The Thursday Take is our weekly column: one racing talking point, put to the test against the Proform database — every British and Irish runner, every result. This week, a question we have wanted to answer properly for a long time: does the betting market underestimate women riders?
The short version
- The market prices women almost exactly right. Across 67,854 British rides since 2018, horses ridden by women won 6,456 races against the 6,360 their Betfair prices expected: an A/E of 1.015. Men came in at 0.999, and put on the same mix of prices they would have been 0.999 too. A/E is simply what actually won divided by what the prices said should win; 1.00 means the market got it right.
- The professionals are priced right. Women with a full licence run at 1.011, well inside the noise.
- The gap lives in the riders still claiming an allowance. Women claimers run at 1.069, nearly three standard errors above par and above it in eight of the nine years. Male claimers on the same prices: 1.000. It survives the cleanest control there is — the same horse, ridden by a man and by a woman.
- It is a lens, not a profit. Backing every female claimer at level stakes returned +0.5%. The market's margin eats almost exactly the edge.
- A 2022 academic finding holds on its own ground, then stops. Women on the British jumps ran at 1.080 in 2018–20, inside the study's years. Since 2021: 0.991.
- Across the Irish Sea, women take half the share of rides — 4.9% against 8.9% — and the market prices them right too. Whether the claimer gap crosses the water, the Irish sample is too small to say.
This week's talking point
September brought the first British Classic ever won by a woman rider, and with it the question that follows every milestone like that: are women riders still underrated? For punters the useful version of that question is narrower and harder. Not "are they as good", which is an argument, but "does the betting market price their mounts correctly", which is a measurement.
It has been measured before. In 2022 a team at the University of Liverpool — Cashmore, Coster, Forrest, McHale and Buraimo, Female jockeys: what are the odds? — used official records for British jump racing from 2001 to 2020 and found horses ridden by women won slightly more often than their odds implied, with the effect growing in the later years. That gives us something better than a hunch to test: a published result, with an end date, that we can check on the years that came after it.
Rider sex is not in any racing results feed, so we needed a list. The British Horseracing Authority's own rider list covers every rider with a ride in Britain since 2018; with a handful of overseas visitors confirmed by hand, every British ride since 2018 is classified. All figures are for groups. No rider is singled out, and we do not publish the list.
How to read the numbers. Every figure is measured against Betfair SP, which carries no bookmaker margin, so it is the fairest price to judge a group against. A/E (actual over expected) is winners divided by the winners the prices implied: 1.00 is par, 1.05 means 5% more winners than the market expected. The ± figure is one standard error — roughly, how far the number could wander by luck alone. A result inside two standard errors of 1.00 is not a finding.
The headline: priced about right
| Britain, 2018 to Sept 2026 | Rides | Winners | Strike rate | A/E | Level stakes |
|---|---|---|---|---|---|
| Ridden by women | 67,854 | 6,456 | 9.51% | 1.015 ±0.013 | −2.6% |
| Ridden by men | 694,579 | 80,242 | 11.55% | 0.999 ±0.004 | −3.2% |
Women's mounts win less often — 9.51% against 11.55% — and the market knows it. What they win relative to their prices is 1.015, about one standard error above par. Backing every one of them at level stakes lost 2.6% after commission; backing every man's mount lost 3.2%. That is the whole difference: a fraction of a point.
The obvious objection is that women ride a different sort of horse, and they do. Half their rides (49.7%) start at 16/1 or longer against 42.6% for men, and outsiders are where prices tend to be least accurate. So we priced it out: take the men's A/E in each code and price band, weight it by where the women's rides actually sit, and men on the women's exact mix of prices come in at 0.999. The mix does not explain the 1.015. It is small, but it is not an artefact of the outsiders.
A different programme, not just a different rider
The pooled figure hides the most important fact in the study. Women and men in British racing are, to a large extent, doing different jobs:
| Share of each group's rides | Full licence | Claiming an allowance | Amateur |
|---|---|---|---|
| Women | 39.8% | 30.7% | 29.5% |
| Men | 78.2% | 14.3% | 7.5% |
Six in ten women's rides come as apprentices or conditionals claiming a weight allowance, or as amateurs. For men it is about two in ten. Any honest comparison has to be made within each of those groups, and that is where the story actually is.
