The Data Universe
A model's power is limited by its fuel. Our framework is built on a comprehensive data universe, capturing every facet of a horse's career and its competitive environment.
Core Performance
The horse's official résumé, including finish positions, times, pace data, and in-race incidents.
Intrinsic Characteristics
The athlete's profile: age, sex, pedigree, and physical data like equipment changes.
The Human Factor
Quantifying the impact of the jockey and trainer, including their historical success rates and specific patterns.
Race Conditions
The playing field: distance, track surface, condition, post position, and quality of competition.
Market & Betting Data
The "wisdom of the crowd" captured in odds and betting pools, a powerful measure of public expectation.
Pre-Race Fitness
Indicators of conditioning, including official workout times and days since the horse's last race.
The Analytical Engine
We transform raw data into predictive power through a systematic, multi-stage process that combines statistical modeling with machine learning.
From Data to Prediction
Our methodology is designed to find a true 'edge' by identifying horses the market systematically undervalues. Follow the four key steps from data collection to a validated predictive model.
1. Feature Engineering
Raw data is transformed into insightful metrics like normalized speed figures, pace ratings, and form cycle trends.
2. Hybrid Modeling
We use an ensemble of models, combining logistic regression with gradient boosting to capture both linear and complex non-linear patterns.
3. Market Comparison
The model learns to predict the *error* in the public odds, focusing its power on finding market inefficiencies and true betting value.
4. Rigorous Back-Testing
The model is validated on historical data it has never seen to provide an unbiased estimate of its future accuracy and profitability.
The Performance Dashboard
The final output is an interactive tool for strategic decision-making. Select a plausible future race scenario to see a real-time probabilistic forecast.
Key Factors
Positives:
Negatives: