Ranked picks with source labels.

Model at 60% vs market 64% on TBR. McClanahan has no recorded IP/ERA stats despite 10 starts—data quality concern. Kochanowicz 11.25 ERA in 4 IP is...

Away pitcher Janson Junk has null ERA/IP/K9/WHIP despite 10 starts listed—data quality issue makes total unreliable. Model leans NYM 58% vs market...

Model predicts 9.08 runs vs market 7.0 — a 2-run gap flags structural miss. CLE offense anemic (2.3 RPG, .600 OPS) despite 34-26 record. Suarez sta...

Model predicts 9.12 runs vs market 7.5 — a 1.6 gap in a 0.855 park factor venue. T-Mobile suppresses scoring; model is overweight on recent Seattle...

Away pitcher Andrew Painter has null stats across all core metrics (IP, ERA, K/9, WHIP) despite data_quality='full'. This is a data integrity issue...

Model has TEX 57.5% vs market 51.3% — modest comparison on a sub-.500 home team. KCR offense is brutal (.518 OPS, 2.8 RPG) but Wacha's 0.0 ERA in 6...

Model has COL at 57.5% vs market 47.7% — a 10pp gap at Coors with Gordon's missing peripherals is suspect. SFG road team with better offense (4.6 R...

Model favors OAK 57.5% vs market 40%, a huge gap. NYY 35-23 with 4.8 RPG offense vs OAK 28-30 at 2.9 RPG. Lopez missing IP/ERA data is red flag. Ma...

Missing pitcher stats (both ERA, IP, K/9 null) makes evaluation impossible. Model sees coin flip (53.5% vs market 48.3%). SDP's 2.9 RPG with .547 O...

Model predicts 9.19 runs vs market 7.5, a 1.69-run gap triggering total_far_from_market flag. Missing pitcher ERA/WHIP/K9 stats despite 'full' data...

Spencer Miles has 1.0 IP this season—TOR starter is effectively unknown. Model relies on away_pitcher_last features that cannot be meaningful with...

Model has HOU 51% vs market's 35% — massive disagreement in extreme pitcher park (PF 0.81). Imai's 13.50 ERA in 2.2 IP is catastrophic; Misiorowski...