The DataBase Ratings is our brand-new free basic ratings service, covering all Australian horse racing.
These are a basic set of ratings designed to give an accurate, data-driven view of each runner’s chances. We do not expect them to win on their own simply by betting into every race. But they do give you a decent head-start on the market with the top-rater winning 28% of the time and losing just 3% at Best Tote.
The DataBase Ratings are produced for us by a data scientist with a tertiary education in actuarial studies. Our analyst has been able to combine his passion for racing with his professional abilities to produce a thorough set of daily ratings exclusively for Champion Bets readers.
How are they produced?
The DataBase Ratings are derived from our private analyst’s racing database, which captures over 40 different race-specific, runner-specific and track-specific form factors from every race. All recognised form factors are covered such as track, distance, conditions, time and weight.
These factors are combined to produce a number of different models that give a quantitative representation of how horses, jockeys and trainers perform in each race, and thus builds up a comprehensive bank of their respective capabilities.
When a race field is released, the models combine to produce the single DataBase Rating to give an accurate guide to each runner’s chances.
Crucially, the combination of these models, and hence the ratings, are ever-evolving.
Using artificial intelligence, the model learns which combinations of data are most successful and retrains itself on a weekly basis, adjusting the ratings calculation to suit. This is critical, as racing (and the edge needed to win) is always changing.
When & Where
Every day for every TAB meeting. If they’re racing, there will be ratings published on the Champion Bets site every morning.
The only exclusions are races which include first-time runners – with no data available, the system cannot calculate an accurate rating.
How to use
There’s no set way to use them: you may want to use them as the basis of a strategy, as a way to round out your exotics, or as a further check against your own ratings and selections.
A few things to note from looking at the data history:
- Races with a clear spread in the ratings (ie, one or two horses well clear of the rest) tend to have good value.
- Overlays are best used for metro racing, where the form is more reliable: and in particular the more specific higher class metro races (ie, stakes races).
- There are no real trends in distance ranges, other than a higher degree of accuracy for long distance races (over 2300m).
- The ratings don’t include variables related to the run the horse will get: the model won’t know that a horse will map to get caught three wide, and also obviously there is no information contained about any track bias / patterns that presents itself on the day.
Post a comment below if you have any questions at all about how they are generated or how you might use them.