A finishing positiondoesn't tell you what happened.
Fourth place after being crowded at the first bend is not the same fourth place as a clear run — but the form line looks identical. We read the stewards' comment on every run and tell you which is which.
43,870
races on record
249,672
runs read from the stewards' comment
197,039
runs with bend-by-bend positions
69.4%
of results confirmed by two or more sources
What you are actually buying
A dataset you can check, not a black box
Every figure below is counted from the database when this page loads. Nothing here is a claim you have to take on trust — each one has a page behind it showing the working, including what is missing.
What happened, not just where it finished
Every stewards' comment parsed into trouble, start quality and running line, graded by severity. Fourth after being crowded at the first bend is a different fourth, and the form line alone cannot tell you which one you are looking at.
Where the dog actually was
Bend-by-bend positions and the first split, run by run. Whether a dog leads and fades or comes from the back is a fact about the dog, not an impression.
Checked against independent sources
Results are compared across the sources that publish them, and every race carries how many agreed. Where they disagree we say so, and where only one source exists — all Irish racing — we say that too.
Published, not locked away
The whole thing downloads as CSV under CC BY 4.0, rebuilt weekly, provenance column included. A dataset you cannot audit is a dataset you cannot trust, so ours leaves.
And the part most sites leave out: no strategy we have tested beats the market at the starting price. We measured every one of them on nine months of this data and published the numbers, including the ones that lose. This is analytics, not tips.
Building something with this data?
The free sample is ninety days, complete in every column. The full archive, live results and the rights to redistribute or train a model are licensed — API from £99 a month, bulk by agreement.
Data API & licensingHow it works
From tonight's card to a clear read — in three steps
We pull the form
Every runner at every UK & Irish track, every day: times, sectionals, traps, going, trainer form and market prices — collected automatically before the card is published.
The model rates the race
A LightGBM ensemble turns 23+ features per dog into a calibrated win probability — and shows which features moved it up or down.
You get the read, not a tip
Which dog the model rates, why, and where it disagrees with the market — plus the public record of how each system has actually performed. Test your own ideas in the Strategy Lab.
What you get
The part of the form that nobody else publishes
Everyone shows you where a dog finished. We show you what happened to it on the way round — and we are open about what we cannot tell you.
Trouble in running, on every run
Crowded at bend 1, forced to check, badly baulked, clear run — read from the stewards' comment on every run we hold and graded by severity. 18% of runs are compromised: the finishing position understates the dog.
Where the dog actually was
Bend-by-bend running positions on 162,135 runs, with the first split where it is published. Whether a dog leads and fades or comes from the back is a fact about the dog, not a guess.
Kennel style, not just strike rate
594 trainers with a page of their own, showing how that kennel's dogs run: how often they break quickly, how often they meet trouble, the line they favour.
Breeding that means something
810 sires with progeny strike rates by distance, from our own records. We also say plainly where breeding does and doesn't predict — measured, not asserted.
Rules & legislation tracker
Parliamentary petitions with live signature counts, GBGB and Irish regulator announcements. The sport is being legislated right now and nothing else follows it in one place.
Honest about the model
Win probabilities with a published, checkable record — and an open statement of what our own research found: the betting market is efficient. We publish the losing weeks too.
Season leaderboards
Who's winning this season
Live trainer and greyhound rankings built from open UK & Irish results data — updated every week, with each position's movement since the previous week.
