Methodology

Negative odds (-110) show how much you must risk to win $100; positive odds (+150) show how much you win on a $100 bet. Every price hides a win probability: for negative odds it's |odds| ÷ (|odds| + 100) — so -110 implies 110/210 ≈ 52.4%. For positive odds it's 100 ÷ (odds + 100) — so +150 implies 40%. Decimal odds are the total payout multiple: -110 → 1.91, +150 → 2.50.

Add up the implied probabilities of both sides of a market and you get more than 100% — typically 104–105%. That extra is the vig, the book's built-in fee. LineLens removes it with the standard multiplicative method: each side's fair probability = its implied probability ÷ the total. A -110/-110 market devigs to exactly 50/50.

When you type odds in manually with no second side, LineLens assumes a typical 4.5% total vig and labels that probability "estimated."

For found bets, LineLens devigs every sportsbook's own market separately, then combines them: the headline consensus is the equal-weight average, and the median, low, high, range, and standard deviation are always shown alongside it. Player-prop markets are devigged per player and per line — Over/Under pairs on the same number only. One-sided markets (anytime-TD style) are excluded entirely because their vig cannot be honestly removed, and different lines (-3 vs -3.5) are never mixed. There is no secret weighting — every book counts equally.

EV = fair probability × decimal payout − 1. It's the average profit or loss per $1 if the same bet were made many times: +2% EV means +$2 per $100 on average. Positive EV does not mean the bet wins — a +EV bet at 40% still loses 60% of the time. It means the price is better than the market's best estimate of the risk.

Books that agree tightly make the consensus trustworthy; books that disagree make it noisy. LineLens labels agreement from the spread of per-book fair probabilities: High = range ≤ 2.0 points and σ ≤ 0.75; Low= range > 4.5 points or σ > 2.0; everything between is Medium. The numbers behind the label are always displayed. Low agreement is information — it is not automatically value.

LineLens records every price change it observes and charts them per book. The earliest snapshot is the earliest price LineLens saw— not the official opening line, which our provider doesn't confirm. A "steam-like move" is flagged when 3+ books move the same direction by ≥1 probability point within 2 hours. A "possible stale price" is flagged when one book sits ≥2 points from the market median after 2+ other books have updated. Movement is a description of the market, never proof of an outcome.

The closing price is the market's most-informed estimate. Beating it consistently is the best available evidence of a good process, independent of short-term results. Moneyline CLV = closing fair probability − your implied probability at placement, in percentage points. Spread/total CLV is shown in points beaten ("you bet -3, it closed -4.5: +1.5 points") with the price change shown separately — never mashed into one score. LineLens uses the last price it observed before the game started as its close, labeled as an approximation.

Parlay probability = the product of leg probabilities, which assumes independence. Legs from the same game are not independent — a game's script pushes them together. LineLens flags these combinations and names the shared dependency, but never adjusts the probability with an arbitrary factor: the honest move is showing the independence-based number with a clear warning.

• Odds can be delayed; always confirm the price at the book.
• Not every sportsbook is included, and coverage varies by market.
• Positive EV is an estimate; the market consensus can be wrong.
• Line history covers what LineLens observed, not full market history.
• Closing lines are approximations, not provider-confirmed closes.
• Past results never guarantee future results.
• LineLens is an analytics tool. It does not guarantee winnings.