Every punter hears it eventually: "just back the favourite — they win the most." Half of that sentence is true. We settled every one of 18,925 UK and Irish races between 16 March 2025 and 12 July 2026 at official starting price, and this is what backing the SP favourite in all of them actually did:
Favourites won more than one race in three — comfortably the best strike rate of any blind strategy. And a £1 level stake on all 18,925 of them lost about £2,016. Both facts at once. That's the whole lesson of betting markets in one table.
The market isn't guessing. Thousands of pounds of opinion compress every horse's chance into its price, and favourites are — on average — priced slightly shorter than their true winning chance once the bookmaker's margin is added. Win 34% of the time at average odds that assume you'll win 38%, and the arithmetic quietly takes ten pence of every pound. Not through drama. Through margin.
This is the difference between the two questions punters constantly swap by accident: "who is most likely to win?" and "which price is wrong?" The favourite is usually the answer to the first. Profit only ever lives in the second.
This is the obvious next move — if favourites lose overall, surely there's a sweet spot? A price band where the crowd over-corrects and the favourite becomes good value? We split all 18,925 races by the favourite's starting price. There isn't one:
| Favourite's SP | Races | Won | Return |
|---|---|---|---|
| Odds-on (under evens) | 3,364 | 58.7% | −7.1% |
| Evens – 2/1 | 6,623 | 37.3% | −9.7% |
| 2/1 – 4/1 | 7,875 | 24.2% | −12.5% |
| 4/1 and bigger | 1,063 | 15.6% | −14.4% |
Every band loses — and it gets steadily worse as the favourite drifts. There is no secret price. The margin is baked into every number on the board, and a weak favourite in an open race is more mispriced against you, not less.
Any tipster, system or model claiming an edge should be measured against this baseline — not against zero. Losing 5p per £1 sounds bad until you know the default strategy loses 10.7p. The honest yardsticks:
1. Return against the favourite baseline over the same races — not
cherry-picked ones.
2. Closing line value (CLV) — do the prices you take consistently beat the
price at the off? If yes, you're seeing something before the market does; if no, short-term
profit is probably luck.
3. Losses left in. Any record that can be edited after the fact is
marketing, not measurement.
Every model pick logged pre-race with timestamped odds. Wins, losses, and the favourite baseline — side by side.
What percentage of favourites win?
34.4% across the 18,925 UK & Irish races we settled (March 2025 – July 2026) — about one in
three, the best strike rate of any blind strategy.
Is backing favourites profitable?
No — the same favourites returned −10.7% at SP (about £2,016 lost on £18,925 staked), because
the market prices them slightly shorter than their true chance.
Is there a price band where it works?
No. Odds-on favourites lost 7.1%, evens–2/1 lost 9.7%, 2/1–4/1 lost 12.5%, and 4/1+ lost 14.4%.
Every band is negative.
How should you judge a tipster or model?
Against this favourite baseline over the same races, on closing line value, and on whether the
record is public with losses left in.
"Favourite" = the runner with the lowest official starting price in each race (18,925 races with a priced favourite, 16 March 2025 – 12 July 2026, UK & IRE flat and jumps; joint-favourites resolved to a single runner). Returns are £1 level stakes at SP, win only, no each-way, no Rule 4 adjustments beyond official results. Non-runners excluded. Past performance does not predict future results.
Update, 13 July 2026. The first version of this study covered 3,185 races — our live forward-test window (April–July 2026) — and found 35.1% winners at −10.4%. It now runs on the full settled dataset back to March 2025: 18,925 races, 34.4% winners at −10.7%. Six times the sample, the same answer. We're leaving both figures on the record rather than quietly swapping one for the other — if you saw the older number somewhere, that's where it came from.
Are odds-on favourites safe? — they win
nearly six times in ten and still lose money.
Do long-shots pay? — 179,202 runners on whether the
big price is really where the value hides.