The price of a bet

Like for like, a bet costs more than it did when the record began: the same field sizes carry a bigger margin today. Fields are smaller than they were, and smaller fields carry less margin, so part of the change in the year-by-year line is the fields, not the book. Across 268,233 races with a full set of starting prices, the mean overround is 118.5 %: 18.5 points of margin in the average race, in fields averaging 9.9 runners. By year the price of a bet ran from 116.7 % in 2014 to 122.4 % in 2005; 2026 so far reads 119.1 % against 122.4 % in 2005. Yesterday's 23 priced races carried a mean overround of 118.3 % — 0.8 points below the last year's 119.1 %, and 0.9 points above what the same field sizes have cost over the year (117.4 %). Every GB and Irish race from 2 January 2005 to 25 October 2026.

The bigger the field, the more a bet costs: the biggest fields carry more margin than the smallest. The favourite wins less often in the biggest fields than in the smallest. Against its price, it does better in the biggest fields than in the smallest: the market marks the big-field favourite down by more than the extra runners cost it. The overround rises with the field: 109.4 % in fields of 5 or fewer, 144.3 % in fields of 20 or more. The favourite by field size: 5 – 7 runners 41.0 % won, 60.2 % placed, fair A/E 1.04 (N 85,502); 8 – 11 runners 31.5 % won, 64.2 % placed, fair A/E 1.09 (N 116,622); 12 – 15 runners 27.0 % won, 56.2 % placed, fair A/E 1.13 (N 67,654); 16 or more 24.1 % won, 55.0 % placed, fair A/E 1.22 (N 23,485). Every GB and Irish race from 2 January 2005 to 25 October 2026.

In every field size the favourite does better against its price than the outsiders do: the favourite–longshot bias runs through the whole table. Against the price, the best cell in the table is the favourite in fields of sixteen or more; the worst, the sixth favourite or longer in fields of sixteen or more. The best fair A/E in the table is the favourite in fields of 16 or more: 24.1 % won and 55.0 % placed at fair A/E 1.22 (N 23,485) — the favourite–longshot bias at its strongest, and the place part is where it shows. The full table is on the page. Every GB and Irish race from 2 January 2005 to 25 October 2026.

A clear favourite wins more often than one at the head of a tight market. Measured against the price, there is little between them: the market sizes the gap about right. A clear favourite (twice the second favourite's fair chance or more) won 51.0 % and placed 76.2 % at fair A/E 1.10 (N 57,257); a favourite 1.4 – 2 times the second won 34.8 % and placed 63.2 % at fair A/E 1.08 (N 68,529); a tight market (under 1.4 times) won 26.8 % and placed 55.5 % at fair A/E 1.08 (N 120,730); a joint favourite won 21.8 % and placed 49.9 % at fair A/E 1.07 (N 45,642). Every GB and Irish race from 2 January 2005 to 25 October 2026.

On the flat, the favourite wins about as often in Ireland as in Britain, and against the price there is little between them. Over jumps, the favourite wins about as often in Ireland as in Britain, and against its price the Irish favourite does better. The favourite in Britain, flat 31.7 % at fair A/E 1.08 (N 147,592); the favourite in Britain, jumps 34.7 % at fair A/E 1.07 (N 83,886); the favourite in Ireland, flat 30.1 % at fair A/E 1.11 (N 28,480); the favourite in Ireland, jumps 33.4 % at fair A/E 1.16 (N 33,305). Every GB and Irish race from 2 January 2005 to 25 October 2026.

On the flat, the favourite wins most often in winter and least often in the autumn changeover; by going, it wins most often on firm and least often on heavy. Over jumps, the favourite wins most often in winter and least often in summer; by going, it wins most often on heavy and least often on good to soft. The flat favourite wins most often in winter (16 Nov – 14 Mar) (32.0 %) and least often in the autumn changeover (15 Oct – 15 Nov) (29.7 %); the jumps favourite wins most often in winter (16 Nov – 14 Mar) (35.6 %) and least often in summer (May – Sep) (33.0 %). Every season and every going is on the page, by code. Every GB and Irish race from 2 January 2005 to 25 October 2026.

On the flat, the favourite does best against its price in November and worst in April. Over jumps, the favourite does best against its price in November and worst in March. The flat favourite runs furthest above its price in November (fair A/E 1.10) and nearest to it in April (1.05); the jumps favourite runs furthest above its price in November (fair A/E 1.12) and nearest to it in March (1.07). Every GB and Irish race from 2 January 2005 to 25 October 2026.

On the flat, odds-on runners win more often than even their short price says. Over jumps, odds-on runners win more often than even their short price says. Odds-on (fair) runners on the flat won 63.5 % at fair A/E 1.08 (N 12,103); odds-on (fair) runners over jumps won 65.1 % at fair A/E 1.09 (N 12,394) — against the fair line of 1.00. Every GB and Irish race from 2 January 2005 to 25 October 2026.

Convert every runner's starting price in a race to an implied chance and add them up. A fair book would total exactly 100 %; a real one totals more, and the excess is the bookmakers' margin — the price of the bet. We compute it per race from the stored starting prices, only for races where every runner has a usable SP, and the figures here are the mean over those races in the window named beside them.

Fair A/E is winners divided by the winning chances implied by the starting prices once the bookmakers' margin is removed. A row above 1.00 won more often than its price said; below 1.00, less often. Favourites tend to sit above 1.00 and long shots below it (the favourite–longshot bias), so compare rows with rows, not with 1.00 alone.

Nightly. Every figure is summed from a store rebuilt from our archive after each day's results, so each carries the date of the last race it includes; the daily line is yesterday's races against the year and against the same field sizes.

By the figures, not by a writer. Each section's headline compares two of its own numbers — the price of a bet now against when the record began, the biggest fields against the smallest, the clear favourite against the tight market, Ireland against Britain — and takes the sentence that matches the result. A difference counts when it is at least 1.0 point of overround, 2 points of strike rate, 0.04 of fair A/E or 0.5 of a runner in mean field size; anything smaller reads as about the same. The numbers are rebuilt nightly, so a headline that stops being true changes its words.