The Edge Board
Where the machine disagrees with the market. Our model grades every game across all four leagues, and these are today's biggest gaps between its win probability and the sportsbook's price. Value the books may be missing, not guarantees.
Lines as of 6:32 PM ET, prices move; check your book.
SA @ PHX
Model backs PHX · 33% to win · market +296
Strong
CLE @ CHA
Model backs CHA · 34% to win · market +282
Strong
BOS @ PHI
Model backs BOS · 56% to win · market -109
Moderate
BUF @ MIN
Model backs MIN · 42% to win · market +164
Moderate
MIL @ NY
Model backs MIL · 15% to win · market +770
Moderate
IND @ DAL
Model backs IND · 44% to win · market +143
Moderate
DET @ STL
Model backs DET · 46% to win · market +124
Slight
LAL @ ATL
Model backs ATL · 60% to win · market -144
Slight
NO @ GS
Model backs NO · 21% to win · market +426
Slight
WSH @ DET
Model backs WSH · 22% to win · market +394
Slight
NYI @ SJS
Model backs SJS · 46% to win · market +120
Slight
MIN @ MTL
Model backs MTL · 50% to win · market -102
None
This is the teaser board.
The per-league Edge Finder shows every edge with fair odds and Kelly stake sizing.
What "edge" means here
For every upcoming game, the model produces a win probability independent of the betting market. We convert the sportsbook's moneyline into its implied probability and compare. When our number is meaningfully higher than the market's on a side, that gap - the "edge", is where a bet has positive expected value if the model is right. Bigger gap, bigger rating. It's a starting point for research, not a lock.