This post will show you the process of calculating speed ratings for any given horse race. I’m not going to go into the complexities of the mathematical theory behind each stage of the process, nor the various formulae for producing the data needed to arrive at the speed ratings.

A driving instructor doesn’t help you by explaining to you in minute detail precisely how the internal combustion engine works, but rather his aim is to teach you how to drive the car safely to your chosen destination.

Similarly, I’ve written this article to teach you how to get started compiling your own speed figures for horse racing.

Besides, a great deal of the analysis data in this article comes from Nick Mordin’s book *Mordin On Time* and I strongly recommend if you are halfway serious about compiling your own speed figures then you search out a copy for yourself.

It used to be available from publishers Aesculus Press but I think nowadays copies are hard to come by, and a search on Google may be in order.

I’ve written a guide to speed ratings for use at UK all-weather tracks called Patterns In The Sand and my main aim when writing that book was to show you what to do with the speed figures you produce, and how you can use them to have a positive effect on your betting. You’ll notice throughout this article I make several references to my book, and if I spark an interest in speed ratings today and you decide to buy it, you’ll have my eternal gratitude.

**But I’m digressing!** To demonstrate how to compile speed ratings I’ll take you through the working example of how I produced the figures for the meeting at Wolverhampton on Monday 1st February 2010.

**This is a summary of the process:**

- Compare the actual race winning time with the Standard Time
- Is the time slower or faster than Standard Time?
- Calculate the Going Variance
- Calculate a speed rating for the winning horse
- Produce figures for the other runners

Using a table as shown below (I use Microsoft Excel but any spreadsheet program is fine), the first thing to do is to enter the scheduled start time of the race into column A. In this case the first race on the card was the 2:05. We then look for the Class of the race, which we can see was Class 6. The Class figure goes in column D.

The figure in column E is taken from the Class table (see below) and 6.9 represents Class 6. This table essentially shows us how many seconds per mile a horse in a given Class can be expected to perform below the Standard Time. As you might expect, the higher class horses are expected to run nearer the Standard Time.

The calculations can be broken down into five steps:

### STEP #1 : Compare the actual race winning time with the Standard Time

Now we look up the winning time for the race. We can see this 6f (actually 5f 216yds, which is 4yds short of six furlongs) race was won in 1m 16.69secs. Deduct the Standard Time for the course and distance. (Standard Times for the all-weather tracks, and how they are compiled, is shown in my book). For a 6f race at Wolverhampton the Standard Time is 1min 10.14secs

Winning time is 1min 16.69secs

Standard Time for a 6f race at Wolverhampton is 1min 10.14secs

The difference is 6.55secs

This is the difference for a 6f race, and as there are 8 furlongs to the mile, 6f is six-eighths of a mile, or three-quarters of a mile, or 0.75 miles. To equate the time difference to that of a mile, we must divide by 0.75 to give 8.73secs. This figure is placed into column C. The Standard Times tables in my book also show the fractions you will need to divide by, in order to equate each race to one mile. For example, a 5f race is five-eighths of a mile, and so you divide by 0.625

### STEP #2 : Is the time faster or slower than Standard Time?

Next check the figure in column C against that in column E. We are comparing the time taken to win this race against the standard time. If C is greater than E then we place a MINUS in column F. If C is less than E then we put a PLUS in column F.

In our example, 8.73 is more than 6.9 so we put a MINUS in column F.

The actual difference between column C and E goes in column G. We put 1.83 in column G for the first race.

The process is repeated for all the races in the meeting, and our table for Wolverhampton on February 1st looked like this:

### STEP #3 : Calculate the Going Variance

The next thing to do is to remove the extremely fast or slow times. So in column G put a line through the two highest, and the two lowest figures (for an eight race card) to leave the middle four times. Add together the four times, and find the average by dividing by four. This will give you the Going Variance for the track that day.

With a seven race card you will be left with three times, so instead divide the aggregated time by three to get your Going Variance. With a six race card, discount the highest and lowest times, to leave you four times to find your variance figure.

Normally you will find the figures in column G are either all positive or all negative, but sometimes there will be a combination. In our example, after weeding out the top and bottom two, we were left with the following figures…. -0.60, -1.26, -1.83, and lastly -1.91

Added together these make -5.6 which divided by four gives an average, and our Going Variance, of -1.40

### STEP #4 : Calculate a speed rating for the winning horse

To arrive at the final speed rating figure for the winner of the race, deduct or add (in our case we deduct as the Going Variance is a negative figure) the Going Variance from the figure in column C, then multiply by five, and subtract the result from one hundred. This is the speed rating for the winning horse, and goes into column B.

In our example, in the first race, we deduct 1.40 from 8.73 to give 7.33. Multiplied by five this gives 36.65 and when we take this from one hundred we are left with 63 (rounded to the nearest whole number).

The final table for our meeting at Wolverhampton is shown below:

You can see that comparatively speaking, the winner of the 4:50 race with a speed rating of 80 was the best performance of the day.

