Skip to content



In this article we are referring to ratings as a single number that measures a horse’s overall performance in a past race. It doesn’t matter what class of race it was, what track they raced at, what weights were carried, how far horses were beaten by or how fast or slow the race itself was… well constructed ratings aim to reflect all of that information in one single number.

When faced with the uncertainty of lining up various form lines for an upcoming race, ratings can help you to understand the differences between them with clarity and confidence.

They are typically used as a form analysis tool to help forecast the performance of each horse in an upcoming race. That forecast is often expressed as an expected rating for “today’s race.”

Those conclusions can be directly used in betting decisions, or taken a step further to develop assessed prices for comparison to the actual betting market and identification of “value” runners that may be good betting prospects. You can see examples of this output via the rated prices made available from different providers on The Betfair Hub.

While ratings aren’t essential to becoming a winning punter, good ratings can save you hours in form analysis time and allow you to be far more consistent and confident in the way that you assess races.


Not all ratings are created equal. The presence of ratings in a horses form doesn’t necessarily make them accurate or reliable. Good ratings require a high degree of racing intelligence to construct and take a significant amount of effort to maintain.

Sure, you can program some rules and have your computer automatically generate ratings, but the reality is they will only be useful to a certain point and are unlikely to offer any competitive advantage in the betting market.

Quality of inputs aside, ratings are traditionally made up of three key elements, as shown in the diagram below.


The starting point to creating any horse performance rating is assessing the strength of a race. The intelligence that goes into this element will largely dictate the quality of the final rating.

The class label of a race i.e. BM70, Open Hcp, Listed Race etc. is a general guide to the quality, but in a rating sense, they are largely unreliable. Every day we see races that are significantly weaker or stronger than the class label suggests, ranging from Maidens to Group 1 races.

Each race is a unique and dynamic event, which must be treated accordingly.

Our form analysis and betting is based around our own WFA Performance Ratings, which were developed in-house and continued to be maintained by detailed reviews on a daily basis. They take a multi-dimensional view of race strength and use a group of key factors to make that assessment. The main factors are represented in the diagram below.

The circular nature of the diagram above shows the nonlinear approach taken to assessing each race. In other words, the strength of each race is not always in proportion to the input of any one individual factor.

For example, a fast time does not by itself necessarily mean a strong race and equally so, a slower times does not always indicate an inferior race. A big margin spread between runners at the finish may indicate a better than average race, but that is not always the case. A race with relatively weak lead up form & ratings does not always constrain that race to a similarly moderate measure of race strength.

The strength of each race is assessed based on all of the individual factors.

Whatever the method used, race strength is typically expressed as a number and this acts as the base for final rating calculations. There are no rules as to the scale or range of numbers that should be used. Some ratings use kilograms consistent with the handicapping weight scale; others use pounds, while our own ratings work on a totally independent scale.

They key to a good rating is being able to accurately identify races that are stronger or weaker than the general market perception. For example, a Provincial Maiden race where the standard set by the winner was more consistent with midweek or Saturday City class; or a Saturday class BM race where the winner has set a standard more consistent with a Listed race or better. Equally important is the ability to identify black type races that are no better than standard Saturday class races or any other race that was clearly inferior to what the class label suggests.

Once race strength has been determined, each horse’s rating can then be calculated by applying individual adjustments.


Any horse that finished behind the winner must have their rating reduced for the margin they were beaten. The traditional approach is to use a certain number of rating points per length. For example, if your scale was 1.5 points per length, then a horse beaten two lengths would have three points deducted from the race strength measure.

A more sophisticated approach would use a variable scale that changes depending on the distance of race and margin itself. For example, a horse beaten one length was probably giving everything in the drive to the line, while a horse well back in the field beaten eight lengths is very rarely pushed out and would have finished closer if asked to give everything it had.


In most races, horses carry different weights, which affect their performance, and this must be reflected in their rating. For example if two 4YO geldings go to the line locked together in a dead heat and one carried 54kg, while the other carried 60kg, it makes no sense to rate these performances as equal. The horse that carried 60kg was clearly the better performance.

Just how much should be allowed for the effect of weight is highly debatable. Traditional theories taught us that 1.5kg in weight affects a horse by one length in performance, but our research shows that this is greatly exaggerated.

