Arbitrage Betting Calculator: Pre-Match & Live Sure Bets

Arbitrage betting (or "arbing") exploits pricing differences between bookmakers. When bookmaker A offers Team X at 2.15 and bookmaker B offers Team Y at 2.05 on the same game, the sum of implied probabilities is 46.5% + 48.8% = 95.3% — below 100%. Bet proportionally on both sides and you lock in a 4.9% guaranteed profit regardless of outcome.

The formula: if implied probability sum < 1.00, you have an arb. The profit percentage = (1 / sum) − 1. For each outcome, the optimal stake is: (total stake × implied probability of that outcome) / sum of all implied probabilities. Our calculator handles this automatically for any number of outcomes.

Three-way markets (soccer 1X2) work identically, just with a third term in the sum. Book A prices the home win at 2.15, Book B has the draw at 3.60, Book C has the away win at 4.00 — a modest, common-sized gap, the kind you'll actually see rather than a headline number:

S = 1/2.15 + 1/3.60 + 1/4.00
  = 0.4651 + 0.2778 + 0.2500
  = 0.9929

Margin = 1 − 0.9929 = 0.71%
ROI    = (1/0.9929) − 1 = 0.72%

Stake Home = 1000 × 0.4651 / 0.9929 = $468.45
Stake Draw = 1000 × 0.2778 / 0.9929 = $279.77
Stake Away = 1000 × 0.2500 / 0.9929 = $251.79
Total      = $1000.01 (rounding)

If Home wins: 468.45 × 2.15 = $1007.17
If Draw:      279.77 × 3.60 = $1007.17
If Away wins: 251.79 × 4.00 = $1007.16

Guaranteed profit ≈ $7.16–$7.17 (0.72% ROI)

In practice, the hardest part of arbitrage isn't the math — it's finding opportunities before they close. Sharp bettors and automated tools monitor hundreds of markets simultaneously. Manual arbitrage bettors typically focus on niches: regional leagues, prop markets, or promotions and boosts that create temporary mispricings. Enhanced odds promotions (bookmaker A boosts a team to 3.00 that normally trades at 2.50) are a reliable arbitrage source, since the other side is available elsewhere at market price.

The main risk isn't losing money — it's account restriction. Bookmakers identify arbitrage bettors by bet timing patterns (always betting late on boosted prices), odd stake amounts (calculated stakes like $47.83 instead of $50), and exclusively betting on mathematically optimal prices. Managing exposure across many bookmakers and mixing arb bets with recreational bets extends account longevity significantly.

Stake limits are the part the calculator can't see. It assumes you can place the full $468.45 or $251.79 at the quoted price — real books often cap exactly that market, especially the lagging side that made the arb exist in the first place, at $50–$200. Check the maximum stake at the actual price before you place the first leg; a calculated split that gets rejected or partially filled on one side turns a locked profit into an open position.

Live (in-play) arbitrage runs the same test as pre-match — implied probability sum below 100% — on odds that update every few seconds while the event is in progress. Each bookmaker reprices its in-play markets off its own model of the match state (score, shots, cards, momentum), and the models don't move in lockstep. A goal, a break of serve, or a red card shifts one book's price before another's, and for a few seconds to a couple of minutes the lagging price on one side combines with an already-updated price on the other to dip under 100% implied probability.

That window is short, and the execution is worse than pre-match. In-play bet slips rarely confirm instantly — books insert a short acceptance delay, often several seconds, to reprice against the latest event before accepting your stake, and "the price has changed, accept new odds?" or an outright rejection are routine outcomes, not edge cases. Live markets also carry lower stake limits than the same market pre-kickoff, and the limit tends to be tightest on exactly the lagging price that created the arb — that's the book's own risk system flagging it.

Because of that, the arb percentage the maths above shows you is a ceiling, not an expected return. Model it as: with probability q one leg reprices or gets rejected and you re-hedge at a worse price, giving back roughly slip%; with probability 1 − q you land both legs clean. Expected profit is:

A "detected" 1.8% in-play margin is worth roughly 0.78% once execution risk is priced in — under half the headline number. And that's before commission. Live arbers frequently pair a fixed-odds book with a betting exchange for the faster-moving leg, and exchanges charge commission on net winnings, typically 2–5%.

Commission alone took a 1.42% detected margin down to 0.14% — before the execution buffer above even touches it. That's the honest picture: a two-book live arb that looks like a clean 1–2% edge on screen is routinely flat or negative once you price in rejection risk and, if one leg sits on an exchange, commission. None of this is a reason to avoid live arbing — it's a reason to run the buffer math before you commit the first leg, not after the second one fails to fill.

This page runs the maths, not a live odds feed — there is no automated scanner or WebSocket connection behind it. Take the two or three prices you're actually being quoted right now, run them through the calculator above, then apply the buffer above before you commit capital to both legs.