### STEP #5 : Produce figures for the other horses

Now that you have a speed rating for each of the winning horses, you now need to allocate a figure for the remaining horses in each race. Write the speed figure next to the winning horse. Next, divide the number of lengths each horse finished behind the winner, by the distance of the race, and deduct this from the winner’s speed figure.

A simple example to start with, is a horse beaten by one length in a 1mile race. One divided by one is one, and so the horse would receive a rating one less than the winner.

Let’s say the horse in question was beaten 2 and a half lengths, in a six furlong race (remember, 6 furlongs is 0.75 miles). 2.5 divided by 0.75 equals 3.33 and rounded to the nearest whole number gives us 3. In this case we would allocate a speed rating of three less than the winning horse.

One last example, and this time our horse is beaten half a length in a 1m2f race. 0.5 divided by 1.25 equals 0.4 which we would round down to zero. The horse would be allocated the same speed rating as the winner.

If we look at the race featured at the beginning of this chapter, the 2:05 race at Wolverhampton over five furlongs, we have calculated a speed rating of 63 for the winning horse Miss Firefly. The second horse Tamino was beaten by a neck (one quarter of a length). Six furlongs is 0.75 miles, so 0.25 lengths divided by 0.75 miles is 0.33. This is nearer zero than one, so I would also give Tamino a rating of 63 for his performance. The third horse Metropolitan Chief was another neck distance back, and so beaten half a length by the winner. 0.5 lengths divided by 0.75 miles is 0.66 and so I would give the horse a rating one less than the winner, ie. a 62.

**1st Miss Firefly 63! W 6f**

**2nd Tamino 63 W 6f**

**3rd Metropolitan Chief 62 W 6f**

NB. I explain my speed ratings annotation in detail in my book, but the exclamation mark denotes a winning performance, the ‘W’ denotes the track Wolverhampton, and the 6f shows the race distance.

As with many new processes, and if you are anything like me, calculating your first set of horse speed ratings will take you quite some time. I was forever re-reading Andrew Beyer’s instructions, and referring back and forth with my own calculations. But I soon got the hang of it, and as you might expect, with some practice you will also see for yourself the task becomes a lot easier and very much quicker.

If you don’t think you’re going to have the time to research and maintain your own database of speed ratings, or you would simply rather let someone else do all the hard work for you, then I’ve got some good news. When you buy my guide to speed ratings Patterns In The Sand you will also get a **free** one month trial of my speed ratings service. Check it out here [...]

So there you have it. That is how you can produce your own speed ratings for each horse in a race. I agree it’s not very exciting nor romantic, and nothing like the compelling stuff that we usually read on the sales pages of all the betting wonder system web sites. There is no ‘secret formula’ and I’m not revealing to you some missing link that I found scribbled inside a dusty old journal in my grandfather’s attic. It is just simple mathematics.

But now that you have a speed rating for each horse, what do you do with the figures? In Mordin On Time the author suggests keeping your figures in the weekly supplement to RaceForm Update or in a notebook. This is a simple way to record your figures, but the problem later arises when you want to retrieve your figures to rate a particular race. The Wolverhampton meeting we have been using as our example involved 108 horses, and during the 2008/2009 winter season more than 3,000 horses ran on the all weather alone.

In my book (see below) I explain how I solved this problem.

### Patterns In The Sand: a guide to speed ratings for all-weather racing

This is my book. I’ve called it Patterns In The Sand because it specialises in the use of speed figures at the UK all-weather tracks.

**This is some of what’s covered by the guide:**

- What are speed ratings?
- How to produce the speed ratings
- How to calculate Standard Times
- An introduction to the all weather tracks
- Winning betting strategies

my wife as bought your book patterns in the sand for me, and she will go up the wall if she finds out you can get the information thats in your book for free instead of 11.95????

Hello Chris, and welcome to the blog. With respect, I could point you to at least a dozen web sites where you can download this book for free. I’m sure there will already be links on various betting forums revealing the download page of this latest edition too. Fortunately (or unfortunately, depending on your outlook) that is the nature of the internet — everything can be found for free if you are prepared to look hard enough. If you really resent supporting this web site and paying for the ebook, then I’ll happily refund you, just drop me a line and let me know.

hi Paul no I don’t want a refund, I am just hoping that your book can help with a little profit on the all weather, I know it will take awhile to get enough data on board so if you have any pointers I would be grateful.

Hi Chris, I’ve included in the book several ideas on how best to put your speed figures to good use, and you should find it helps you uncover good value betting opportunities. I’m always here to offer opinion and assistance too. It does take some patience and effort to build your initial database, but in the meantime you can take advantage of speed ratings services such as Inform Racing at http://informracing.com

i want to learn speed rating

Hello Ashish and welcome to the blog. This article here should help you calculate your own speed ratings, and if you want a more detailed approach and ideas on how to make profits from speed ratings, I have written a guide. More details at http://www.allweatherspeedratings.com

How do we use these speed ratings, eg how to turn them into bets?

