How do i rate a horses performance

[Amorga81]

Hi all

This forum is excellent and it is great to see so many people helping each other out.

Sorry if this has been discussed before.

I have been looking at creating speed ratings but am struggling to work out how the rating should work.
I have read a few books where it mentions creating standard times for each class and some have mentioned for each age restriction, but this is where my logical head is struggling. As like a running race it comes down who can run the race in the fastest time, so why would it matter what class you have run in or what age you are?
Sorry if this is a silly question but I might have blinkers on
So when creating my standard times I have thought about just using class 4 races at all ages by distance and ground. This is because all tracks have at least class 4 races (ascot doesn't go lower than class 4) this way it sets a level playing field. So if a group 1 horses time is compare against the class 4 standard time he would more than likely get a higher rating.

Could I have your thoughts on this?

Thanks in advanced
Andy

 

[LukeyBoy]

Compiling Speed figures.

The lbs per length which I devised for each course and distance can be obtained from my standard time list (see below) along with the furlong distances to use. I have highlighted the use of the lbs per calculation (Distance adjustment) below. So in this example, the lbs per length calculation...

www.theukbettingforum.co.uk

 

[mick]

Hi A Amorga81 my knowledge of the subject is limited but on this part of the forum reading davejb and TheBluesBrother might help answer your questions.

 

[Outlander]

Hi all

This forum is excellent and it is great to see so many people helping each other out.

Sorry if this has been discussed before.

I have been looking at creating speed ratings but am struggling to work out how the rating should work.
I have read a few books where it mentions creating standard times for each class and some have mentioned for each age restriction, but this is where my logical head is struggling. As like a running race it comes down who can run the race in the fastest time, so why would it matter what class you have run in or what age you are?
Sorry if this is a silly question but I might have blinkers on
So when creating my standard times I have thought about just using class 4 races at all ages by distance and ground. This is because all tracks have at least class 4 races (ascot doesn't go lower than class 4) this way it sets a level playing field. So if a group 1 horses time is compare against the class 4 standard time he would more than likely get a higher rating.

Could I have your thoughts on this?

Thanks in advanced
Andy

You are probably going to struggle just using class 4 races as there won’t be enough data , ideally you need to adjust each time for the class of horses running in the race, the going , weight and wfa if you include weight in your calcs before you can then use them to create standards but again this is not simple the BHA class numbers like class4 are not that reliable , you can easily get class 4 races chick full of class 2 or even class 1 horses lightly raced and starting on the ladder, Frankel beat Nathaniel in a class 4, the biggest help I can give you is read the compiling speed figures thread as much as you can , and if you are just starting out read my thread

UK All Weather Speed Ratings Template

explanation to follow

www.theukbettingforum.co.uk

this has templates that do it all for you and give you an idea how things work
if you download the version on post#33 it has a full list of my old standards I calculated, I have improved them significantly since but might be a good starting point.
I don’t have a list of my new standards because I have them all in individual course sheets.
But if your just starting the old standards probably an improvement on RP standards so might prove useful.
To get full use of the template though for calculating ratings etc you do need HRB results data to paste into the templates.

 

[davejb]

Personally I don't think class has a bearing really (see a bit further on though) - the standard time definition that I tend to think of is the time taken for a horse rated 100 to cover the distance when carrying 9-0. (126 lbs, for those who don't think in st-lbs) on good ground. (Think I got that right!) .
I think I got this next version from the RP themselves in response to a query i raised with them a good while back 'the average of the 10 best times over the past 5 seasons for a mature horse rated 100 carrying 9-0 on good ground'.... which is pretty similar.

The class angle, for me, comes into it when you realise that better horses will tend to contest better races and there will be a tendency - though not a rigidly observed phenomena - for the better races/classes/horses to set better figures and faster times.

I'd be tempted by the RP 'best 10 rated 100 plus' approach, but for some tracks you might well not find any 100+runners ever winning there, so for me I'd be looking at the BHA ratings of the better winners that they do have - I'm look at compiling standards currently (still bashing data into shape for the next few days) and I'll be seeing what winning times might adjust to if I adjust for weight and the winner's OR..... I don't have a preference as yet, but I'll see what the data actually looks like when I work it two or three different ways, as my aim is simply to get figures that allow me to compare performances on different tracks as accurately as I can. It is, after all, about producing ratings that will give a decent clue to the races.

Currently O Outlander is the chap to go to on here about this, I'm just a bloody minded sort who has to bang his head against the wall for a while before admitting he's not as good as he thought at stuff! as you've already heard, just sticking to one class won't give you enough data.

Dave

 

[The Hare]

Excellent replies already.

The compiling of Standard Times is by far the hardest part of the task of compiling useful speed figures.

You firstly need to work out a method of how you are going to assess the figures. With weight or not? If you read the excellent Compiling Speed Figures from TheBluesBrother. anything by O Outlander and davejb. This will give you all the information you can ask. I doubt there is anything in books that surpass this or throughout the web.

So once you decide whether weight or no weight, you then need to determine how many lbs you will give to each length (or lbs per second which would be best if you have the actual finish times, which are available on BHA website and IrishRacing website). These will vary as distance increases and also the lengths per second are changed based on going and track type for All Weather Racing - which is why actual times work better.

The best way then has to be as per O Outlander and adjust all completed times based on weight carried (if using weight) and the ability of the horse (ideally using a ratings service such as BHA, Racing Post or Timeform or any others that are based on distance and weight carried)> Without weight your only option is class, again using just Class 4 races removes a lot of information. You need to calculate a measure of the difference in times between each of the Classes and adjust the finishing time accordingly.

There are also the complications of the actual distance as rails are moved around on round courses which will need to be accounted for in the finishing time. As you can see a lot of work involved in finding the actual information and trawling a few sites. The BHA results are best for Rails movement, results from HRB, Racing Post, Timeform or some others.

These adjustments give you a kind of equality to all times so that only the going should have much affect on the finishing times after that.

Then you will need to take a decision on how too choose the standard. Put them in order and choose an average of some number. The problem with the average of the 10 best times is that distances have a varied number of races, some very few. An average of the fastest 10 would give you much higher ratings for those distances with very few races. A better method would be a percentile or average of the first 20% say of all races at a distance.

Using 2 and 3yo races also requires adjusting for Weight for age so another complication.

It isn't easy and needs constant checking as new distances arrive and ground conditions alter as courses are updated drainage and the like.

Good Luck.

Kevin