Season 2026 · week of 14 Sept
Top trainers
View all| # | ± | Trainer | Runs | Wins | Win % |
|---|---|---|---|---|---|
| 1 | — | P Guilfoyle | 32 | 20 | 62.5% |
| 2 | ▲1 | Patrick Guilfoyle | 57 | 27 | 47.4% |
| 3 | ▼1 | R Peckham | 55 | 26 | 47.3% |
| 4 | — | Michael J O'Donovan | 43 | 19 | 44.2% |
| 5 | ▲1 | A W Sear | 23 | 9 | 39.1% |
| 6 | ▲5 | J L Morris | 61 | 23 | 37.7% |
| 7 | — | L Dowling | 69 | 26 | 37.7% |
| 8 | ▼3 | Liam Dowling | 105 | 39 | 37.1% |
| 9 | ▲7 | Brendan Matthews(NI) | 28 | 10 | 35.7% |
| 10 | — | Robert G. Gleeson | 28 | 10 | 35.7% |
Top dogs
View all| # | ± | Greyhound | Runs | Wins | Win % |
|---|---|---|---|---|---|
| 1 | — | LEATHEMS BINGO | 54 | 22 | 40.7% |
| 2 | — | COOLRUSS IZZY | 47 | 19 | 40.4% |
| 3 | — | MICKYS TRIBUTE | 23 | 18 | 78.3% |
| 4 | — | PIEMANS GOALIE | 38 | 18 | 47.4% |
| 5 | — | DEVON RUTH | 44 | 18 | 40.9% |
| 6 | — | MADDOX ROSE | 51 | 17 | 33.3% |
| 7 | — | BRUNCH PAL | 30 | 16 | 53.3% |
| 8 | ▲7 | BROWNS BARBIE | 45 | 16 | 35.6% |
| 9 | ▼1 | BOUNCING MONARCH | 21 | 15 | 71.4% |
| 9 | ▼1 | TIFFIELD TARQUIN | 21 | 15 | 71.4% |
Rankings are computed from public race results for research and information. Trainer figures attribute each dog to its current known trainer. Not affiliated with GBGB or any track.
Test any idea against 8 years of real prices
Build and backtest your own greyhound system across ~81 parameters on 8 years of real UK & Irish races — evaluated at real Starting Price, point-in-time, with look-ahead-free walk-forward. Start from the exact systems we run live, make them your own, publish them if you like.
- ~81 parameters: your filters, your allocation, your exit rules.
- Evaluated at real Starting Price — never simulated market prices.
- Walk-forward, embargoed, warmup — the model never sees the race it scores.
- Overfitting checks (Deflated Sharpe, PBO) that penalise fork-and-tweak honestly.
- We show negative-result systems too — that's the point: a transparent lab, not a tipster.
Illustration of the Strategy Lab UI — not a performance claim.
Still reading form on gut feel?
Swap guesswork for a calibrated read you can check against the record.
Without TrapStats
- A hunch from names, traps and colours
- One lucky week treated as a system
- Look-ahead bias — using data you wouldn't have had
- No record of what actually held up
With TrapStats
- 23+ features weighed for every runner
- Model probability set against the market price
- Point-in-time, walk-forward backtests
- A public paper track record — good months and bad
Pricing
Start free. Upgrade when the model has earned it.
Every new account starts with 7 days of Pro — no card, no auto-renewal. It ends by itself.
Free
Rankings, trap stats, results and 3 rated races a day
- 3 rated races a day, any track
- Weekly trainer & dog rankings
- Trap stats for every track
- Results, racecards & dog profiles
- Every race rated
- Reasons & model-vs-market
- Strategy Lab
- API access
Pro
Every race rated, the reason behind each pick, model-vs-market
- Every UK & Irish race rated, every day
- Win & place probabilities + reasons
- Model-vs-market price gaps
- Live paper-tracked systems
- Strategy Lab backtests
- Pattern & angle analytics
- API access
Sharp
Backtest your own systems on 8 years of real prices
- Everything in Pro
- Strategy Lab at real SP
- SHAP breakdowns per pick
- Pattern & angle analytics
- CSV export · priority support
- API access
API
The data in your own tools
- Everything in Sharp
- REST API: racecards, probabilities, results
- Bulk export for backtesting
- Adjustable rate limits & quota
- Dedicated support
Frequently Asked Questions
See tonight's rated races.
Free account, no card — your 7-day Pro trial starts the moment you sign up, and ends by itself with nothing to pay.
Create free account