Stop guessing.Start modelling.
A community of greyhound enthusiasts building, testing and sharing analytical systems. Calibrated win probabilities, model signals and plain-English reasons for every UK & Irish race — open data and AI models, never advice or promises.
30+
UK & Irish Tracks
500+
Daily Predictions
23+
Features / runner
7-day
Free trial
How it works
From data to a clear read in 3 steps
We Collect the Data
Our scrapers pull live form, sectionals, trap stats, going conditions, and trainer data from every UK and Irish track — automatically, every day.
AI Analyses Every Runner
LightGBM models process 23+ features per dog — form, track bias, sectional positions, weight trends — to produce calibrated win probabilities.
You Get the Analysis
See exactly which dogs the model favours, how much each system allocates (Kelly allocation method), and why (model feature importance & per-pick reasoning). Built-in backtesting lets you test systems on historical data.
Features
Everything you need to analyse a race
Professional-grade analytics powered by machine learning — professional analytics, all in one place
Win Predictions
LightGBM models trained on thousands of races to predict calibrated win probabilities for every runner.
Risk-Weighted Allocation
Mathematically grounded allocation modelling (Kelly method) computed from model signal, test portfolio and risk tolerance.
Model transparency
See which features drive the model globally, plus per-pick reasoning explaining why each selection was made — no black boxes.
Backtesting Engine
Walk-forward backtests on real historical data with Betfair BSP market prices to validate any system.
Real-Time Data
Live feeds from GBGB and GRI with Betfair BSP integration. Fresh data every single race day.
REST API
Full programmatic access for automated research and simulation. Build your own tools on top of ours.
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 27 Jul
Top trainers
View all| # | ± | Trainer | Runs | Wins | Win % |
|---|---|---|---|---|---|
| 1 | — | R Peckham | 32 | 14 | 43.8% |
| 2 | ▲1 | K M Grayson | 61 | 24 | 39.3% |
| 3 | ▼1 | A W Dean | 25 | 9 | 36.0% |
| 4 | ▲1 | J L Morris | 37 | 13 | 35.1% |
| 5 | NEW | B P Cooper | 20 | 7 | 35.0% |
| 6 | ▼2 | M A Burton | 64 | 22 | 34.4% |
| 7 | ▲1 | B Fairbairn | 100 | 34 | 34.0% |
| 8 | ▼2 | G D Holland | 83 | 28 | 33.7% |
| 9 | ▼2 | K J Ferguson | 54 | 18 | 33.3% |
| 10 | — | E Soppitt | 82 | 25 | 30.5% |
Top dogs
View all| # | ± | Greyhound | Runs | Wins | Win % |
|---|---|---|---|---|---|
| 1 | — | COOLRUSS IZZY | 22 | 13 | 59.1% |
| 2 | — | MICKYS TRIBUTE | 15 | 12 | 80.0% |
| 3 | — | LEATHEMS BINGO | 28 | 11 | 39.3% |
| 4 | ▲4 | SHAKE THE BOTTLE | 14 | 10 | 71.4% |
| 5 | ▲6 | BRUNCH PAL | 17 | 10 | 58.8% |
| 5 | ▼1 | FABULOUS SHOW | 17 | 10 | 58.8% |
| 7 | ▼2 | REENADERRY ROAD | 20 | 10 | 50.0% |
| 8 | ▲5 | BAGGINS LAD | 21 | 10 | 47.6% |
| 9 | ▼3 | PIEMANS GOALIE | 24 | 10 | 41.7% |
| 10 | ▲9 | SPADE WORK | 32 | 10 | 31.3% |
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 analytical 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, with point-in-time data and look-ahead-free walk-forward. Start from the exact systems our platform runs live, then make them your own and share them with the community.
- ~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 analysing on gut feeling?
Swap guesswork for rigorous, reproducible research.
Without TrapStats
- Guessing based on names and colours
- Overfitting to a handful of past races
- Look-ahead bias — using data you wouldn't have had
- Zero record of what actually generalises
With TrapStats
- 23+ features analysed per runner
- Model signal measured against the market price
- Point-in-time, walk-forward backtesting
- Share and fork systems with the community
Pricing
Simple, honest pricing
No hidden fees. Cancel anytime. Start free.
Free
Get started with basic predictions
- 3 races per day
- 1 track (Romford)
- Basic race data
- Community support
- Kelly allocation
- Pattern & profitable-angle analytics
- Backtesting
- API access
Pro
For analysts and researchers who want full model-driven analysis
- Unlimited races
- All UK & Irish tracks
- Win predictions + Kelly allocation
- Email support
- Pattern & profitable-angle analytics
- Backtesting
- API access
Sharp
Advanced analytics for quant hobbyists and researchers
- Everything in Pro
- Advanced backtesting (ML strategies)
- Pattern & profitable-angle analytics
- Backtest CSV export
- Priority support
- API access
API
Full API access for automated systems
- Everything in Sharp
- REST API access
- Bulk export for backtesting
- Adjustable rate limits & quota
- Dedicated support
Frequently Asked Questions
Ready to stop guessing?
Create a free account and explore the predictions, analytics and tools. Start with a free account — no credit card required.
Start Free Today