The Power Of Ranking Ratings

What I want to share with you today is something from one of the best analytical racing minds in the US. He looks at racing in a completely unique way and his ideas have been significant in getting me to where I am with my betting today.

The chances are that, being focused on UK racing, you’ve not heard of him before. And his name is Dave Schwartz. You can check out his website here.

Now while I recommend that everyone read Dave’s work, it is developed for US racing and will need to be adjusted to work in the UK and IRE. What we want to take away are the concepts and ideas to put them to work for us in our style of racing.

The technique that you’re about to learn Dave calls Reynold Ranking. It’s a superb approach to using the ranks of horses for different factors to find the strongest runners in the field.

And yet, whilst being so powerful, it’s incredibly quick and easy to use!

First of all let me explain where this method came from…

Dave was digging into his research, as he is often found doing, and during one particular session he came across an English physicist from the 19th Century called Osborne Reynolds.

Osborne was a prominent engineer in the field of fluid dynamics. Don’t worry, you won’t need to know anything about this to use the approach.

It was 1883 when Osborne invented his number… the Reynolds Number. This number has a lot of uses, such as allowing ship designers to use small scale models to determine how they will work on the open sea.

The way the Reynolds Number works is by multiplying together the diameters of different pipes. with a number of other factors. He managed to show that when the result of these numbers reaches a “magic number” the flow of water becomes erratic.

Okay, so you’re probably wondering how this can help you with horse racing.

Well… while reading this Dave’s brain went off in a different direction with the though…

“If it’s possible to use this approach to determine when the flow of water becomes erratic, then surely we can use it to determine which horses performances are likely to be erratic!”

Let’s assume you’re trying to build a model to find selections on Handicap Chase races at Cork.

You are using four factors to try and find out what types of horses win these races. For this example I’m going to use the RnkPFP, RnkACSPCL, RnkCst10 and RnkSHorPro factors from the Racing Dossier.

The first thing we do is pull out a list of these factors for winners in at Cork in Handicap Chases with 15 runners or less. I’ve used 2013 data to do this.

Below is a sample of the data that we get:

6 3 3 11
4 2 1
2 10 7 8
2 10 7 8
10 3 12
8 6 1 7
15 5 12
4 1 4
7 1
7 8
10 8 5
10 11 7
2 1 2 1
4 5 6
3 5
10 8
8 4 3 2
4 3 1 1
7 6 7
13 9
13 9
7 9 7 3
1 4 1 6
6 3
7 7 1 4
10 9 13

What does this tell is?

Well, at first glance not very much because there is no single factor that can be used to reduce the number of runners.

But this is where Dave saw the opportunity.

He created Reynolds Ranks by combining every factor with the other factor, simply multiplying them together. So you would do:


RnkPFP x RnkCst10

RnkPFP x RnkSHorPro

RnkACSPCL x RnkCst10

RnkACSPCL x RnkSHorPro

RnkCst10 x RnkSHorPro

Once we’ve done that for all the winners we can look at the worst (highest) score that a winning horse has achieved.

RnkPFP x RnkACSPCL RnkPFP x RnkCst10 RnkPFP x RnkSHorPro RnkACSPCL x RnkCst10 RnkACSPCL x RnkSHorPro RnkCst10 x RnkSHorPro
110 80 180 70 80 60

Now we start to have a base line to look for potential selections in a race!

If we take a race from the Racing Dossier with these factors:

We can then multiply the factors together as we did above, but we now have a baseline for the worst winners in each factor.

Horse RnkPFP x RnkACSPCL RnkPFP x RnkCst10 RnkPFP x RnkSHorPro RnkACSPCL x RnkCst10 RnkACSPCL x RnkSHorPro RnkCst10 x RnkSHorPro
Lucky William 28 10 4 70 28 10
Fosters Cross 5 7 7 35 35 49
False Economy 35 30 70 42 98 84
Nearest The Pin 6 60 66 10 11 110
Saint Gervais 33 6 15 22 55 10
The Black Russian 60 110 10 66 6 11
The Amarillo Kid 156 168 120 182 130 140
Back To Balloo 90 27 72 30 80 24
Dazzling Susie 16 64 104 16 26 104
Theatre Mill 44 44 33 16 12 12
Formidableopponent 112 126 112 72 64 72
Instant Impacked 39 156 143 36 33 132
Beckwith Star 63 91 42 117 54 78
Gentleman Duke 48 4 16 12 48 4

The above table shows the result for the horses in this race for each factor multiplied together.