Hi Callum, and welcome to the blog. Clearly, the most obvious way to use the speed ratings would be to bet the top-rated horse in each race. You’ll score plenty of winners by doing this, but you will almost definitely lose money over time. You will improve your profitability if you specialise, and I’ve researched several strategies for using speed ratings to make a profit, and outlined these in my book Patterns In The Sand. More details on the book here http://allweatherspeedratings.com

dear sir, please explain me about how to decide horse probable odds with method

Could this book work outside the UK.

Hi Bill, welcome to the blog. Yes, speed figures can be calculated for any horse, on any track, if you have the time posted for the race, and a set of standard times for the track, to use as a benchmark.

hello Paul

I am in hot water with financial woes. its hard to find a software that will consistently work. i know i dont total depend on software but i am over my head with bills/debts and collection. if i could just get tips to make a killing one time wheither its pick 3-6 or superfecta i would be happy. could you subject anything. thank you I use Andrew Beyers and it gives me a guide, but i need as much experience as a handicapper. there is no perfect expert handicapper but i do want to come better at handicapping. LOL

Hi Edward, and welcome to the blog. A good place to start, especially if you’re looking for reliable speed rating handicapping figure, would be a service such as Inform Racing. Put “inform racing” into your favourite search engine, and you’ll find them. Regards, Paul

Hi Paul,

I’m interested in compiling my own set of speed figs for racing outside the UK. I read in one of the posts above that it is possible to do. What about creating my own standard times, does your bokk explain this in detail and can I do this for grass and a/w racing or best to stick with a/w racing. Thanks

Thanks

Hi Michael, and welcome to the blog. The methods explained in the book can be transferred to racing in other countries. If you have any problems please feel free to drop me a line to paul [at] skybluekangaroo [dot] com. You can easily apply speed ratings to turf tracks, its just that I’ve found results not as reliable due to the inherent changes in going, and even moving running rails which can alter distances.

Hi Paul,

Like your self I have discovered the stone cold perfection of speed based analysis, although I’m more interested in times above or below standard times having spent a couple of years working out real actual times in 100ths of a second seperation.

I’m under contract now to a major tipster so I cant reveal exactly how I work: & a great deal of the fag is carried out by a computer bot running to my spex.

Like yourself I’m amazed at the lack of interest in speed data in the UK, with predictors more interested in what a horse has for breakfast than in how fast they run.

The one piece of knowledge I can impart, though you’ve probably worked this out for yourself, is that horses run at the same speed ALL THE TIME & it is only minor variations in amount of runners, course wind speed, variations in turf moisture etc, that a/c for speed variation. It is also noticeable that horses really do run at different speeds on different courses, this being down (specially on the flat) to different course undulations etc etc.

I found your site stimulating & much closer to my view of flat racing. I dont do Nat. Hunt & it baffles me why anyone bets on it, although I do enjoy it as a spectator sport.

Please keep my details to yourself: its been great to find someone like-minded.

Hi Paul. I’ve been reading some of your articles online (there great reads by the way) and I’ve been interested and actively punting on horses and greyhounds for about 12months now. After the Cheltenham Festival I’ve got the bug for it again and would like to make my picks more precisely and hopefully make some pennies (obviously).

Towards the end of the festival I started using trends and stats (provided online by a lad called Matt Bisogno) for the races and I was doing much better. I thought I could transfer this to the weekly British racing program, compiling my own stats and trends but theres a lot of races and a lot of horses… it’s pretty overwhelming!! Would speed ratings be the best place to start and how long did it take to compile yours?

Thanks for your time, and good luck in all your punting

Nick

Hi Nick, and welcome to the blog. The downside to getting an edge over the market, and availing yourself of information and data not commonly, well, available… is that it takes time and effort. Building your own database of speed figures does take time, although I’ve restricted myself to the all-weather tracks and horses. Nonetheless it does take regular input. But the upside is that you stumble across the horses that come in under the radar. If you are short on time then Inform Racing provide an excellent service for speed ratings and if you’d rather someone does most of the heavy lifting for you as regards trends analysis, feel free to take a look at my Horse Racing Trends service.

I would like to ask you a question regarding track variants. Andy Beyer uses the $10,000 par to create class pars for a track. Do you know which class in British AW racing is equivalent to US $10,000 par? regards vasilis

Can you please tell me What is “Standard time” and where can i get it for Indian race tracks?

Hello Anil. The Standard Time for a track is the benchmark time for that particular distance. I’m afraid I don’t have figures for Indian race-courses, but a search using your favourite search engine might reveal some answers. Failing that, you would need to gather a significant sample of race-times and compile your own Standard Times by finding the average or mean times for each distance.

You are using the old Nick Mordin method to calculate ratings for the beaten horses using distance beaten and race distance, which according to Andy Beyer’s methods is fundamentally incorrect.

This method has to be used with an amount of caution (but that applies to any method), especially when considering longer races. But I wouldn’t go so far as to say the process is fundamentally flawed. What would be your opinion, and what would be your suggested alternative approach?