Where the difference lives: the claimers
| Britain, by licence | Women: rides | Women: A/E | Men on the same prices | Women: level stakes |
|---|---|---|---|---|
| Full licence | 27,015 | 1.011 ±0.019 | 0.998 | −6.3% |
| Claiming an allowance | 20,820 | 1.069 ±0.025 | 1.000 | +0.5% |
| Amateur | 20,019 | 0.971 ±0.023 | 1.003 | −0.8% |
The professionals are priced right. The amateurs run a touch below par on both codes. And the claimers are where the market has not caught up: women claiming an allowance win 7% more often than their prices say, nearly three standard errors above par, while male claimers on the same prices sit at 1.000.
It is not one lucky season. The women's claimer figure is above par in eight of the nine years from 2018 to 2026, the exception being 2024 (0.982). Split it in half and it holds in both: 1.053 in 2018–21 and 1.090 in 2022–26.
It is not perfectly tidy either, and we are not going to make it look tidy. By size of allowance it shows at 3lb (1.088) and at 7lb or more (1.130), but the 5lb claimers sit at 1.003, par. By price it is above par in every band from odds-on out to 33/1, then the 33/1–100/1 band comes in at 0.837. Real effects are usually a little ragged like this; invented ones are usually suspiciously neat.
The cleanest control there is: the same horse. 15,158 British Flat horses since 2018 were ridden by both a man and a woman. On those horses alone, claiming women ran at 1.064 and claiming men at 1.007; among riders with a full licence it was 0.998 against 0.990. Take the horse out of the question and the claimer difference is still there.
The plausible reading: the market prices a claim as a weight allowance — so many pounds off — and prices the rider off their record. A young woman's record is short and thinly watched, and the market seems to mark it down a little too hard. That is a story that fits the numbers, not something the numbers prove.
Why this is not a system. Backing every female claimer at level stakes since 2018 returned +0.5% after commission. The edge is real, repeated and roughly the size of the market's margin, so it cancels the margin and no more. It is a reason to look twice at a horse, not a reason to back it.
The 2022 study, tested on the years after it
| British jumps | Women: A/E | Men: A/E |
|---|---|---|
| 2018–2020 (inside the study's years) | 1.080 ±0.044 | 0.996 |
| 2021–2026 (after it) | 0.991 ±0.032 | 1.001 |
On its own ground and in its own years, our data agrees with the Liverpool study: women on the British jumps were underestimated, at 1.080. Since 2021 they have run exactly to their prices. The effect the paper saw growing did not keep growing; it went. On the Flat, which the study did not cover, the figure is small and steady: 1.019 before 2021, 1.010 since.
The paper also found women finishing better than their odds implied. Measured within price bands — so the outsiders women ride more of cannot distort it — women beat a point or two more of their rivals than men at the same price in most bands. Same direction, modest size.
None of that makes the study wrong. It was right about the years it measured. Markets learn, and this looks like one that did.
Why other comparisons find no difference at all
Most published comparisons of women and men riders measure against the bookmaker starting price. That price carries a margin, and the margin falls hardest on outsiders — which is exactly what women ride more of. On bookmaker SP our own rated Flat rides give an A/E of 0.862 for women and 0.862 for men: identical. On Betfair SP the same rides give 1.013 and 0.999. The difference is small enough that the bookmaker's margin hides it completely. If you want to measure a small edge, measure it against a price that does not have one built in.
Across the Irish Sea
Irish racing is a separate jurisdiction with its own riders, so it is a natural second test. The BHA's list covers anyone who has ridden in Britain; for riders who have only ridden in Ireland we built the classification ourselves, from the racecard title or a published source for every rider, with the harder cases settled by hand. To compare like with like, the table below puts both countries on exactly the same basis.
| 2018 to Sept 2026 | Britain | Ireland |
|---|---|---|
| Women's share of all rides | 8.9% | 4.9% |
| Women's rides that are amateur | 29% | 40% |
| Women's rides with a full licence | 40% | 26% |
| Women, all rides: A/E (men on the same prices) | 1.015 (0.999) | 0.987 ±0.029 (1.001) |
| Women claimers: A/E (men claimers on the same prices) | 1.070 (1.001) | 1.038 ±0.066 (1.022) |
| Women amateurs: A/E (men on the same prices) | 0.969 (0.997) | 1.000 (1.016) |
| Women on the jumps, 2018–20 | 1.080 | 1.009 |
Half the share. Women take 4.9% of Irish rides against 8.9% in Britain, and neither figure has moved much since 2018. The Irish women who do ride are even more concentrated in the amateur ranks: four in ten of their rides carry an amateur title, and only a quarter are professional rides. They are also home-grown — riders based mainly in Britain account for under 2% of Irish women's rides.