Different ratings use different allowances for weight. The most basic approach is to come up with a number of points per length (as the traditional theory of 1.5 per length did.) A more sophisticated approach would involve a variable scale where the effect of weight would change depending on race distance and perhaps some other key factors such as the weight carried itself and pace of race etc. They key point though is that the laws physics say that weight carried does have an impact on performance so it must be allowed for.

Weight carried adjustments also need a base on which to calculate. If one horse carried 58kg and another carried 55kg then how much adjustment is applied to each?

The traditional approach is to use the handicapped minimum weight in the race and subtract it from the horse’s actual weight carried. Using the example above, if the minimum weight was 54kg then the horse with 58kg would have a positive adjustment of the number of points equal to 4kg in weight (because it carried +4kg above the minimum.) If a horse was handicapped on the minimum 54kg and then had a 3kg claiming apprentice it would effectively be carrying 3kg less than the minimum. In this case, the horse’s rating would receive a negative adjustment of points equal to 3kg.

Our approach makes an adjustment by comparing weight carried to each horse’s weight for age based on their age / sex, time of year and distance of the race. This is an important rating concept, as weights cannot be considered in the same way for all horses. An early 3YO carrying 58kg for example should not be viewed in the same way as a 5YO carrying 58kg. The early 3YO not as physically developed as the 5YO and may be carrying 7kg more than WFA (depending on the time of the year and distance of race), while the 5YO is carrying 1kg below WFA. Normalising for those differences means that the ratings of horses can be directly compared regardless of age, sex or the time of year. Traditional ratings handle this aspect by applying a WFA adjustment to past ratings as a horse ages.


Some approaches to constructing ratings also apply a further adjustment to allow for instances of back luck in running or other disadvantages that affected a horse’s performance. They could include but are not limited to:

  • Being caught wide in the run
  • Striking interference at any stage in the race
  • Being advantaged or disadvantaged by the track pattern
  • Jockey tactics / decisions that had a negative impact on performance

Allowances for these factors can give a more complete measure of a horse’s performance, but require a massive amount of effort to maintain. If you make the adjustments for one race then you have to do it for all races. There is also the issue of whether you are accurately adjusting for the real impact of these factors and doing it consistently on a week-to-week basis.


Once the above factors have been determined, each horse’s final rating can be calculated:

Race Strength + Beaten Margin Adjustment + Weight Carried Adjustment + Subjective Adjustment (if any) = FINAL RATING

That rating then becomes the measure of the horse’s performance in that race which can be referred to as a part of its form history for next start and races beyond.


Horse AS Wgt FP Marg RS Marg Adj Wgt Adj Final Rating
Farson 3G 56 1 0 91.5 0.0 1.0 92.5
First Approval 3C 59.5 2 0.5 91.5 -0.9 4.0 94.6
Almas Rossa 3F 56.5 3 2 91.5 -3.3 3.0 91.2
Astley 3C 56 4 3.8 91.5 -6.5 1.0 86.0
Congressional 3G 55 5 5.1 91.5 -9.7 0.1 81.9
Ontoff Ofthe World 3G 57.5 6 5.6 91.5 -9.4 2.1 84.2
Hard Promise 3C 58.5 7 7.4 91.5 -11.8 3.1 82.8
Charlie Cheval 3G 56 8 7.6 91.5 -12.1 1.0 80.4

Race Strength (RS) = 91.5: this is the base measure given to the race (using our own rating scale)

Marg Adj = Margin adjustment applied to each horse for its beaten margin

Wgt Adj = Weight adjustment applied to each horse for the weight it carried relative to our WFA scale.

Final Rating: This is the sum of race strength, margin adjustment and weight adjustment as shown in the formula above.

What’s worth noting with this example is that the winner of the race doesn’t always end up with the highest rating. In this case, First Approval was beaten 0.5 lengths by Farson, but actually ended up with the highest rating in the race (94.6) courtesy of carrying +3.5kg more in weight.


There is no such thing as a magical number that makes your racing predictions infallible… racing is far too complex for that and the circumstances a horse faces from one race start to the next can be very different.

However, ratings are an incredibly useful tool in helping to efficiently make more confident and consistent race assessments. When used with the right perspective in conjunction with your own flexibility and judgement they can provide a real competitive advantage in identifying horses that are underestimated by the betting market (to back) or overestimated (to lay.)

After that, it’s all up to the horses and jockeys… and of course a bit of luck!

Go To Betfair App