Do the math the book already did
Strip the vig from any odds
Every price you see in a betslip has the bookmaker’s margin baked in. This tool takes it back out — for any odds you type in, or head-to-head across the crypto books we sample.
Here is exactly what this tool computes, with a worked example — and below it, the same dated samples the “compare books” mode uses. (The interactive calculator needs JavaScript; the reference below always works.)
Standard US pricing -110 / -110: each side implies 110/210 = 52.38%; the book totals 104.76%. Overround = 4.76%; hold = $10 kept / $220 handle = 4.55%; fair odds = +100 / 2.00. A 3-way example: 2.50 / 3.40 / 2.90 sums to 103.89% → a 3.89% overround; multiplicative fair odds ≈ 2.60 / 3.53 / 3.01 (our computation).
| Book | Event | Market | Odds (as captured) | Overround | Sampled |
|---|---|---|---|---|---|
| Duel | France vs Morocco — FIFA World Cup | football-1x2 | 1.62 / 4.10 / 6.40 | 1.74% | 2026-07-08 |
| Duel | Spain vs Belgium — FIFA World Cup | football-1x2 | 1.66 / 4.10 / 5.60 | 2.49% | 2026-07-08 |
| Duel | G2 vs T1 — LoL MSI 2026 (live) live | esports | 1.75 / 2.14 | 3.87% | 2026-07-08 |
| Duelbits | Spain vs Belgium — FIFA World Cup | football-1x2 | 1.64 / 4.05 / 5.45 | 4.02% | 2026-07-08 |
| Duelbits | France vs Morocco — FIFA World Cup | football-1x2 | 1.61 / 4.00 / 5.90 | 4.06% | 2026-07-08 |
| Cloudbet | Norway vs England — FIFA World Cup | football-1x2 | 4.35 / 3.75 / 1.81 | 4.90% | 2026-07-08 |
| Cloudbet | Spain vs Belgium — FIFA World Cup | football-1x2 | 1.61 / 3.95 / 5.60 | 5.29% | 2026-07-08 |
| Cloudbet | France vs Morocco — FIFA World Cup | football-1x2 | 1.58 / 3.80 / 6.35 | 5.35% | 2026-07-08 |
| Cloudbet | Argentina vs Switzerland — FIFA World Cup | football-1x2 | 1.71 / 3.45 / 5.55 | 5.48% | 2026-07-08 |
| Cloudbet | Tennis match-winner, in-play (Wimbledon week; teams not recorded) live | tennis | 2.08 / 1.74 | 5.55% | 2026-07-08 |
| Thunderpick | Atreides vs Eternal (live, esports hub) live | esports | 1.15 / 4.80 | 7.79% | 2026-07-08 |
| Thunderpick | GamerLegion vs Xtreme (esports hub) | esports | 4.24 / 1.89 / 2.90 | 10.98% | 2026-07-08 |
| Thunderpick | REKONIX vs MOUZ (esports hub) | esports | 6.11 / 2.04 / 2.19 | 11.05% | 2026-07-08 |
| Thunderpick | Power Rangers vs BB (esports hub) | esports | 10.18 / 2.55 / 1.60 | 11.54% | 2026-07-08 |
| Duel | Dota 2 EWC 2026 match (live) live | esports | 10.00 / 3.20 / 1.38 | 13.71% | 2026-07-08 |
Every row is a real price we captured, with its capture date. Overround = (Σ 1/odds − 1) × 100. Sample counts are still below our n≥6 bar for publishing medians — treat these as first data points, not averages.
The five methods, honestly labelled
| Method | How it removes the margin | Best for |
|---|---|---|
| Multiplicative | Divides each implied probability by the book total — margin removed proportionally. | Balanced 2-way markets (spreads, totals, tennis) |
| Additive | Subtracts an equal slice of the margin from every outcome. Can go negative on big longshots — the tool warns when it does. | Quick estimates on near-even markets |
| Power | Raises implied probabilities to the exponent that makes them sum to 1 — more margin comes off the longshots. | Lopsided odds; favourite–longshot bias |
| Shin (1993) | Models the margin as the book's defence against informed money (H.S. Shin, The Economic Journal 103, 1993) and prices it out accordingly. | 3-way / 1X2 markets; the academically grounded pick |
| Worst-case | Takes each outcome's lowest fair probability across the four methods above — a practitioner convention for conservative value checks. | Screening bets where a false positive costs money |
Vig, hold, overround, juice — one thing, two bases
UK books say overround or margin: the excess of summed implied probabilities over 100%. US bettors say vig or juice for the same excess — and hold, strictly, is that excess divided by the book total, i.e. the book’s expected keep per dollar of handle. That is why the same -110/-110 market reads “4.76%” on one site and “4.55%” on another; neither is wrong, they just use different bases. This tool always shows both, labelled. The margins we publish on the index and the market pages are overround.
Why a crypto-book preset matters
Generic de-vig calculators start from a blank field. Ours also carries the dated prices we actually captured from Duel, Cloudbet, Thunderpick and Duelbits — so you can see who was cheapest on the same fixture without hunting the odds yourself. In our 8 July 2026 sample the spread between books on one World Cup fixture was wider than most bettors expect: 1.74% at the cheapest book against 5.35% at the priciest — on identical outcomes. Check the margin history as re-measurements accumulate.
Common questions
What does de-vigging odds do?
It removes the bookmaker's margin from a set of odds to reveal the fair (no-vig) probabilities and prices. If a book quotes -110/-110, both sides imply 52.38% — a 104.76% total. De-vigging normalises that back to 50/50, i.e. fair odds of +100 / 2.00.
Which de-vig method should I use?
Multiplicative is the standard default for balanced two-way markets. For lopsided odds or three-way (1X2) markets, power or Shin shift more of the margin off the longshot — Shin (1993) is the academically grounded choice. Worst-case takes each outcome's lowest fair probability across methods, giving the most conservative value estimate.
What is the difference between vig and hold?
Both describe the same juice against different bases. Overround/vig measures the excess over 100% of the book total; hold measures the book's expected keep as a share of total handle. At -110/-110 that's 4.76% overround but 4.55% hold. Sites disagree on labels, so this tool shows both.
Does the calculator send my numbers anywhere?
No. Everything runs in your browser — there is no server call, no tracking of your inputs, and nothing to sign up for.