Why analysts keep coming back to racing as a dataset, and the honest limits that come with each of its strengths.
Which UK and Irish racecourses stage the most races? We counted every race since 2011. The answer is not the famous ones.
The Hong Kong Jockey Club publishes veterinary notes on its horses. We hold 53,966 of them. This is what they contain and how to use them in an analysis.
The tables behind racing data, how they join, and two traps: IDs that only mean something inside one market, and profile totals that quietly look into the future.
The distance between runners is the messiest column in racing results. This is how UK and Hong Kong write it, and how to turn it into a number.
Where AI genuinely helps with racing data, where it just invents things, and why a language model needs the table in front of it.
Working SQL and Python for the features most racing models start from, with the leakage traps that catch people out and the thin-history problem.
Where UK sectional times exist, and where they do not. Coverage by year, by country and by course, measured on every runner since 2017.
A horse racing data pipeline is two jobs, not one: a big one-off load of history and a small daily job. Here is how to build both and what the daily job costs.
USA and Canada horse racing data is live on HorseRaceDatabase: what is available, what is not included and where to start.
Recent Comments