So when I say daily newspaper, I really mean website although if you want to use a newspaper you still can.
One of the best ways to make a profit from your betting is to find horses that are improving and bet on them before they start winning.
As soon as they start winning then the rest of the betting public will start to bet on them and their odds will drop, meaning that there will be no value left in betting on them.
Of course, the key to doing this is to be able to highlight those potential improving horses before anybody else spots them.
And today I’m going to show you a quick and simple way of doing just that.
First of all head to the Sporting Life. You can get straight to the days racecards by going to http://www.sportinglife.com/racing/racecards
Click on a race that you’re interested in. You’ll see a screen like below:
You want to click on the Ratings tab to go to a page like:
The first thing you want to do is make sure that the majority of horses have six ratings. You can see above that all the horses have six ratings except for three, of which one has five.
The ratings going from most recent on the far right to least recent on the far left.
In order to calculate the level of a horses improvement over these last six ratings we want to subtract a horses rating from the rating it had in it’s previous race.
Let’s take Heinrich as an example.
His sixth rating is 65. In the next race he achieved a rating of 62 so…
62 – 65 = -3
Then he achieved a rating of 62 again so…
62 – 62 = 0
Then he achieved a rating of 59…
59 – 62 = -3
Then he achieved a rating of 55…
55 – 59 = -4
Moving from race to race Heinrich has achieved an improvement of:
-3, 0, -3, -4
Next we want to add all of these figures together to get the total number for improvement and it is -10.
This immediately tells us this horse seems to be declining in performance!
Now all you need to do is repeat this for every runner in the race to get an idea of how much improvement each horse is making.
But… generally it’s much easier to make an assessment of a visual representation rather than numbers.
It’s possible to put these figures into a graph so that for our example race it would look like:
When the race is visually represented you can see immediately that there are not many improving runners in this race.
The only horse that is showing sign of improvement is horse number 2, which is Cadmium, whilst horse number 4 and 1 are showing the largest signs of decline. These are the runners It’s Taboo (4) and Heinrich (1).
I was going to explain how you could make a graph like this in Excel. But, then I thought why the heck… I’ll just make it for you.
Which I’ve down and you can download the spreadsheet below. There are two versions, one is an XLSX file and the other is an XLS file for older versions of excel.
All you need to do is enter the ratings and the graph will automatically calculate for you.
Click To Download In xlsx
Click To Download In xls
Let me know how you get on by leaving a comment below.
This looks as if it could be very helpful. I will certainly give it a go.
Thank you Michael
Thanks Tony.
Hi Michael
I downloaded the spreadsheet but its not making me a graph any help would be appreciated
What version of Excel are you using?
excell 2011
Cracking bit of kit.
Many thanks
Thanks Jimbo, I hope you find it useful.
Excellent information Michael many thanks for that and the excel spreadsheet. Most useful.
Thanks Wendy, pleased to hear it is going to be useful 🙂
Hi Michael
Fantastic help once again many thanks for all the support and help. Very quick question. I use both puntology and monte in my selections. If this form shows as a negative value do they still work with the monte system ? Was just thinking how good it would be adding this as an additional column if it took negative numbers. Thanks again
Thanks Bob. There should be no issue adding negative figures into the Monte Carlo software.
Don’t we get the same answers by subtracting the latest from the 6th ago
Well spotted Peter, you do but there are two reasons to do it this way. The first is that by doing it manually you get a feeling for the recency of decline/improvement. The second is that you can change the bar graph for a line graph to visually represent the decline/improvement 🙂
Very interesting info but I am a bit confused. I selected the 17:25 Wolverhampton race today (Monday 3rd Nov) to try it out and entered all the ratings into your spreadsheet (thanks for that by the way). I found that the Royal Trooper ratings (all in the 80s and 90s) gave him a huge plus score compared to the other runners. The ORs were so different to the others that I checked through the form on the At The Races website, only to find that the last 6 races were rated in the 50s and that the higher scores were from this time last year. What does this mean? Is it just a mistake on the Sporting Life page or are there other factors involved?