But now we can use our “worst winner” scores to remove any horse with a RnkPFP x RnkACSPCL of more than 110 or a RnkPFP x RnkCst10 of more than 80 etc…

That reduces the field of fifteen runners to just:

Horse RnkPFP x RnkACSPCL RnkPFP x RnkCst10 RnkPFP x RnkSHorPro RnkACSPCL x RnkCst10 RnkACSPCL x RnkSHorPro RnkCst10 x RnkSHorPro Sum
Lucky William 28 10 4 70 28 10 150
Fosters Cross 5 7 7 35 35 49 138
Nearest The Pin 6 60 66 10 11 110 263
Saint Gervais 33 6 15 22 55 10 141
Back To Balloo 90 27 72 30 80 24 323
Dazzling Susie 16 64 104 16 26 104 330
Theatre Mill 44 44 33 16 12 12 161
Gentleman Duke 48 4 16 12 48 4 132

We have reduced the field by half almost instantly to find only those runners who have the profiles we know are the strongest for horses that win these types of races.

I have then added a new column called Sum, which is the sum of the scores, and the horse with the lowest score is the strongest. Which would be Gentlemen Duke.

Very quick, very simple and… very powerful.

But imagine if we had added together the totals for the winning horses when we were profiling these race conditions?

Then you would know that from the previous winners, the worst score for a single horse was 315 with nearly 75% of the winners having a score of under 200.

That means we can remove Back To Balloo and Dazzling Susie. We can also consider Nearest The Pin not to be too strong a threat.

Horse RnkPFP x RnkACSPCL RnkPFP x RnkCst10 RnkPFP x RnkSHorPro RnkACSPCL x RnkCst10 RnkACSPCL x RnkSHorPro RnkCst10 x RnkSHorPro Sum
Lucky William 28 10 4 70 28 10 150
Fosters Cross 5 7 7 35 35 49 138
Saint Gervais 33 6 15 22 55 10 141
Theatre Mill 44 44 33 16 12 12 161
Gentleman Duke 48 4 16 12 48 4 132

You are now left with five possible contenders in this race with the strongest being Gentlemen Duke, Fosters Cross and Saint Gervais.

I used to use this approach on it’s own to make regular profits. Like anything, it will require a bit of practice. But once you have it mastered you will be able to profile and rank races using the Reynolds Ranking method in a matter of minutes!

Michael Wilding

Michael started the Race Advisor in 2009 to help bettors become long-term profitable. After writing hundreds of articles I started to build software that contained my personal ratings. The Race Advisor has more factors for UK horse racing than any other site, and we pride ourselves on creating tools and strategies that are unique, and allow you to make a long-term profit without the need for tipsters. You can also check out my personal blog or my personal Instagram account.


  1. Hi
    A very interesting article.I take it that you have to produce a list of factors for each race type at each course.
    Where do I find the list of factors to enable me to produce these figures.
    I have racing dossier so I can use the same ratings that you have.

    1. You would need to have enough historic ratings to be able to perform the analysis on the race but… this approach can work with a surprisingly small small sample of races if you go niche enough into the conditions. I have had it work effectively with just 10 past races. If you don’t have history of RD ratings Leslie try putting a post in the forum and asking if anyone will share theres with you.

    2. Hello

      This is the type of analysis which I have loved for a long time. It is mathematical in the sense that a study of

      factors identifies a pattern from which it is possible to make a prediction. Unfortunately ,for me , it is a few

      centuries late, but not too late, to be interested in articles like this.

      Best wishes

  2. Hi

    I employ a procedure similar to this in which I rank all horses in a race and chose the best three as my selections. The strike rate of these selections averages over 40% (yesterday it achieved 60%). Having narrowed the field down to three selections the art is in selecting the winner from the three.

