How a Winning Arena tip is made

Step 1 – rate every team

Each team gets two ratings. The attack rating says how many goals it scores compared with an average side, after taking into account the defences it has faced. The defence rating does the same for goals conceded. We use a year of results, settled at 90 minutes, and every match loses half its influence after 120 days – so this season counts most, but last season still helps. Teams with only a handful of games are pulled towards the league average until the sample grows.

For cup ties and leagues with little data, a second set of ratings links competitions through the games played between them. You can browse every rating on the team ratings page.

Step 2 – expect the goals

For a single match, a team's expected goals are its attack rating × the opponent's defence rating × the league's average for a home or an away side, which builds in each league's own home advantage.

That ratings figure makes up 45% of our final goal estimate. A statistical expected-goals feed supplies 25%, and the total implied by the over/under 2.5 price supplies the last 30% – the part that picks up late team news and line-ups.

Step 3 – give every scoreline a chance

Two goal figures are enough to give each team a chance of scoring 0, 1, 2 or more goals (the Poisson method), and combining the two gives every exact score up to 7–7. Real football has a few more 0–0 and 1–1 draws than the formula expects, so those are lifted slightly and the table rebalanced to 100%.

Every market is then just a sum of scores: a home win is every score where the home side leads, over 2.5 is every score with three goals or more, BTTS is every score where both teams have scored. Half-time markets assume 44% of the goals arrive before the break. The full tables are on the score grids page.

Step 4 – check the bookmakers

Prices carry information that numbers alone can miss, so for the match result we combine 58% of our own chance with 42% of the bookmakers' chance once their margin is removed. Goal lines and BTTS are treated the same way where prices exist.

Every tip then gets its fair odds – 1 divided by its chance. When the best available price beats fair odds by 6% or more, it appears on value bets. Tips priced below 1.15 are never featured, because the reward is too small for the risk.

Step 5 – pick the tip

For each match we publish the selection with the best balance of chance and price. The banker of the day is the strongest of those at a sensible price; the accumulator bets combine strong tips from different matches and always show the chance of the whole slip landing.

Settlement, in public

Tips lock at kick-off and are settled on the score after 90 minutes plus stoppage time. Extra time and penalties do not count, and postponed games are void. Results go straight into the prediction archive, the market profit table and the calibration check.

What no tip can know

A red card in the tenth minute, a goalkeeper's nightmare, a coach who rests half his team without warning, a side with nothing left to play for – numbers only know what has happened and what prices say. That is why every tip carries a probability, and why even our strongest tips will sometimes lose.