Games

This is what the betting market thinks, with the bookmaker’s cut taken out. We take every sportsbook’s own two prices on a game, strip its margin ?, and publish the middle number across books. Then we rank the slate by how certain the favorite is.

It is not our forecast, and we are not claiming to beat it. We have no game-winner model. We built one for UFC and tested it against 2,680 real closing prices over six seasons: the market beat it in every single year. This number is our best available estimate precisely because the game market is efficient. There is no edge being sold here, and anyone telling you they can price a moneyline better than the market owes you the same test.

Useful for: seeing a whole week at a glance, filling in a confidence pool, and checking whether a price you were offered is roughly fair. Not useful for: finding a bet we think wins.