Collect race data
Runner history, track and distance records, class, conditions, prices and structured race information.
Beyond form. Beyond instinct. Driven by trained data.
Horse racing is not decided by one statistic or one past performance. Every start, result, price movement and racing condition contributes to a larger pattern. RACEMODEL continuously evaluates those patterns to provide disciplined, probability-based decision support.
17Racing signals
Full fieldEvery runner compared
15 minMarket refresh cycle
Place-firstSelective decision model
How it works
Four layers turn raw racing data into a clearer race view.
RACEMODEL combines historical performance, model probability, core quality and market behaviour before comparing the entire field.
Score runner quality
Model probability and core strength are calculated while market-derived risk remains visible.
Monitor movement
Steam, flat movement and drift show whether the market confirms or weakens the model view.
Compare full field
Every runner is reviewed in race context to identify candidates, dangers and weak-core traps.
Continuous improvement
Every race adds evidence. Every result improves the next decision.
RACEMODEL is built as a continuous validation cycle. Historical data, live market behaviour and completed race outcomes are reviewed together so the system can be tested, refined and strengthened over time.
Collect every race
Capture runner history, conditions, prices, market movement and final outcomes.
Train and score
Calculate model probability, core quality and structured racing signals.
Compare the full field
Evaluate every runner in race context instead of relying on one isolated statistic.
Monitor live confirmation
Track steam, flat movement and drift as the market changes before the race.
Learn from the outcome
Use completed results for validation, reporting and future model improvements.
The process is continuous: every new race creates more evidence for testing, validation and future refinement.
Key findings
Probability becomes more useful when quality confirms it.
Current validation shows meaningful separation when high model probability is combined with stronger core quality.
Historical validation continues to change as more races are added and does not guarantee future outcomes.
80%+
Runner with 75% model probability shows 80%+ placement rate
High probability provides a strong first filter for placement profiles.
99%
Runner with 75% model probability + Core score 5+ shows 99% placement rate
Core quality adds an important confirmation layer to the probability signal.
Race Explorer
See the whole race, not one isolated runner.
Compare model probability, core score, odds, market movement and status across the full field.
Upcoming racesDesktop preview
| Track | Race | Name | Last 5 | Model Prob % | Core | Win Odds | Place Odds | Steam % | Status |
|---|---|---|---|---|---|---|---|---|---|
| Belmont | R3 | Golden Legacy | 2-1-4-2-1 | 82% | 6.4 | $4.20 | $1.62 | +18.6% | Model Top 2 |
| Belmont | R3 | Midnight Runner | 3-5-2-1-4 | 68% | 5.1 | $5.80 | $1.95 | +2.4% | Watch |
| Belmont | R3 | Coastal Spirit | 7-4-5-3-6 | 54% | 3.8 | $7.50 | $2.35 | −12.1% | Drift Risk |
Live race data, filters and full runner detail are available in Race Explorer.
Open Race Explorer →
Built for decision support
✓Full-field comparison
✓Probability + core quality
✓Steam, flat and drift context
✓Weak-core and S4 trap awareness
Analytics, not tips
RACEMODEL is designed to help users interpret racing data. It does not promise winners, guaranteed returns or certainty. The objective is fewer, higher-quality decisions supported by evidence.
Early access
Help shape the next version of RACEMODEL.
Join the early-access list for product updates, dashboard releases and testing opportunities.