Observed market data
How Much Did Price Shopping Reduce 1X2 Overround?
Using the same 10 matches from the 2025/26 Premier League opening weekend, we compared the source file's average opening prices with the maximum listed opening price for each outcome. The question is narrow: how much did the combined implied probability change when stronger prices were selected outcome by outcome?
What “price shopping” means in this study
Football-Data.co.uk records both average and maximum bookmaker odds. For each match, we calculated one book from the average home, draw and away prices, then a second book from the maximum home, draw and away prices.
Book percentage = (1 ÷ home odds + 1 ÷ draw odds + 1 ÷ away odds) × 100
Reduction = average-price book − best-available-price book
The maximum prices can come from different bookmakers. The best-available book is therefore not a bookmaker's own market. It is a cross-bookmaker benchmark showing the arithmetic effect of selecting the strongest listed price separately for each outcome.
Match-by-match results
| Match | Average-price book | Best-available book | Reduction |
|---|---|---|---|
| Liverpool v Bournemouth | 105.15% | 100.54% | 4.61 pp |
| Aston Villa v Newcastle | 105.58% | 100.94% | 4.64 pp |
| Brighton v Fulham | 105.28% | 102.95% | 2.33 pp |
| Sunderland v West Ham | 105.38% | 102.82% | 2.56 pp |
| Tottenham v Burnley | 105.87% | 101.95% | 3.92 pp |
| Wolves v Man City | 105.44% | 100.90% | 4.54 pp |
| Chelsea v Crystal Palace | 105.55% | 102.15% | 3.40 pp |
| Nott'm Forest v Brentford | 105.43% | 102.19% | 3.24 pp |
| Man United v Arsenal | 105.85% | 101.61% | 4.24 pp |
| Leeds v Everton | 106.02% | 101.36% | 4.66 pp |
Every match produced a lower book when maximum rather than average prices were combined. The reduction ranged from 2.33 percentage points on Brighton v Fulham to 4.66 points on Leeds v Everton.
Summary statistics
Another way to express the same result is in overround. The average-price set contained about 5.56 percentage points above 100%; the cross-bookmaker best-price set contained about 1.74 points. This does not translate directly into a guaranteed bettor return or bookmaker profit figure.
Where did the reduction come from?
We also measured the reduction in implied probability produced by replacing the average price with the maximum price separately for home, draw and away outcomes. Averaged across the 10 matches, the contribution was approximately 1.32 percentage points from home prices, 1.13 from draw prices and 1.36 from away prices.
Those figures sum to roughly the 3.81-point average reduction in the total book. In this small sample, the effect was therefore spread across all three outcome types rather than being confined to one side of the 1X2 market.
Worked example: Aston Villa v Newcastle
The average opening prices were 2.30, 3.56 and 2.94. They produced a 105.58% book. The maximum listed prices were 2.38, 3.75 and 3.10, which produced a 100.94% book.
Average-price book = 105.58%
Best-available book = 100.94%
Reduction = 4.64 percentage points
This is one of the largest reductions in the sample, but it should not be read as a claim that those maximum prices were simultaneously available indefinitely or at unlimited stakes.
What the result does — and does not — show
- It shows that the listed cross-bookmaker maximum prices materially reduced combined implied probability in every match in this sample.
- It does not show that one bookmaker offered a 101.74% average book.
- It does not prove that the same reduction exists in other leagues, market types or time periods.
- It does not account for account restrictions, stake limits, price changes or practical execution.
- It does not identify profitable bets. A smaller book and a correct probability estimate are different questions.
Data and reproducibility
The source is the Football-Data.co.uk 2025/26 English Premier League CSV (division E0). Betting Maths calculated the implied probabilities and book percentages. The same derived dataset used for our opening-weekend and opening-versus-closing studies contains every input used here.
Why this matters mathematically
Overround is often discussed as though a market has one fixed margin. In practice, the book percentage depends on the exact prices being combined. This sample provides a concrete demonstration: using different recorded prices for the same outcomes changed the arithmetic substantially without changing the matches themselves.
