Spreadsheet FCProbabilities, not promises.

How it works

Four steps, the same every matchday. No gut feelings, no inside tips.

  1. 1. Start from the numbers

    For each game we collect odds from many bookmakers, plus stats such as goals, shots and expected goals (xG). The betting market's average price is a very strong forecast: it already contains what thousands of people and bookmakers know.

  2. 2. Turn odds into fair probabilities

    Bookmakers build a margin into their odds, so the chances they imply add up to more than 100%. We remove that margin to get a fair chance for each outcome, and a fair price (fair odds = 1 รท fair chance).

    We also run our own stats model (form, attack and defence strength, xG) as a second opinion. To be honest: in our tests it hasn't beaten the market, so the picks rely on the market's fair chance, not on the model.

  3. 3. Choose one pick per game

    For each game we pick one bet whose fair odds are close to 2.0 (about a 50% chance): match result, double chance, Asian handicap, over/under 2.5 goals or both teams to score. If no single bet fits, a two-leg bet builder from the same match.

  4. 4. Only pick when the price looks generous

    We compare the fair odds with the standard prices (no boosts) at five UK bookmakers. A pick is only posted when the best price is at least 2% above the fair odds. If no bookmaker is that generous, there's no pick for that game.

How we keep score

Every pick is logged with a timestamp before kick-off, at a flat stake of 1 unit, and checked against the closing odds after the game. The full record, losers included, is on the Results page. We're paper trading first: no real money goes on these picks until a long record shows the method works.

A 2% edge on paper is small, and prices move. Even a good method loses often. Picks near odds of 2.0 lose about half the time.