Explanations

Neural Rating

This is essentially a rating based on a broad spread of statistical factors. Achievements are measured relative to the horse population as a whole (and sometimes market expectations), rather than being specifically based on individual performance. Twenty factors are used within the calculation and this rating is likely to be the most closely aligned with market expectations for an upcoming race. That is, the neural rated price, when looked at over a very large sample, will generally have a strike rate which is in line with the winning probability associated with that price.

Class or Weight rating

Every class of race is given a numerical value (in kg) and runner ratings are determined by taking the assigned Class value for the race, adding the WOL (weight over limit) amount and subtracting a beaten margin adjustment. The beaten margin adjustment involves converting the amount a horse has been beaten in lengths, to an equivalent weight value (in kilograms). This type of rating has the benefit of not being influenced by race pace, time or track conditions etc.

Race Time rating

Each runner is rated purely using their overall race time. The rating is a performance measure which compares the race time against a benchmark or Par time for the track and distance. Each rating contains adjustments for par time 'strength', potential weight for age improvement, race pace, weight carried and prevailing track conditions. These adjustments are made so we can compare 'apples with apples', regardless of when, where, or how races have been run. This type of rating does not have any race Class considerations.

Sectional Time Rating

Each runner is rated purely using their last 600m sectional time. All other comments relating the overall race time rating (above) apply to the sectional time rating. The unique aspect of this rating is that we have individual sectional times (spanning several years) on which to base the rating, available for all runners, across most metropolitan and provincial race meetings conducted throughout Australia (we use digital timing, off race videos to obtain these ourselves).

Pf Score

This encompasses each of the previous four ratings methodologies and takes the form of a 'consensus score' from 0 to 100. How each runner has 'ranked' in each of the four ratings categories (or three, where there is no sectional rating available) contributes to the final consensus score. This value appears for all runners listed in the meeting form guides (provided you have a current monthly ratings subscription) and links to the relevant Ratings Suite published for that meeting (as well as providing a link for individual meeting purchases).

Run Style

The run style indicator categorises where each runner has typically settled in its past performances. Our research has shown that regardless of early speed ability a horse will try and take up its 'usual' position in running. Early speed ability is the factor which will determine how easily it is likely to be able to attain this position and how much energy it will have available to finish the race off.

Predicted Settle Position

As the name suggests, the predicted settle position is where we are predicting the runner to settle in the upcoming field of horses. This is a useful indicator when looked at in combination with run style. For instance, a horse with L (leader) as a run style, which is predicted to settle 4th in running, will most likely have to work hard to hold a prominent position.

Reliable Data Quantity

This is a simple yes/no indication of whether a particular race has sufficient 'relevant' form data for the ratings to be 'reliable'.

How are the various Rating Ranks arrived at?

The Meeting Quick Guide format shows the top four ranked runners in each rating category. These rankings are NOT based on career best performance or any sort of average rating figure etc. Our ranking (and pricing) algorithm is based on what is known as a 'paradox of odds' approach. We determine a likely 'performance envelope' for each horse based on past ratings and have each horse participate in what amounts to thousands of simulated contests, to determine a winning probability for each runner in an upcoming race.

What is the best way to use this information?

ABettorEdge has a series of articles (blog posts) which give some pointers as to how the various ratings, run style and predicted settle position information, may be used in 'concert' to produce a satisfying punting outcome. These articles can be found over at the ABettorEdge blog.