HORSERACING DATABASE
FR / EN / ZH

When you have all the information,
racing is no longer a game of chance

10 years of French horse racing history (1996-2005) structured and organized for quantitative analysis.

Vector Data Model

A perfect relational mesh connecting major entities for high-frequency extraction.

Table : Races

Racetrack, Weather, Prize, Distance

Table : Actors

Odds Spread, Shoeing, Handicap

Table : Yields

Human Entities

Horses, Jockeys, Trainers

1. The Fundamentals

An Unprecedented
Volume

A raw relational database. Every race, every actor, every yield has been captured, normalized, and indexed to enable instantaneous joins.

115K+
Archived Races
1.6M
Finish Lines
124K
Unique Horses
5K / 12K
Jockeys / Trainers
-- Core Algorithm : Extraction Multicritères
SELECT c.id, a.cheval_id, r.quarte_ordre
FROM courses c
INNER JOIN acteurs a ON c.id = a.course_id
INNER JOIN rapports r ON c.id = r.course_id
WHERE a.deferre = 'Des 4'
AND a.ecart_pmu_pmh > 10
AND c.meteo = 'Bon'
AND a.recul = TRUE;
2. The Engine (Algorithms)

Query
40+ Parameters

Having data is not enough. The Huggy system integrates querying algorithms designed to instantly cross-reference over 40 dynamic fields.

  • Extreme Granularity

    Shoeing, blinkers, 25m handicap starts, weather, autostart type, sex, age, pedigree history.

  • Indexing Performance

    Database integrity is ensured by an optimized architecture, preventing relational table corruption during heavy queries.

3. Quantitative Modeling

From Archive to
Prediction

The past models the future. By isolating Typical Profiles over 10 years, the engine deploys statistical algorithms to validate your strategies.

Bernoulli's Law

The system does not promise certainty; it calculates mathematical expectations. Using the binomial distribution, Huggy computes the exact probability of profiting over a sequence of N races.

P(X ≥ k) = ∑ i=kn (ni) pi (1-p)n-i

Monte-Carlo Simulation

Generate thousands of random trajectories based on the historical variance of your queries. Identify the maximum Drawdown before risking real capital.

  • > Yield Calculation
  • > Standard deviation of losing streaks
  • > Bankroll Optimization
4. 2026 Horizon: AI Fuel

A Goldmine for
Machine Learning

In the AI era, access to structured data is the most precious resource. This archive possesses two irreplaceable assets:

1. Organic Alpha

Current markets are saturated with algorithmic arbitrage. This database contains the Smart Money Spread captured before the bot invasion.

2. NL2SQL Corpus

With 40+ parameters per table, it's the perfect playground to train Reasoning Models to transform business questions into SQL joins.

Licensing & Acquisition

Native export in SQL (.sql) or CSV for Data Warehouse integration.

Analyst Segment

R&D Pack

950€
  • 2 years of sample data
  • Full relational structure
  • Limited PMU/PMH Spread
The Complete Dataset

Historical Deep-Dive

Full Archive

4 500€
  • Full 1996-2005 Database
  • PMU/PMH Spread Included
  • Native SQL structured access
  • 40+ Variables per race

Enterprise / Quants

Reasoning Engine

45 000€+ / IP
  • Intellectual Property Transfer (Source Code)
  • Bernoulli & Monte-Carlo Module
  • Original Query Builder
Notes on a horse

Natural Language Output

Institutional Contact

For technical audit requests, R&D samples, or full license acquisitions.

Direct Email

contact@huggydatabase.com