Value bets & arbitrage
Pinnacle's line is the sharpest public estimate of what an outcome is worth — but it is quoted with the bookmaker's margin baked in. GoldEdge takes that margin out, flags every Bet365, Betano and Stake price paying more than the fair odds, and catches the moments those books disagree enough to lock a profit outright.
| Line | Home | Draw | Away |
|---|---|---|---|
| Quoted | 2.10 | 3.40 | 3.90 |
| Fair | 2.156 | 3.491 | 4.004 |
Bet365 pays 2.30 on Home, against a fair 2.156+6.67%
Coverage
The first thing worth knowing, before anything persuasive: whether your book and your sport are in here.
Reference
Pinnacle
Low margin, high limits, winners welcome — which is what forces its line to be accurate. Every fair price is derived from it.
Scanned
Bet365 · Betano · Stake
Compared against the fair line for value, and against each other for arbitrage.
Cross-checked
OrbitX exchange
The order book behind the price. A soft quote that cannot beat the exchange's lay side is not really value.
Sports & markets
Soccer · Tennis · Basketball
1X2, over/under, moneyline, Asian handicap, draw no bet, double chance and first half — whatever each book quotes.
The product
This is the whole application. Every panel below is a page you land on after signing in — the route above it is where it lives.
/valuebets/dashboard
Every value bet currently open, ranked by edge. Take one and your record is kept against the price you actually got — not the price on screen when it was found.
/valuebets/live
Value on matches already running, with the score beside it. Soccer and totals settle themselves as the match ends; first-half and tennis markets propose a result you confirm.
/valuebets/drops
Fixtures no soft book has matched yet, with the sharp line tracked anyway. A price that has fallen 9% or more from its opener is money moving — visible before there is anything to bet into.
/valuebets/history
What you took, what it returned, and whether you beat the closing line — broken out by signal, market and odds band, so you can see which slice is actually working.
ROI by signal. Example figures — your record is your own.
/arbitrages/pregame
Price sets across books adding up to under 100%, with the stake split worked out — and the profit reported for the worst outcome once stakes are rounded, which is the only figure actually guaranteed.
/odds-comparison
Every matched fixture's prices side by side across all four books, with the no-vig fair line beside them. The shopping view — where the best price is, before you decide whether it is value.
Fixtures and figures in these panels are worked examples — the board itself is behind the login.
The arithmetic
One book too generous against the true probability, or two books disagreeing enough that every outcome can be backed at a profit. Both are division you can check.
Pinnacle's three prices imply more than certainty. Remove that overround proportionally and what is left is the fair price.
| Pinnacle | Price | Implied | Fair |
|---|---|---|---|
| Home | 2.10 | 47.62% | 2.156 |
| Draw | 3.40 | 29.41% | 3.491 |
| Away | 3.90 | 25.64% | 4.004 |
| Total | 102.67% | 100.00% |
100% true probability2.67% margin
Two books, two sides of one market, adding up to less than certainty. Split the stake in proportion and the return is the same whichever way it goes.
| Leg | Book | Price | Stake |
|---|---|---|---|
| Over 2.5 | Bet365 | 2.10 | 48.78 |
| Under 2.5 | Pinnacle | 2.00 | 51.22 |
| Book sum | 97.62% | 100.00 |
97.62% covered by the two prices2.38% gap
The hard part
Spotting a sum under 100% is arithmetic. Not acting on a false one is the product — a bad arbitrage loses both stakes, where a bad value bet loses one.
Pinnacle gives a tennis moneyline action once a set is complete; Bet365 and Stake void on retirement. Legs only combine inside one rule class, and a book whose rules haven't been checked is excluded rather than assumed.
Over and under must sit on the same total, and quarter lines are dropped outright — half the stake pushes, which turns a locked return into a coin flip.
A book quoting a draw is excluded from a two-way market. That one check is what stops regulation-time basketball being priced against including-overtime.
Every leg is under three minutes old, from at least two distinct books, on a fixture matched with high confidence. Margins above 10% are discarded — past that a palpable error is likelier than an opportunity.
Limits
Everything above is checkable. This is the part you would otherwise have to find out for yourself.
There is no model and no opinion — only arithmetic on public prices. A +6% bet loses about as often as its odds say it should. The edge shows up across hundreds of bets or not at all.
Accounts that consistently beat the price get restricted. Staking sensibly and spreading across books lengthens the runway; nothing removes the ceiling. Pinnacle is the exception, which is part of why it is the reference.
Get one and miss the other and you are holding a live bet, not a locked position. Prices move in seconds, which is why every leg here carries its age.
Real value sits in low single digits and real arbitrage lower still. Anything advertising much more than that is usually a stale price, a mismatched fixture, or a lie.
Questions
A price that pays more than the outcome's true probability justifies. If the fair odds are 2.156 and Bet365 offers 2.30, each stake returns slightly more than it risks on average. It says nothing about whether this particular bet wins.
An arbitrage — a surebet — is a set of prices across different bookmakers whose implied probabilities add up to under 100%. Backing every outcome in the right proportion returns more than you staked no matter what happens. A value bet is an edge that pays out over many bets; an arbitrage is locked in on a single fixture, which is why the checks around it matter more than the arithmetic.
Pinnacle operates on low margins and high limits, and welcomes winning bettors rather than restricting them. That forces its line to be accurate, which makes it the best public proxy for true probability once the margin is removed. Every fair price on the board is derived from it.
Pinnacle is the sharp reference. Bet365, Betano and Stake are scanned for value and combined for arbitrage, across soccer, tennis and basketball — 1X2, over/under, moneyline, Asian handicap, draw no bet, double chance and first-half markets, depending on what each book quotes.
Every one is checked before it is shown: the legs must be under three minutes old, on the same line, with matching settlement rules and a complete set of outcomes, on a fixture matched with high confidence. Margins above 10% are discarded rather than shown, because past that a palpable error is likelier than an opportunity. What survives is the arbitrage, not the noise.
No. Take what you want and ignore the rest; untaken bets stay out of your profit and loss. Many users filter to a narrow slice — one sport, one market, a minimum edge — and work only that.
No. There are no predictions and no opinion. It is arithmetic on public prices: the Pinnacle line, de-vigged, compared against what Bet365, Betano and Stake are offering right now.
Scans run continuously, every few seconds. Value prices are short-lived — a book that is slow to move is exactly the one worth betting into — and arbitrages are shorter still, usually gone within minutes, which is why every leg carries its age.
Free, no account
The odds board, the calculators and the guides are all public. The whole claim is that this is arithmetic you can reproduce, which is worth very little if you have to register to test it.
Live data
Odds comparison →
Every matched fixture across all four books with the no-vig fair price beside them. The actual scanner output, unlocked.
Calculators
Run the maths →
Arbitrage stake splits, no-vig fair odds, expected value and Kelly staking — the same code the engine uses, running in your browser.
Guides
How it works →
Value betting, arbitrage, removing the vig and closing line value — with every example worked from prices you can check.
Comparisons
Versus the paid tools →
How this sits against RebelBetting, OddsJam, BetBurger and Surebet.com — including where they are the better choice.
Shipping
Changes deploy the day they are ready — the latest below, newest first.
A research tool, not a tip service. Positive expected value is an edge across many bets, not a prediction about any one of them — lines move, and a good price still loses often. Arbitrage depends on both legs being placed and settling as expected; a voided leg leaves a live position. Bet only what you can afford to lose, and stop if it stops being a decision you're making calmly. 18+.