Hi Mark, it looks like Sporting Life have a data error on that horse this morning. If you click on the horses name it has the correct OR’s listed and it’s the summary for that horse which is broken.
Michael,
Does the BSE software factor this in?
Thanks, Ian
Ian, the BSE software takes this strategy and supercharges it 😉
Noticed same as Mark also in Wolverhampton race 1.45 pm today for Look Here’s Al. Looks like they have the most recent the wrong way round? If that’s the case is the info on site trustworthy?
Also when looking through the example above Cadmium has a race in which it was rated 56 which has been adjusted to the 51 shown above I presume because an apprentice rode it? Even if that’s the case shouldn’t the rating for the purposes of showing improvers still be 56 in which case Cadmium would have had a negative figure too?
Thanks, Ian
It’s a tough one because most sites get it wrong sometimes including RP. Ideally you should always cross-check multiple sites. And yes, ideally you would change Cadmium to 56 if you’re happy to spend the extra time. As with any data based approach, the better the data that is put in the better the information you get out.
Michael is there a site that can be relied on for accuracy e.g. if I use the RP – which you’ve said can also be wrong – then your worked example would have Cadmium as -5 and the biggest improver would be Silvee which finished second at 20/1, next would have been Superior Duchess, last at 100/1 and next Flumps which finished 3rd at 20/1.
Two places at 20/1 out of 3 ain’t bad but it wouldn’t have picked Cadmium as the winner.
It’s opened my eyes to another way of analysing a race but if none of the data sources can be relied upon for the official ratings……..what does BSE use?
All the sites potentially have data issues, unfortunately that is one of the constant struggles with UK racing data.
Currently we use Racing Post data feeds and so that is what BSE uses, but BSE is based on my own speed ratings and factors in other things such as confidence intervals.
thanks for the spreadsheet michael.
i look forward to trying it out
regards
george
Thanks George, I look forward to hearing how you get on.
definitely look into this, thanks for invite to your website
looks a good way for looking for value e/w bets as well
thank you so much for the advice…
ill definitely look to doing this before making a decision for a bet in future,
its makes so much sense..
especially if i can find an improver at odds that are worth betting on e/w
Thanks Michael 🙂
I tried to use the ratings in The Sportinglife sometime last year but found that the figures shown in the “Rating” list were often incorrect. I took this up with The Sportinglife and there was some admission that there was a problem. Does that problem still exist?
Look at today’s ratings, for example the next race up as I write – 1615 at Kempton.
Ronaldhino’s last 4 ratings are shown as 120 121 120 120. Click on the name of the horse and the last 4 OR are 120 121 80 80.
Rayvinn Black has ratings of 119 119 119 119 119. Checking on the horse shows just 4 ratings at 119 then a couple of gaps. The Racing Post has 119 119 119 126.
Next, look at Samarian – 118 118 135 133 130 127. Clicking the horse shows that LTO the rating was 122 and all the others are pushed back one. The Racing Post has 125 135 133 130 127 122.
One more, The Game Is A Foot – 95 95 95 106 106 106 agrees with the horse’s actual record as per the SL but the racing Post has 95 102 102 106 106 106
There is merit in the approach of trying to identify improving horses but if the OR figures given by The Sportinglife are incorrect then it can’t be used as a basis for the approach you outline. I would far trust the Racing Post figures.
The approach can be used with any publication online or offline that provides the information, you should use whichever you feel most comfortable with. I chose the Sporting Life in this example because it is much clearer to visualise with their clean layout.
Hi
Was intrigued by the method you have described and wrote my own program to obtain the measure of improvement you suggest.
I tested 10 races and the following were the results I obtained:
Horses with top rating placed first – 3
3rd – 1
4th – 3
6th – 2
7th – 1
The winners had improvements of 15, -2 and 14. I then checked all improvements of 10 and above. There were 8. As the total SP of the two horses with 15 and 14 scores was 15 points, the net profit was 9 points at SP. The profit on Betfair SP was 14.47 points net.
It therefore seems to pay if you only only back those horses who have an improvement score of 10 and above.
