Loading the board.
Loading the board.
The chance each starting quarterback throws a pick this week. Books post this at 0.5, so what matters is whether he throws one at all — not how many he averages.
SEA @ DENThu 8 PM59%2
LAC @ KCSun 4 PM57%1
NE vs NYJSun 1 PM54%
LAR vs ARISun 4 PM51%
ARI @ LARSun 4 PM49%1
NO @ NYGSun 1 PM48%
LV vs BUFSun 4 PM47%1
TB vs PITSun 1 PM47%2
PIT @ TBSun 1 PM46%1
NYG vs NOSun 1 PM46%2
IND vs TENSun 1 PM45%2
PHI vs CARSun 1 PM45%
DAL @ GBSun 8 PM44%2
TEN @ INDSun 1 PM44%1
GB vs DALSun 8 PM43%1
BUF @ LVSun 4 PM43%
JAX vs HOUSun 9 AM41%1
CAR @ PHISun 1 PM41%
ATL vs CHISun 1 PM39%
KC vs LACSun 4 PM39%1
CHI @ ATLSun 1 PM39%
DEN vs SEAThu 8 PM38%
SF vs WSHMon 8 PM36%
CLE vs BALSun 1 PM34%
BAL @ CLESun 1 PM33%
WSH @ SFMon 8 PM32%
NYJ @ NESun 1 PM28%1
HOU @ JAXSun 9 AM27%1We haven't graded this board against a walk-forward run yet, so read it as a projection, not as an edge. What we have measured is the market: across 1,563 posted 0.5 lines from 2023 to 2025, betting every over blind returned −6.3% and every under −2.3%, both at the best price of five books. Both sides lose to the hold — this is a tight market. It's also why the number above is a chance and not an average: a quarterback averages 0.67 picks a game but throws one in only about 48% of starts, so an average read against a 0.5 line is an over on nearly every passer in the league.
How often our number landed on the right side of the line a book actually posted. Every quoted line counts, not just the big disagreements — and a push settles as neither. There are no units here: two seasons of walk-forward say these lines are efficient, so we grade accuracy and don't claim an edge we can't show. Green means it beat what picking a side with no model at all would have scored.
Every number here comes from the same drive-level simulation as the matchups board, so a player's line and his team's projected points are the same game. Not betting advice — 21+, play responsibly (1-800-GAMBLER).