Generating accurate merit ratings for racehorses has long been a manual exercise. Complex though it is to computerise the process, it is achievable as our ratings prove.
Our system incorporates numerous unique facets. Trainer debutant ratings to assess the most probable rating an unraced horse is capable of. Mathematical equations to decay past performance ratings in view of the length of time since they were recorded and a horse’s form since. Goodness of fit statistical testing to derive the most likely performance ratings given a particular race result. Exclusion scenarios to rule out improbable fits. And a correction process to constantly improve the accuracy of previous ratings based on new evidence.
How long does it take the computer to rate a race? 0.03s – that’s automation for you.