Thank you for the comment Alan, and great to see that you went straight out got it tested. I wouldn’t draw any conclusions of the best way to use this process from a ten race test, I would recommend a minimum of 100 winners in your sample before during conclusions.
I would also recommend breaking down your stats by race conditions as well as the types of runners in a race. You will find that making profits with this approach in AW flat sprints is very distance to long distance National Hunt races. You will also find that using it in races where all runners are negative is different where a higher percentage are positive.
But great you made a profit on the first day 🙂
Thanks for you comment Michael. The reply seems to have lost its meaning in the transposition of your dictation but I get the drift.
The article describes a method of how to find ‘Improving horses’. and illustrated the article with an improving horse that won. The conclusion drawn was that the best improver won. Now you have added other factors to the equation! Naughty!
As usual you are correct in the context of size of the sample. Prior to reading your reply I was formulating other methods to add to the basic concept as nothing is as simple as described.
The improving horses today over nine are:
Redcar………15.30…That Be Grand
Southwell…..13.40…Day Star Lad
……………….14.10…Powerfullstorm…Penelope Pitstop…Sweet Talker
Exeter……….14.15…Somersby…Oscar Hill
Kempton…….19.10…Exceedexpectations
These were selections from 12 races.
Thanks Alan, you’re right I always assume that readers know I suggest considering other factors and my methods are primarily for finding the strongest contenders but I should reiterate it in each post where it’s relevant.
Thank you for the shortlist for today 🙂
Michael which are the most profitable race conditions can be used this graph ?
To be honest I haven’t gone through and tested all the different race conditions separately so I wouldn’t be able to say which would be most profitable. What I would suggest is that you use this on the race conditions that interest you the most. The more you use it the more you will understand those race conditions and how it works with this approach and you can adjust the method slightly to suit them.
Once again another cracker Michael.
Due to time constraints I have only trialled it in races with 8 – 10 runners and only at 4/1 or better SP
Results to date
3/11
2nd 5/1
2nd 10/1
4/11
W 20/1
KEM 1810 Hanalei Bay
KEM
Which won at 20/1? Could only see one 20/1 winner Acclio in the 2 o’clock at Redcar but Strike A Light the favourite was the biggest improver.
finger trouble
KEM 1840 Inoko
They will only be selections if they go off 4/1 or better and minimum 8 runners
Stuart
Hi Ian
True, Strike a Light was the biggest improver but at SP 5/2.
I take the first improver with a SP of 4/1 or better and Acclio was it.
KEM 1810 Hanelei Bay lost
KEM 1840 Inoko won 9/2
All round a good day
Stuart
HI Michael
What effect does a horse coming off a layoff have if his in the negative but his fresh and trained for the race
I haven’t looked at that specifically Douglas, so it would be something to test. My feeling is that you should bump up their score slightly but… I’m always cautious when a horse is racing again for the first time since a long layoff unless I’ve identified them as racing their best races in the first two or three runs coming back.
Many thanks for the article.
You could make the Improvement column the following:
=SUM(IF(OR(C11=””,B11=””),0,C11-B11),IF(OR(D11=””,C11=””),0,D11-C11),IF(OR(E11=””,D11=””),0,E11-D11),IF(OR(F11=””,E11=””),0,F11-E11),IF(OR(G11=””,F11=””),0,G11-F11))
Then you no longer need the Calculations table.
Great, thank you for the improvement Garry 🙂
Hello Michael!
You are really great and very dedicated to horse racing.
Read a lot of Your excellent stuff here. Even, I was subscriber of some of Your services earlier.
I could write a lot of nice words about You.
But, reason because I am writting right now is the simple one and concerning Your excell file (calculator) for improving (declining) horses.
Final result for every horses is, simply, difference between last OR and first OR for that horse! (do not need to enter all data-all OR form race to race).
Just to know 🙂
Regards,
Goran
Hi Goran, thank you for your message. You are absolutely right, that is a much quicker way of doing it. The reason I didn’t say that is because entering the ratings forces you to take note of them and understand they how work. You’ve obviously already done that though otherwise you wouldn’t have spotted the quicker calculation 😉