What Is A Horse Racing Rating?
Ratings are a very mis-understood medium in betting. They are however exceptionally powerful if used correctly. Today I want to make sure that you don’t fall into the biggest trap that most punters do when using ratings.
You cannot get away from ratings in horse racing. They are part and parcel of what horse racing is. If you look at the Racing Post there is the Official Rating, RPR, Top Speed amongst others. This means that if we want to progress our betting we need to understand what these ratings mean. And by understand I don’t mean have an idea what it is measuring, but really understand how ratings work.
I am going to kill the biggest myth right now…
- The top rated horse is NOT necessarily the best horse in the race
At first glance this sounds like an illogical statement. After all, if it’s top rated then surely it is the best horse in the race?
No, this means it is top rated for one rating which tells us that this horse is the best horse in this race for whatever it is that this rating is measuring. However it may be joint best. Ratings are estimates of either an aspect of a horses performance in a race or a prediction of an aspect of a horses performance in a race. Either way the key word to note is that they are estimates. This means that there is going to be a margin of error in them. The error can either work in our favour or against. By this I mean the horse could be slightly under-rated or over-rated.
If we take into account that there is a margin of error around a horses ratings then we can immediately see why a horse may not be top rated.
In the screenshot above you can see a random race I took from the Racing Post. Notice that Desert Sunrise and Forceful Flame are top-rated for RPR. Then we have three runners in Seven Of Clubs, Sakhee’s Rose and Kodatish that are joint second-rated. If we were following the thinking of most punters then we would assume that these runners are not as good as the first two.
However we aren’t normal punters. We’re better than them and we know that these ratings are estimates. If we allow just a 5% margin of error, which is probably too little, then the horses rated 84 could have a real rating as low as 80 and as high as 88. The horses rated 83 could have a rating as low as 79 and as high as 87.
So the top two horses are likely to have a real rating between…
80 – 88
The second top three horses are likely to have a real rating between…
79 – 87
Looking at those new ratings would you be so confident that the top two rated are the best horses in the race?
Based on this new information we can see that in fact the top 5 runners are very closely rated in this particular factor. In fact the other runners in the race are not that far below either when we apply the margin of error. Suddenly this race becomes very competitive if we are only looking at the RPR for information. This is backed up by the fact that a horse rated 82 won the race.
This is the reason that we cannot rely on a single rating to make our selections. We need to look at a range of ratings that measure different aspects of a horses performance and allow for a margin of error in the ratings creation. When using ratings we ask ourselves…
- Which horses are best suited to all the conditions?
- Which horses are best suited to the most important conditions?
- Which horses have the best overall figures?
Start using this in your betting and you will start to see that races are more competitive than they first looked. However this works both ways. You will also begin to see races where the competition is a lot lower than it initially looks and you can target these races for your betting!
Interesting read – thanks. Underlines that, allowing for often bizarre market moves, only about 1/3 of ‘favourites’ actually win. Does a similar rating spread apply to ‘Official Rating’ & how is this rating derived?
It applies to any rating. The based explanation on how the official handicap rating is created is at the official site 🙂 http://www.britishhorseracing.com/resources/media/ratings.asp
There is also the problem of how to rank/rate any race or runner, there are so many different ways to achieve the same objective. You could use margin of error as in the article, standard deviation with a confidence level, fuzzy logic, reynolds method etc etc. Then you need the raw data, to actually be able to calculate the rating in the first place and know how to create the rating from the data, then know which rating to use for any individual race.
Absolutely, there are a number of pieces to the puzzle and I will be looking at them all.
…& even having done all that the horse goes on and wins anyhow! At least not often / not often enought (depending what you are trying to achieve)
read your last few posts with great interest. After many years trying I developed my own system, which I used on two races today, then used another system using ratings and betting forecast and amazingly the results were the same!!!!. The ratings i used were or/rpr/ts/lto.Was this just a fluke?
Thanks Peter. Two races is not enough data to get an idea of how effective the ratings are, you would need a lot more than that to know whether it was luck or due to good ratings.
Sorry should have said my own system uses no ratings, just betting, conditions, class
and weight, but I do take your point about the no. of races.
What you are looking at here are ‘spot ratings’. They are purley a comparison of each horse’s current rating by the unofficial handicapper (Racing Post ratings on this case) agaonst the weight they are set to carry (allowing for weight for age).
The mpst important fact in looking at a race is “is this horse currently improving or is its form deteriorating”?
A spot rating cannot possible tell you that
That is the key!
If you want to find out any more then send me a message 😉
Thanks for the comment Gareth. In my investigations the single most important factor is preference over the conditions.