↻ Refresh in progress — pulling new lines & re-running predictions… this page updates automatically when it finishes.
Date:
Best Bets — 2026-07-02
Generated 2026-07-02 01:23:13 AM ET
PAPER RESEARCH ONLY — not betting advice. Every ML model here is NEGATIVE against the opening line since ~2026-05-17; the season-long “break-even” came from early-season games only. These are research flags, not an edge.
Filter: at least one production ML model clears the canonical gate — EV ≥ 7% vs the OPENING consensus line (max_model.betting_gate; one strict rule everywhere, Dave 2026-07-01). Cards sorted strongest-consensus first.
Picks · 4 sources per game
One card per game (consensus sorted top). Each row shows what Alibi, Flyer, Today's Picks, and Side Model picked for each side, with BET + (+pp) format and a Cons cell counting how many of the 4 agree.
— no qualifying games —
No games currently meet the best-bets filter. Flagged when our model's win probability implies at least a 7% edge over the OPENING consensus line (median across books). Paper research only. Check the Picks page for the full slate.
BET shade gradient (model edge vs current market): 5pp 6pp 7-9pp 10+pp. Cell shows BET TEAM at the model fair odds + edge magnitude.