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, with the horses-in-training sales filling yards all over Britain and Ireland, a piece of advice that has been doing the rounds: buy one out of England and keep backing it.
The short version
- The first run is fine. A British import's first run in an Irish Flat handicap returns A/E 1.13 at Betfair SP, though that is within the range of chance and one 205/1 winner is all of the profit.
- Keep backing them and it turns. The second run is 0.85, the third 0.77. Together, 0.81.
- The horses do not get worse. The prices get shorter. The strike rate holds at 4.7%, 4.3%, 4.3%. The chance the price gives them goes 4.1%, 5.0%, 5.6%.
- The first-run bump has nothing to do with Britain. Any Flat handicapper's first run for a new yard, on any route, returns 1.10 on 15,116 runs, above par in 8 years out of 9. It looks a touch long. It is the one clear result here.
- "It's the trainer" is mostly hindsight. Yards that looked brilliant at 1.64 on their recruits ran 1.06 over the next four years.
This week's talking point
The horses-in-training sales are when a lot of ordinary handicappers change hands, and a line from Racing TV's Fran Berry caught our eye. He said horses bought out of England "definitely have two races in them … Just back them and keep backing them." Mark Costello, editor of The Irish Field, picked it up in his newsletter, said his own gut told him the opposite, and said the data on recent imports in Irish handicaps was not easy to come by. He suspected the real answer was the trainer.
We hold that data, so here it is. His remark followed Serenity Blue, who won at Dundalk on 9 October on his first run for Ado McGuinness, at about 15/4. Three weeks later he went back there 7lb higher, at about 9/4, and finished fourth. One horse proves nothing, but keep that price in mind.
How to read the numbers. A/E (actual over expected) is winners divided by the winners the Betfair SPs said to expect. 1.00 means the prices were exactly right; 0.85 means 15% fewer winners than the prices implied. Every runner counts, fallers and pulled-up horses included. A British import here is a horse whose last yard was British-based and whose new yard is Irish-based, running in a Flat handicap in Ireland. British and Irish racing, 2018 to 2 November 2026.
Two races in them?
| British import, Irish Flat handicap | Runs | Wins | Strike rate | Price says | A/E |
|---|---|---|---|---|---|
| First run for the new yard | 941 | 44 | 4.7% | 4.1% | 1.13 |
| Second run | 886 | 38 | 4.3% | 5.0% | 0.85 |
| Third run | 857 | 37 | 4.3% | 5.6% | 0.77 |
| "Two races": first and second together | 1,827 | 82 | 4.5% | 4.6% | 0.98 |
| "Keep backing": second and third together | 1,743 | 75 | 4.3% | 5.3% | 0.81 |
Read across the strike rate and the price. The horses win about as often on their third run as on their first. What changes is what you are paid for it. So Berry's point about the horses may well be fair, and it is still not a reason to keep backing them, because the market has noticed too.
Be careful with the good news. The first run shows a +10.8% return at Betfair SP, but that is one horse, Miss Nifty, at about 205/1 at Naas in June 2025. Take her out and it is about −10.6%. Winning first time gives no licence to keep going either: a second run after a win is 0.88 (44 runs), much the same as after anything else. And 41% of these horses start at 50/1 or bigger. Most are outsiders and they run like it: an average runner in an Irish Flat handicap wins 7.7% of the time, an import on its first run 4.7%, which gives Costello's gut about Irish competitiveness some support.
The real test: is it the border, or the move?
A first run for a new yard might just be a first run for a new yard. So we ran the same test on every route.
| Flat handicap, last yard to new yard | Runs, first run | First run | Second | Third |
|---|---|---|---|---|
| Britain to Ireland | 941 | 1.13 | 0.85 | 0.77 |
| Ireland to Ireland | 2,285 | 1.11 | 0.92 | 0.94 |
| Ireland to Britain | 1,314 | 1.13 | 1.00 | 1.08 |
| Britain to Britain | 10,481 | 1.08 | 1.04 | 1.03 |
| Every route, all Flat handicaps | 15,116 | 1.10 | 1.01 | 1.02 |
That is the finding. The bump is not about crossing the Irish Sea. Move a Flat handicapper to any new yard, Britain to Britain included, and its first run returns about 1.10. By year it runs 1.06, 1.05, 1.15, 1.02, 1.32, 1.11, 1.12, 0.97, 1.10, so above par in 8 years out of 9. It holds in every class from 2 to 6, on turf (1.11) and on the all-weather (1.08), whether the horse ran well last time or not. And it sits in a particular part of the market:
| First run for a new yard, Flat handicap, by Betfair SP | A/E |
|---|---|
| Under 3/1 | 1.00 |
| 3/1 to 6/1 | 1.03 |
| 6/1 to 10/1 | 1.17 |
| 10/1 to 20/1 | 1.12 |
| 20/1 to 50/1 | 1.30 |
| 50/1 or bigger | 0.87 |
The favourites are priced right and the rags are not worth it. The gap is in the middle, the horses nobody is quite sure about. What Ireland adds is the second half of the story. In Britain, the second and third runs hold up (1.04 and 1.03 on the home route). On Irish courses, recruits from anywhere fade: 0.90 and 0.89.
How sure are we? The 1.10 is the only thing here clear of noise, at about three standard errors. It was the control, not something we went hunting for, but it is still one cut among many, and it needs saying that 2022 (1.32) flatters it: without that year it is about 1.07, and 2025 was under par at 0.97. It looks a touch long; do not read it as a system that makes money. The import fade is weaker again: first run against later runs is a gap of about 1.7 standard errors, and the wider Irish fade (0.90) about two. Suggestive, not proven.