    The Strike Rates of previous selections at every course when backing all three horses are an aid as is the value of the profit or loss generated.

    The base data is all extracted from Sporting Life

  3. Fascinating article Michael. I do a similar thing with the weight distance and class figures, and would love to take it further, but have been unable to so far. It does absolutely narrow the field down though. It would be nice to get more on this topic though.

      1. Hi Michael – no not stuck really, as I find it quite effective, it is just that I love the idea of reducing a horse to a number and am always willing to learn more if it will improve thiings, and in fact even if it doesn’t improve matters I still like to hear it. I’ve been around a long time and have never heard of the Reynolds Rankings before, so naturally it is of great interest and my ears pricked up.

        I’ve always created my numbers on the laws of physics, ie the extra or less weight carried on it’s last best run, ie where it came first or second. The distance it runs in furlongs and yards, even down to the last yard, all very relevant at the end of the race when they are running out of steam. The class figures in handicaps 0-75 etc. But I find it hard to factor anything else into those figures like the going, the jockey etc.

        1. I’m not a big connections user myself but going I think is very important. Looking at it from a law of physics perspective, if a horse is running in sloppy mud then it’s going to become more tired quickly and the weight carried will affect it quicker than if it’s running on good ground.

    1. Hi Wendy and Leslie

      I am a system developer and have developed a system that produces the selections I publish on the internet.

      This will involve me in despatching daily the factors I use to create my selections. You then run a program I will e-mail to you initially in order for you to create the same list

      There are six factors I use to create the final score. I have the facility in the program for you to exclude any of my factors. The system is to be developed further to allow for the addition of users factors.

      I have no one other than myself to test the efficacy of the system and you would be assisting me in the use of the internet facilities. If you would like to avail yourselves of this program e-mail me at **email address has been removed due to the security of publicly posting an email address in the comments section.

      To Michaels comment re Sporting Life – I download raw data from many sources on the internet and use it to create data for my systems from which I rate, record, produce statistics and monitor the multiple systems I have.



    2. Hi Wendy & Leslie

      Sorry e-mail address is :

      **email address has been removed due to the security of publicly posting an email address in the comments section.

      1. Hi Alan
        I would be very interested in looking at your at your system and help with checking the proformance.
        How do I get in touch with youas your Email address has been deleted from your comments.
        Regards Leslie

  4. a very intresting read michael but what i want to know will this work with the rating used in the racing post and how would they be used if at all could you do a write on on this so that punters like myself can further understand it

    1. Thank you James. You would need access to historic ratings which can be hard to get for RPR. I do have night before ratings historically for RPR so I could use if there is interest?

  5. Michael, is there any definitive set of filters used for Reynolds Ranking? The reason I ask is that in this article, you use the above filters, but I notice that on each card I have saved, when I click on the Reynolds button, the listed filters used are different. The one in particular which I found odd was the listing of SP Odds Rank as a filter in creating a RR number. I was trying to work out how you would have the SP Odds Ranking until the race has run, so how could it be used beforehand in creating this possible power ranking? Just curious. cheers

    1. The Reynolds Ranking on the race cards will use any of the ratings that it can on the race card selected. You can see which of the ratings it is using by clicking on the Reynolds symbol and seeing the ratings checked. If you don’t want it to use any of them, you can simply uncheck them. Changing race card with different ratings it will automatically update to use those ratings instead. You can see in the image attached.

      Reynolds Ranking

      If you choose a race card which has SP Odds Rank on it, then this will be included pre-race as well as post-race, but there will be nothing in the column for the horses SP Odds Rank, so it won’t contribute anything to the RR score.

      1. Thanks so much Michael. That sorts my question of the cards. Yes, I was curious about SP, so at least I know now that if a column is blank, it is simply not used in the calculation. It still does beggar the question about what ratings to use definitively. I want to be able to add a RR to the data I am compiling, but am unsure as to exactly what ratings to use, thus my question about a definitive set of ratings to use to create a RR. What do you advocate personally? Thanks again

Back to top button