Every term on the live board, in plain English, two sentences each.
What the spread or total should be given the score, clock, possession and both teams' scoring rates, with no sportsbook margin included. Everything on the live board is built around the gap between this number and the book's.
The spread, total or moneyline the sportsbook is actually offering right now. It includes the book's margin and its read of where public money is going, which is why it can sit points away from fair.
Book line minus fair line, in points. A spread gap of -1.6 means the book's number is a point and a half friendlier to the home side than the model says it should be.
The sportsbook's built-in margin, usually priced as -110 on both sides of a spread. You must win 52.4% of even bets at -110 just to get your money back — that's the tax every edge has to clear.
The win probability a bet needs to show a profit at its odds. At -110 it's 52.4%; at -150 it's 60%; at +150 it's 40%. The board shows break-even next to the model's cover probability on every verdict.
The model's cover probability minus the break-even probability, in percentage points. Positive edge inside the vig is a wait; enough edge past it is a bet.
The model's estimate that a side beats the book's number — e.g. that the favorite wins by more than the posted spread. It comes from the fair line and a per-sport margin distribution.
How many more points the model expects one team to outscore the other by over the rest of the game. Fair spread is current margin plus expected remaining margin.
The point value of having the ball, derived from the team's scoring rate per possession. In football it's worth roughly two to three points and swings the fair spread every time the ball changes hands.
How fast possessions or scoring are being used relative to the pregame expectation. A slow football game bleeds total value; a fast basketball game inflates it.
A quarter where one or both offenses produce nothing. Each no-show quarter drags the fair total down by a fixed penalty, because books consistently cut live totals slower than the clock does.
The most common NFL margins, because of field goals and touchdowns. A fair spread crossing 3 or 7 is worth far more than the same move from 9 to 10, and verdicts treat those crossings with extra weight.
How much the current state should be trusted as the clock runs down. A six-point lead with one minute left is near-certain; the same lead at half is barely information. It scales how aggressively the fair line follows the score.
A ceiling on how far one hot stretch can move a team's scoring rate. Without it a 12-0 run would rewrite the whole model; with it, runs move the number only as much as their real information value.
Baseball's expected runs for each combination of runners and outs. Bases loaded with nobody out is worth over two runs on average; the fair total re-prices the moment the state changes.
Baseball's spread, almost always 1.5 runs. Because it's really a win-probability bet, the fair price comes straight from the model's win probability rather than a margin line.
The model's chance each team wins the game, from the fair margin and time remaining through a per-sport distribution. Shown next to the book's implied probability so you can see the disagreement.
The win probability baked into the book's moneyline odds after removing the vig. When the model's win probability sits several points above implied, the moneyline is the market with the edge.
A setting that mixes the book's number into the model's baseline. It defaults to zero — pure model — because the product exists to price games without borrowing the market's answer.
A staking formula that sizes a bet from your edge and the odds. The board uses a fractional Kelly with a hard cap on bankroll percentage, because full Kelly is a rollercoaster nobody enjoys.
The point late in a game where so little time remains that verdicts stop being meaningful. Past the cutoff the board shows numbers but stops printing bets.
The book pausing its live line while it re-prices — usually after a score or a review. Settled books make stale gaps, so the board flags movement and expects the number to catch up.
A bet that lands exactly on the number and refunds. Totals and spreads on whole numbers carry push risk, and a fair line sitting exactly on the book's number is a pass, not a coin flip.
How the model treats a game headed to OT. In football a tied game late prices as close to 50/50 with a small possession edge; totals get an OT bump because extra time adds expected points.