Do Steamers Actually Win? Horse Racing Labs
The team at Racing-Odds.com have analysed over four million price movements. The aim is to find out whether it's worth backing "steamers" on a regular basis. A steamer is regarded as a horse whose odds significantly shorten in the build-up to the race. There are a number of reasons why this happens including market confidence, bookmaker liability and whispers in the yard. On a daily basis, thousands of horse racing odds change across the bookmakers we track. Some of these moves are minimal. Others can be significant. The trick is to often identify when it's the money that's talking. We wanted to know that when a horse becomes a steamer, how often does it actually win?
Read more
Our quest led us to analyse every price movement recorded across UK and Irish racing since mid-March. That meant 4.2 million price points, 4,135 races and 13 bookmakers. We matched each runner's market move to where it finished. The short answer is that the market is often a big pointer. The longer answer is more interesting.
The headline numbers
We measured each runner's move from its opening show to its final pre-race price, averaged across bookmakers, and grouped the 8,630 qualifying runners by how far they moved. Then we counted the winners.
| Market move | Runners | Won | Win rate the opening odds implied | Win rate | Win rate the final odds implied |
|---|---|---|---|---|---|
| Steamers (shortened 20%+) | 1,459 | 350 | 12.9% | 24.0% | 20.8% |
| Mild steamers (10-20%) | 694 | 140 | 15.5% | 20.2% | 18.8% |
| Stable (under 10% either way) | 1,559 | 212 | 14.7% | 13.6% | 15.2% |
| Mild drifters (10-20%) | 657 | 55 | 13.4% | 8.4% | 12.1% |
| Drifters (lengthened 20%+) | 4,261 | 146 | 11.7% | 3.4% | 7.5% |
Read the last two columns together and the pattern is unmistakable. Horses whose odds held steady won almost exactly as often as the bookmakers' margin says they should. Horses the market backed won more than even their shortened final price implied. Horses the market abandoned won less than half as often as their drifting price suggested.
Steamers beat the market twice
The 1,459 horses that shortened by 20% or more won 24% of their races. Their opening prices had implied a win rate of just 12.9% - the early market underestimated these horses by nearly half. That is the first and most obvious finding: a strong steamer is carrying genuine information, whether that is stable confidence, a well-followed tipping line or professional money.
The second finding is the one that surprised us. Even at their final, much shorter price, steamers still outperformed: 24% wins against the 20.8% their closing odds implied. In a market where the built-in margin normally grinds every group a few points below its implied rate - our "stable" control group won 13.6% against an implied 15.2%, exactly that drag - steamers overcame the margin entirely. The money kept being right after the move had already happened.
One honest caveat: you cannot bet on this in advance, because you only know a horse was a steamer once it has finished steaming. Part of the effect is also the well-documented tendency of shorter-priced horses to be better value than longshots. This is market intelligence, not a system. But as a signal to weigh alongside the form book, it is the strongest one we measured.
The strongest signal points the other way
Drifters are the real story in this data. The 4,261 horses that lengthened by 20% or more won just 3.4% of their races - against the 7.5% their weakened final price still implied. A horse that is friendless in the market does not merely win less often; it wins less than half as often as its drifting odds would have you believe. If the steamer finding says "the money knows", the drifter finding says it twice as loudly.
Newmarket's most-backed horses this season
The July Festival makes this a good moment to look at Newmarket specifically. Across 88 races at headquarters since mid-March we logged 67,375 real price movements. Of the fifteen most-backed horses at the track this season, five won, which was a 33% strike rate, right in line with the season-wide steamer pattern.
Booziebrunch was the most backed of all, crashing from around 5/1 into 13/8 for a 26th June handicap - and won. Nuptown Girl (10/1 into 9/2) and Powdering (14/1 into 13/2) landed handicaps the same evening. Musical Times was backed from around 9/4 into 5/6 for a fillies' novice and delivered. Dark Tornado was still 11/1 at the off after opening near 19/1 - the weakest-looking steamer of the five - and won at that price.
The gambles that landed, and one that didn't
Season-wide, the biggest single gamble we recorded that came off was Bacio at Royal Ascot: backed from around 14/1 into 11/4 for the Palace of Holyroodhouse Handicap on June 19, and home in front. Windsor's 22nd June evening card produced two in one night - Just Jump and Beelzebub both collapsed from around 13/1 into around 3/1, and both won. But the two biggest steamers of the entire season are the cautionary tale: Documenting, backed from around 33/1 into a 100/30 starting price at Newbury on 23rd June, led into the final furlong and was beaten three quarters of a length into second; and Poppieholla, backed from around 19/1 into 5/2 at Worcester on June 11, finished eighth. The money is right more often than the odds imply. It is not right every time.
What to watch at the July Festival
Over the three days of the July Festival, the pattern above is worth keeping on your screen. The horses to note are not necessarily the favourites but the movers: anything cut by a fifth or more of its price between the opening show and the off has, this season, won at nearly twice the rate its early odds implied. The drifters - however tempting the bigger price looks - have been the most reliable losers in racing. We track every move across 13 bookmakers live on our market movers page, updated through the day.
How we did it
We recorded every odds change across 13 bookmakers for every UK and Ireland race from March 16 to July 2, 2026 - 4.2 million price points over 4,135 races. For each runner we took its opening quote and final pre-race price at each bookmaker, averaged the move across bookmakers, and included only runners priced by at least four books that came under starter's orders (8,630 runners). A steamer is a runner whose average price shortened by 20% or more; a drifter lengthened by 20% or more. Win rates are compared against the win rate implied by the group's average final odds. Implied rates are not margin-adjusted, which is why the stable control group sits slightly below its implied figure - that gap is the bookmakers' margin, and it makes the steamer outperformance more notable, not less. Prices are our recorded bookmaker averages and may differ slightly from official SPs.