Methodology

The math, in plain English.

Bookmakers price events to make money on the average bettor. Our models price the underlying probabilities. When the gap is big enough, that bet is +EV. Stake all sides across multiple books and you have arbitrage. Both compound over volume.

Last updated 25 May 2026.

Positive EV

What is +EV?

Imagine a coin flip. Heads lands half the time, so the fair price is 2.00. Stake $100 and you get $100 back on average.

Now imagine a bookmaker offers 2.10 odds on heads. Over many flips, you'd still win 50% of the time, but you'd be paid at 2.10. That extra 0.10, the difference between fair and offered, is your edge.

We don't bet coin flips. We scan thousands of sports markets where bookmakers misprice odds, and surface those mispriced opportunities the moment they appear. That difference, over hundreds of bets per day, compounds into consistent profit.

Arbitrage

What is Arbitrage?

Two bookmakers price the same event differently. Occasionally the gap is wide enough that backing both outcomes, one at each book, locks in a return whichever way it lands, provided both legs get on at the quoted prices.

The margins are thin. One to three percent of what you stake is the realistic range, so the return comes from repeating it rather than from any single bet.

Both legs have to get on before either price moves, which is why every alert is timestamped to the second. Speed matters more than stake size here.

The formula

Expected value is one line of arithmetic:

EV = (chance of winning × profit if it wins) − (chance of losing × stake)

Run it on the coin flip above. Heads pays 2.10, so $100 staked wins $110 profit half the time and loses $100 the other half: (0.5 × $110) − (0.5 × $100) = +$5 per $100 staked. Positive means the price is better than the outcome deserves. That is the only thing we look for.

Why volume matters more than any single bet

Variance dominates small samples. A +EV bet at +5% can lose ten times in a row before the math asserts itself. That's why we post 1,000+ bets per day across both feeds. The Law of Large Numbers applies: with enough bets at positive expectation, realised returns converge to expected returns.

Practical implication for the member: staking flat per bet at a small fraction of bankroll is how the math compounds. Most members size at 1–2% of bankroll per +EV pick, scaled by edge size via Kelly fractioning.

How we price probabilities

Our pricing models combine:

Sharp-book consensus: books with the highest limits and lowest margins (Pinnacle, Betfair Exchange) are the closest proxies to fair price. We weight them heavily.
Devigging: removing the bookmaker margin from their own odds to recover implied probabilities, then aggregating across multiple books.
Market-specific models: for player props and niche markets where sharp-book lines don't exist, our own statistical models price the event from historical and situational data.

The output is a fair probability per outcome. We compare every book we cover against the fair price, every second. The gaps that clear our threshold are posted.

What we don't do

No tipsters. No human cherry-picking. No “locks of the day.”
No guaranteed results. Past performance does not guarantee future results.
No proprietary mystery box. The methodology is on this page. The execution is in the volume and the speed.