4% — +8.3pp edge on the dog in a market overrating home advantage"
Line Movement
Current
H -1.5 / 10.0 / ML -136/113
Movement
No significant movement detected — market likely settled near open with low liquidity
Both teams hit hard by injuries: ATH -6.9% (Rooker, Severino out), LAA -7.8% (Soler, G-Rod out) — net roughly equal, but missing star bats depresses scoring potential significantly
OMEGA total projection (22.5) vs market (10.0) indicates severe market underreaction to whatever scoring model sees — but the model itself may be overfit given the data quality warning
Monte Carlo projects 10.4-9.9 (20.3 implied total) — still far above 10.0, but high standard deviation (4.7) in scoring suggests extreme uncertainty
Whale signals extremely lopsided toward home (82% volume, $589K) despite Bayesian edge favoring away — whales are betting the public narrative, not the quantitative signal here
Home moneyline carries negative expected value at current odds. calibrated shows the market prior (57.6%) is significantly higher than the fused posterior (49.3%) — the -136 price is nearly 9 percentage points too rich. No value backing the home side.
Bayesian fusion exposes a market that prices the A's as heavy -136 favorites, while the fused posterior calls it a coin-flip with slight lean to the Angels. The 10.0 total is the most glaring number — both OMEGA (22.5) and Monte Carlo (20.3) project nearly double that, though the model is clearly overfit (no pitcher data inflates expected scoring). Whale signals hammer home at 82% volume ($589K), but with zero sharp-book data and LOW model agreement, that's public confirmation bias, not sharp money. The only positive-EV play is the Angels moneyline at +113 — +8.3pp edge on a 50.7% probability — but data quality caps conviction at LEAN. Pass on the over: the 15.2pp Bayesian edge is seductive, but no pitcher data + marine layer risk + historically bad totals performance make it a data mirage. The spread (+1.5) is the safest lean: +3.9pp edge at -110 with an 8-run cushion in a sport where single games are high-variance.
Reduced sizing. Edge exists but sample or signal strength is moderate.