AI horse racing data analysis
Xiaomage Sports organises Hong Kong race data with a statistical model — past runs, draw, weight, pace and jockey/trainer pairing — into a per-race data report.
Updated 2026-10-27
Data used
- Past runs: last 6 placings, margins, class and distance
- Draw: 3-year record at the same course and distance
- Weight carried and body-weight changes
- Pace and running style: leader, midfield or closer
- Recent jockey and trainer data and pairings
Method
The model runs offline before the race and seals its output (same input, same result). Reports list the top-rated runner, model rating, rating tier and supporting/risk factors, then compare with published results. Read the method
Limits
- Data and information for learning and reference only; no result is guaranteed.
- Debutants have limited evidence.
- Going, weather and late changes may not be reflected in pre-race data.
FAQ
- What does the model rating mean?
- A 0–100 rating relative to the same race. It compares runners and is not a forecast of the result.
- Where does the data come from?
- Public racecards and historical results, reviewed against published results after each meeting.
- Is it paid?
- Full data reports are currently free after signing in.
Unofficial information compiled by Xiaomage Sports. Not affiliated with, endorsed by or acting for The Hong Kong Jockey Club. Fixtures and venue terms follow the organiser’s same-day notices.