Crystal Palace 1–4 Manchester City

← Back to the board

What the chances were worth

The scoreline says 1–4; by chance quality the match ran 0.57–1.45.

What the model expected

Its probabilities for this match, logged before kick-off and never changed. The result is highlighted.

Before kick-off the model favoured Crystal Palace to win (38%); it finished with Manchester City to win. A miss — it gave the actual result 35% beforehand.

One match tells you nothing on its own — whether these percentages are any good is the whole point of the calibration page, where every one of them is scored.

The brain read

Manchester City created the better chances — the xG read 0.6–1.4 and the shot count agreed (9–18, 1–9 on target). Manchester City got into the box far more often (12 shots inside it to 6) and Manchester City held 73% of the ball. Manchester City's finishing ran hot — 4 from 1.4 xG, 9 of 18 shots on target (a finishing read inferred from shots on target and goals vs xG — not measured shot placement). On the balance of chances, the 1–4 reads fair.

A second opinion written by rule from the numbers on this page — where the shot picture and the xG disagree, it says so rather than picking a winner. Backward-looking only; it claims nothing about any future result.

The match in numbers

As returned by the data feed at full time. Bars show each side’s share of the stat.

The market’s read

Shortest-priced scorelines: 1-1 7.54 · 1-2 7.72 · 0-1 8.11 · 0-2 8.72 · 1-3 11.61

Consensus of bookmaker prices captured 2026-08-28 21:10 UTC. These are the closing prices — the market’s last word before kick-off. The Move column is how each market’s implied chance has shifted since the first price we captured: green = the chance has risen (the price shortened, backing coming in), red = the chance has fallen (the price eased out — what punters call the price ‘drifting’).

Head to head