Devigging 101: How to Strip Sportsbook Margin and Calculate True Market Probability

If you look at a standard two-way sports betting market — such as an NFL point spread set at -110 for the favorite and -110 for the underdog — the implied probabilities do not add up to 100%.

At American odds of -110, a winning bet requires risking $110 to win $100. Converting those odds to implied probability yields:

Implied Probability = 110 / (110 + 100) = 52.38%

When you add the implied probabilities of both sides together (52.38% + 52.38%), the total equals 104.76%.

That extra 4.76% is the sportsbook's margin, commonly known as the vig (vigorish), juice, or overround. It is the fee the bookmaker charges for facilitating the wager, and it is the exact reason why recreational bettors who win 50% of their spreads lose money over time.

To accurately measure Closing Line Value (CLV) or determine whether a line offers positive Expected Value (+EV), you must first strip away this artificial margin. This process is called devigging.

Here is the mathematical breakdown of how devigging works, the three primary methods used by quantitative modelers, and how we apply devigging behind the scenes in the TrueLine engine.

1. Why Raw Odds Distort True Probability

When a sportsbook sets a market, the displayed odds represent two distinct components:

Displayed Market Price = True Fair Probability + Bookmaker Vig

If you evaluate a bet using raw implied probabilities, you are comparing your projection against an inflated number that includes the bookmaker's cut.

For example, if a domestic Canadian sportsbook sets an NFL game total at Over 47.5 (-105) / Under 47.5 (-115):

• Over 47.5 implied probability: 105 / 205 = 51.22%
• Under 47.5 implied probability: 115 / 215 = 53.49%
• Total market implied probability: 51.22% + 53.49% = 104.71%

The raw market claims there is a 104.71% chance the game finishes. To find the actual probability of the Over or Under hitting, we must remove the 4.71% overround.

2. The Three Core Devigging Methods

Not all devigging formulas are created equal. Depending on market liquidity and odds asymmetry, sharp modelers use three distinct techniques to calculate true fair odds.

| Method | What it does | | --- | --- | | Proportional | Splits vig equally relative to implied size. | | Additive | Subtracts a flat margin percentage from both sides. | | Power (Shin) | Accounts for asymmetric sharp/favorite bias. |

Method A: Proportional / Multiplicative (The Standard Model)

The simplest and most widely used method is proportional devigging. This approach assumes the sportsbook distributes its margin proportionally across both sides of the market based on their implied probabilities.

P_true = P_implied / (sum of all P_implied)

Using our Over 47.5 (-105) / Under 47.5 (-115) example:

• True Over probability: 51.22% / 104.71% = 48.92%
• True Under probability: 53.49% / 104.71% = 51.08%
• Combined fair probability: 48.92% + 51.08% = 100.00%

Converting these true probabilities back into fair American odds (using odds = (1 - P) / P × 100 for underdogs and odds = -P / (1 - P) × 100 for favorites) gives:

• Fair Over line: +104.4
• Fair Under line: -104.4

If you can place a wager on Over 47.5 at any price better than +104.4, you hold a positive expected value (+EV) wager relative to this market.

Method B: Additive Devigging

Additive devigging splits the total overround equally down the middle, subtracting half the margin from each selection regardless of the odds.

P_true = P_implied - (Total Overround - 1) / N

Where N is the number of selections in the market (N = 2 for standard spreads and totals).

• Overround: 104.71% - 100% = 4.71%
• Margin per side: 4.71% / 2 = 2.355%
• True Over probability: 51.22% - 2.355% = 48.865%
• True Under probability: 53.49% - 2.355% = 51.135%

While additive devigging works well for balanced -110 / -110 markets, it tends to break down on heavy favorites and longshots (such as moneyline markets at -800 / +550), where subtracting a flat margin can result in distorted or negative probabilities on severe longshots.

Method C: Power Model & Shin Method (Sharp Market Dynamics)

In heavy favorite/longshot markets, bookmakers do not apply margin equally. Research shows that recreational bettors over-bet longshots (the "favorite-longshot bias"), forcing sportsbooks to add disproportionately higher margin to the underdog side to guard against asymmetric liability.

To correct for this, quantitative analysts use the Power Model or the Shin Method.

The Power Model solves for an exponent k such that the sum of all implied probabilities raised to k equals 1.0:

(P_1)^k + (P_2)^k + ... + (P_n)^k = 1.0

By solving for k iteratively, the Power Model strips more margin from the longshot side and less from the heavy favorite, providing the most mathematically rigorous fair probability when devigging sharp market makers like Pinnacle.

3. How Devigging Unlocks True Closing Line Value

Devigging is the bridge that connects raw line movement to actual betting edge.

Suppose you place a pre-game NFL wager on the Detroit Lions spread at +3.5 (-110). By kickoff, sharp market volume pushes the closing line at Pinnacle down to +2.5 (-105) / -2.5 (-115).

To measure how much Closing Line Value you generated, you do not compare +3.5 against +2.5 in a vacuum. You devig Pinnacle's closing line to find the market's consensus fair win probability:

1. Pinnacle closing implied probabilities:
   • +2.5 (-105) → 51.22%
   • -2.5 (-115) → 53.49%
2. Proportional devigged fair odds:
   • +2.5 fair price → 48.92% win probability (+104.4 fair odds)
3. Evaluating your ticket:
   • Your ticket gives you +3.5 at -110 (52.38% break-even threshold).
   • Because +3.5 captures the crucial key number of 3 (~14.8% frequency in the NFL), your true win probability at +3.5 is roughly 54.10%.
   • Your realized edge: 54.10% - 52.38% = +1.72% pure EV edge against the sharpest closing market in the world.

Without devigging the closing market, calculating your true edge relative to market consensus is impossible.

4. Behind the Scenes: The TrueLine Scoring Engine

When you submit a daily prediction on TrueLine Daily, our platform does not simply compare your guessed total or spread against a raw media score.

Behind the scenes, our real-time scoring engine executes a multi-step devigging pipeline at game kickoff:

[Sharp Odds Feed API]
         │
         ▼
[Raw Line: Pinnacle / Benchmark]
         │
         ▼
[Devigging Engine: Power/Shin Calculation]
         │
         ▼
[True Fair Probability Matrix]
         │
         ▼
[User Prediction Comparison] ---> [CLV Score & Leaderboard Points]

1. Kickoff odds ingestion: At the exact second of kickoff, our cron worker pulls closing lines across main and secondary markets from sharp global benchmarks.
2. Dynamic devigging: The engine strips the synthetic hold using a hybrid Shin/proportional devigging algorithm calibrated specifically to the sport's scoring variance (e.g., tight corridors for NFL totals vs. wide distributions for CFB totals).
3. CLV scoring differential: Your submitted prediction slider is mapped directly against the devigged fair probability distribution curve.
4. Multiplier application: If the prediction was made on a secondary market (such as the WNBA), the engine applies a 2.0x point multiplier to the calculated CLV score, rewarding players who successfully model markets with higher inherent noise.

By anchoring our daily puzzle mechanics in devigged fair market value, TrueLine Daily provides a realistic simulator for sports analytics and market prediction.

Want to test your ability to read market odds and spot closing line value? Play today's daily puzzle at truelinebets.com/play, try our free No-Vig Calculator, or track your real betting performance on truelinebets.com.