Doesn't that contradict what we said on Tuesday?
On Tuesday we looked at horses leaving big jumps yards and said a first run for a new yard was no bonus: 0.86 on 1,617 runs. Different horses, different result. Those were jumps horses, in all kinds of race, leaving the 15 yards with most leavers. This is the Flat, in handicaps, from every yard. Run this week's rule over jumps and the figure is smaller: 1.03 in handicaps (7,642 runs), inside what chance covers, and 1.05 across every jumps race (16,873 runs), about two standard errors. The Flat handicap is where the first-run effect shows clearest, and nothing here changes Tuesday's point that leaving a big jumps yard earns a horse no bonus.
"The answer might be the trainer"
Costello's suspicion is the one everyone has, and it is easy to test properly. We took every yard with 30 or more first runs for new recruits in 2018 to 2022, in any race on any route, sorted them on that record, then looked at what the same yards did from 2023 on.
| Yards' record on recruits, 2018 to 2022 | Yards | A/E then | Runs since | A/E since |
|---|---|---|---|---|
| 1.30 or better | 57 | 1.64 | 1,662 | 1.06 |
| 1.15 to 1.30 | 24 | 1.22 | 909 | 1.16 |
| 0.85 to 1.15 | 42 | 1.02 | 2,054 | 1.03 |
| Under 0.85 | 53 | 0.66 | 1,927 | 0.95 |
The list that looked like geniuses at 1.64 ran 1.06. There is something in it: the top two groups together ran 1.10 since, against 1.00 for the rest. So he has a point, but a smaller one than the records suggest, because most of a brilliant record is luck that does not repeat. The ten yards he named as his own shortlist fit: 1.10 up to 2022, 1.06 since, when all recruits were running about 1.04. A good judge's list, picked with hindsight, adds a little. That is no slight on any of those trainers. It is how hard this is to do from the outside.
The Brexit line in his piece does stand up. British-trained runners in Irish handicaps fell from 247 in 2018 to about 100 a year now, while Irish-trained runners in British handicaps went from 734 to 825. The traffic is largely one way. And British strike rates in Ireland bounce around too much to read: 15% in 2022, 9.6% in 2025, and a thin 3 winners from 89 runs so far this year.
Build it yourself
All of this runs in the System Builder with a handful of rules.
- Race Type is one of Turf, A/W, and Handicap = Yes.
- First Run for New Trainer = Yes. Swap it for Second Run for New Trainer or Third Run for New Trainer to see the fall.
- For the Irish version, Course Country is one of (IRE), (NI).
- For imports only, add Trainer Nation = Ireland and LTO Course Country = (GB). That works on the first run only: the Builder cannot see where the yard was two runs back.
- Then open the Breakdowns tab for Year and Trainer, and remember the table above before you fall in love with a trainer's row.
| System Builder, 2018 to 2 November 2026 | Runs | Wins | A/E | Return |
|---|---|---|---|---|
| All Flat handicaps, first run for a new trainer | 15,116 | 1,098 | 1.10 | +6.0% |
| Second run | 14,687 | 1,169 | 1.01 | −10.4% |
| Third run | 13,848 | 1,210 | 1.02 | +2.0% |
| Irish courses, first run | 3,246 | 176 | 1.13 | +3.9% |
| Irish courses, second run | 3,072 | 159 | 0.90 | −23.0% |
| Irish courses, third run | 2,842 | 158 | 0.89 | −16.0% |
| Imports only, first run | 950 | 43 | 1.10 | +9.1% |
The import line here (950 runs) differs slightly from our table above (941) because the Builder reads the course of the last run, where we used the last yard. The returns swing a long way by year, from about −13% to +36% on the top line, so read A/E first.
The take
Half of it is right. A horse going to a new yard, in a Flat handicap, does seem to run a little better than its price on the first go, and that is worth knowing as the sale horses start to appear. But it is not a British thing, it is not a licence to keep going, and it is not much of a trainer thing. Keep backing them in Ireland and you are paying more each time for a horse winning no more often. They may well have two races in them. The price only leaves you one.
Last week we looked at why "won this last year" costs you in a jumps handicap. Next week: another talking point, another test.
Questions we get asked about this
Do horses improve when they move from Britain to Ireland?
Their strike rate does not rise from run to run: 4.7%, 4.3% and 4.3% over their first three runs in Irish Flat handicaps (2018 to November 2026). Their first run beat its Betfair price (A/E 1.13, 941 runs), but later runs fell short because the prices shortened (0.85, then 0.77).
Is a first run for a new trainer a good bet?
In Flat handicaps it has looked a touch long: A/E 1.10 on 15,116 runs, above par in 8 of 9 years, mostly between 6/1 and 50/1. Without 2022 it is about 1.07. Over jumps the figure is closer to fair: 1.03 in handicaps.
Which trainers are best with new recruits?
Less can be read from a trainer's record than it seems. Yards with an A/E of 1.64 on recruits from 2018 to 2022 ran 1.06 from 2023 on. There is a small trainer effect, but most of a very good record does not repeat.
Have fewer British horses run in Irish handicaps since Brexit?
Yes. British-trained runners in Irish handicaps fell from 247 in 2018 to about 100 a year now, while Irish-trained runners in British handicaps rose from 734 in 2018 to 825 so far in 2026.
Can I test this in the System Builder?
Yes. Race Type (Turf, A/W), Handicap = Yes and First Run for New Trainer = Yes gives A/E 1.10 on 15,116 runs since 2018. Add Course Country (IRE), (NI) for the Irish version, and swap in Second or Third Run for New Trainer to see the fall.
