"Data-poor ARTS pick'em leans John Yannis ML -115 (+3.4 EV) on 54% model prob vs 53.5% breakeven"
Line Movement
Current
omega_model: Spread John Yannis -0.5 (-110) / Jamie Siraj +0.5 (-110); ML John Yannis -115 / Jamie Siraj -105; Total 2.5 (-110 both sides)
Movement
N/A
Marginal +EV on home side at model -110/-115 pricing vs our 54% probability.
No market odds, injuries, or models available; Bayesian fusion defaults to 50/50 but omega_model applies 54% home win prob for slight +EV at generated lines. Data quality 16% forces Lean tier only with -8 pt reduction. Moneylines preferred over losing totals/props calibration. Props are unvalidated MMA-style projections capped at 60.
Reduced sizing. Edge exists but sample or signal strength is moderate.