I've lost enough on "sharp" lines to know: when every book agrees, pause.
Fair play value bets are wagers where the market has converged on a single number — low or zero dispersion (gap between books' lines) — and that number sits at or near a key threshold where a half-point move swings the expected value (long-run profit per dollar). The market has "played fair" by agreeing on the price; the bettor finds value by comparing that settled price to a model or historical tendency.
What are fair play value bets?
Fair play value bets are wagers where the market has converged on a single number at a key threshold, creating an edge for models that disagree by a half-point.
What the consensus lines show this week
Our live feed captures home spreads across NFL and college slates. Several NFL games show zero dispersion at the reporting book, meaning the market has settled.
Chiefs -5.5 over Buccaneers . One book. Zero disagreement. Market settled.
Ravens -3.5 over Vikings . One book. Zero disagreement. Market settled.
Falcons -3.5 over Colts . One book. Zero disagreement. Market settled.
Bills -2.5 over Browns . One book. Zero disagreement. Market settled.
Seahawks -4.5 over Titans . One book. Zero disagreement. Market settled.
Cowboys -1.5 over Cardinals . One book. Zero disagreement. Market settled.
Commanders -1.5 over Lions . One book. Zero disagreement. Market settled.
On the dog side, Giants +3 at Dolphins . One book. Zero disagreement. Market settled.
Saints +1.5 at Rams . One book. Zero disagreement. Market settled.
Bears +1.5 at Bengals . One book. Zero disagreement. Market settled.
Eagles +1.5 at Patriots . One book. Zero disagreement. Market settled.
Where the key-number edges live
The Falcons -3.5 and Ravens -3.5 sit on the 3.5 hook (half-point off a key number) — one half-point off the most common NFL margin.
If a model projects the true line at -3, the favorite at -3.5 is a fair play value bet because the market has agreed on the hook.
The Chiefs -5.5 sits between the 3 and 7 key numbers. It is less sensitive to a half-point move but still reflects a settled market view on Kansas City's superiority.
The Giants +3 is the mirror: an underdog getting the key number 3 in a market that has agreed on the price.
College lines: wider gaps, same principle
College football produces larger spreads and different key numbers. North Carolina -7.5 at TCU and the duplicate entry at -8 show the market settling around the 7/8 zone — a college key number where touchdown-plus-field-goal combinations land.
San Jose State -38.5 at USC and the -38 variant reflect a talent gap the market has fully priced. Dispersion is zero but the line is far from any decision threshold, so "value" here means trusting the market's assessment of the mismatch.
Other settled college lines: NC State -5.5/-6 at Virginia . Jacksonville State -7 at North Dakota State appears twice at the same number — a true duplicate, not a dispersion signal. Sacramento State -8.5/-9.5 at Eastern Michigan . Hawaii -5.5 at Stanford appears twice at the same number . Memphis -5.5 at UNLV appears twice at the same number .
How to use this tonight
Check Falcons -3.5. If you think Atlanta wins by 4 or more, bet it. Track the result in a notes app.
What the duplicates tell you
Several games appear twice with a half-point difference: NC State -5.5/-6, Sacramento State -8.5/-9.5, Colorado -7/-6.5 , New Mexico State -31/-31.5 . This happens when our feed captures the same game from two sources with slightly different timestamps or book subsets. The half-point gap is the dispersion — it tells you the market has not fully converged even if each individual source shows zero internal dispersion. Treat these as "near-settled" and wait for the line to unify before calling it fair play.
Hawaii and Memphis appear twice at the same number (-5.5 both times). Those are true duplicates, not dispersion signals.
Bottom line
A fair play value bet is not a guarantee — it is a disciplined filter. Find games where the market agrees on the price, that price sits at a key-number decision point, and your model disagrees by at least a half-point. Bet it, log the CLV (beating the closing line), and let the sample decide. The lines above are this week's candidates; the backtest these lines free lets you test the logic on past slates before you risk a dollar.
Bet responsibly — set limits, never chase losses.
NFL ATS cover-margin distribution
Distribution of (final margin − closing spread) across an NFL season. Roughly normal with mean ≈ 0 and standard deviation ≈ 13 points, which is why most ATS edges live in the ±1.5 point window. Chart ELO standings winner: free-llm-a on team-summary::graded-record (ELO 1516, artifacts/chart-elo/standings.json).
Model calibration: predicted vs observed
Predicted win probability bucket vs the empirical win rate inside that bucket on the test set. Points on the y=x reference line are perfectly calibrated; points below mean the model is overconfident in that bucket.
