· Independent research
Closing-line value in football betting
A test of whether one bookmaker’s opening odds can be beaten, measured against a sharper bookmaker’s prices. It found no edge.
- Period
- June to September 2026
- Data
- football-data.co.uk: eight leagues, 2019/20 to 2024/25, and 14 further leagues over the same seasons for the final test
- Status
- Closed
Closing-line value compares the odds a bet was placed at with the odds just before the match starts. Pinnacle’s closing odds are a common benchmark for it. The question was whether Bet365’s opening odds are sometimes better than Pinnacle’s view of the fair price, by enough to matter.
The data covered 17,292 matches in eight club leagues, from the 2019/20 to the 2024/25 season, taken from a public odds dataset. A rule that backed Bet365’s opening odds whenever they were at least 1% above Pinnacle’s fair opening price picked 1,527 bets.
Against Pinnacle’s closing odds as quoted, those bets showed +4.55% closing-line value. But quoted odds still include the bookmaker’s margin. With the margin removed, by three standard methods, the figure is between −2.87% and +1.35%, a range that includes zero. On its own, then, the +4.55% figure did not show an edge.
Against Pinnacle’s closing odds, as quoted
With the margin removed, range across three methods
A pre-registered test followed, on 14 leagues the rule had never seen, over the same six seasons. The rule and the way of removing the margin were fixed in advance, and the test was run once. It found no edge: the 3,716 bets the rule picked won slightly less often than the margin-free closing odds implied (Z = −0.46), so I stopped there.
The lesson was mostly about the benchmark: whether the same bets beat the closing odds depended on whether, and how, the margin was taken out of them.