The book's net in each result, from all the crowd's money on the game (spread + moneyline). It keeps the losing stakes and pays the winners' profit. Shaded = the result that makes the book the most; green outline = what happened. Book's side = the spread bet that wins in the book's best result (weeks 3-7: 41-27, 12-3 in primetime). It's a cross-check on the verdicts below, not a separate pick.
Game
Spread
Fav covers
Fav wins, dog covers
Dog wins outright
All $ on game
Book's side
Result
Spreads
Game
Side
Robinhood ask
Pinnacle fair
Edge
Crowd share
$ on this side
All $ on team (spread + ML)
Tape $
$ traded
Verdict
Result
Book take
Book gave
Totals
Game
Side
Robinhood ask
Pinnacle fair
Edge
Crowd share
$ on this side
All $ on team (spread + ML)
Tape $
$ traded
Verdict
Result
Book take
Book gave
No model here. Each row is one side of a spread or total, so a 12-game slate is 48 rows (24 with spreads or totals only). Click any column header to sort; click again to flip. Robinhood ask = what a $1 contract on that side costs. Pinnacle fair = the sharpest book's de-vigged chance at the same line (a ~ means Pinnacle's line was a whole number, Robinhood only lists half points, so the nearest strike is shown and the fair price is nudged for the half point).
Edge = fair minus ask, in points. It is shown for reference only: betting the side Robinhood sells below Pinnacle's price lost money here. LEAN = the only trends that have paid so far, on 4 MLB days: the MLB run-line favorite (-1.5) and the MLB over. Small stakes only; NFL spreads (rebuilt Oct 16, weeks 3-6): the side with less of the crowd's spread money covered 29 of 42; the side with more covered 13 of 42. How much money is in the market decides it: in a spread market of $300k+ the less-money side covered 18 of 20, under $300k 11 of 22. STRONG BET = 🌙 primetime (Thursday, Sunday night, Monday), the side with less money: 10-0, always the underdog so far. BET = Sunday daytime, the side with less money in a $300k+ market, favorite or underdog: 10-2. Smaller markets are a PASS: check back near kickoff, when most of the money is in. NFL totals: LEAN the crowd's side (26-15); primetime LEAN the over (7-3). Four weeks of data, so size accordingly. All $ on team = spread + moneyline money on that team added together, and its share of all the money on the game (red = the team with most of the money). ML $ (under the crowd share) = the moneyline crowd on the same team; ⚠ ML split = moneyline money is on the other team (e.g. KC-SEA week 7: 85% of the spread $ on SEA, 77% of the ML $ on KC). The verdict still fades the spread money, because in those split games fading the spread $ covered 17-8 and fading the ML $ 8-17. Crowd $ on this side = share of the taker dollars on this exact market from the trade tape (each trade credited to the side the bettor backed); $ on this side = those taker dollars; Tape $ = both sides together; Book take = once graded, the dollars on a side that lost (what the winners collected), Book gave = the dollars on a side that won; the tiles sum them. $ traded = the whole ladder of lines for this game.
Rows lock at first pitch / kickoff (volumes only ever go up before that) and grade from the final score. The scorecard at the bottom is the point: does the side with more of the crowd's money cover less than its price says?