Priced right. Overall, Irish women run at 0.987 against 1.001 for men on the same prices, and the amateurs at par. Nothing there the market has missed.
The claimer gap: too small a sample to say. Irish women claimers run at 1.038, but so do the men on the same prices, near enough (1.022), so the gap is 0.016 against Britain's 0.069. Before reading that as "it does not cross the water", look at the precision. Ireland has 241 expected claimer winners in this period against Britain's 1,775, so the Irish figure is about two and a half times less precise — a gap the size of Britain's would be only about one standard error there. The honest answer is that the Irish data can neither confirm nor rule it out.
The study's jumps effect never appeared. The Liverpool paper was about British racing only. In Ireland, even inside its own years, women on the jumps ran at 1.009 — par. Whatever the British market was missing in 2018–20, the Irish one was not.
One caution on the Irish figures. The Irish women's professional ranks are small and dominated by very few riders — the busiest accounts for nearly a quarter of all Irish women's rides, twice the British equivalent — so the professional figure there describes individuals rather than a group, and we have not printed it. The overall and claimer figures above are broad enough to read.
It is not one rider
In Britain the picture is broad and shallow. Take out the five busiest women and the figure is 1.011. Of the 90 women with 100 or more British rides since 2018, 52 sit above par — more than half, and nowhere near all. This is the market being slightly off on a group, not one rider carrying it.
Build it yourself
As of this week, this is in the product. Jockey Sex is now a rule and a breakdown in the System Builder, built from the same list as this article.
- Jockey Sex sits on the Connections tab, Female or Male. It is known before the race, so it works in today's qualifiers as well as back-tests.
- Open the Grid and put Jockey Sex against Jockey Claim: that is the claimer table above, in one click, on any race type, course or period you choose.
- Put Jockey Sex against Course Nation for the Britain and Ireland comparison.
- Use the Holdout view on anything you build here. It splits a system into an early period and a recent one; the jumps result in this article is exactly the kind of thing it catches.
- In Data Export the Jockey Sex column is available to tick alongside the rider's claim, if you would rather rebuild these tables in your own spreadsheet and check our arithmetic.
The take
The market has learned to price women riders where it sees them most: in the professional ranks, it gets them right. The one place it still lags is the young rider claiming an allowance, where women win about 7% more often than their prices say, year after year, on the same horses as the men. That is real, and it is worth exactly the market's margin — a reason to look twice, not a reason to bet.
The 2022 study was right about the years it measured, and the market has since caught up with it. That is the useful lesson for anyone building angles: an edge with a publication date tends to have an expiry date too. The way to know which kind you are holding is to test it on the years after it was found.
Last week we asked whether backing the beaten favourite still works — it wins twice as often as an ordinary runner and returns dead par, bar one narrow exception. Next week, one of the big autumn handicaps.
Questions we get asked about this
Are women jockeys underestimated by the betting market?
Barely, and only in one group. Across 67,854 British rides since 2018, women's mounts returned an A/E of 1.015 against Betfair SP, against 0.999 for men on the same prices. Professionals are priced right (1.011). The difference is concentrated in women claiming an allowance, at 1.069.
Is it profitable to back horses ridden by women?
No. Backing every woman's mount at level stakes lost 2.6% after commission, against 3.2% for men. Even the strongest group, women claimers, returned +0.5% — the edge cancels the market's margin and no more.
Do female apprentice jockeys outperform their odds?
In Britain, yes, modestly: women claiming an allowance won 7% more often than their Betfair prices implied (A/E 1.069, about three standard errors above par), above par in eight of the nine years since 2018, and on the same horses as male claimers. In Ireland the sample is too small to confirm or rule out the same effect.
What about Ireland?
Women take 4.9% of Irish rides against 8.9% in Britain, ride a more amateur programme, and are priced right by the market overall (0.987). The jumps effect found by the 2022 Liverpool study for Britain did not appear in Ireland, even in the study's own years.
Why do some studies find no difference between male and female jockeys?
Because they measure against bookmaker starting price. Its margin falls hardest on outsiders, which women ride more of, and it swamps a small difference: on bookmaker SP our data gives 0.862 for both. On Betfair SP the same rides give 1.013 against 0.999.
