Horse racing - the IQ system

[zabadac]

I came to the forum for the football ratings, but I am also interested in the horses.
We seem to have problems now with the virus but some racing has survived, in South Africa and Ireland.
I 'm confident nevertheless that in three weeks time the measures taken by the authorities will bear fruit and we are going to see some return to normality.

Now about this idea of mine it is based on the method of experts pools.
That is people who are employed by governments and industry to give their opinion on certain issues and these opinions are averaged out to reach conclusions.
Naturally these people have to be experts in the subject under investigation and they are handpicked.

I could do this in football many years ago, that is collect opinions and average them out.
But I can't do it any more.
Nowadays the internet tipsters no longer give opinions for selected matches or groups of matches.
One may advise me to bet on a premiership match, the other picks one from Italy and one from Germany and so on.
So I can't figure out how to create a pool of football experts.
But the results I had with this in the past were not bad.
It was the late eighties or early nineties and we had the pools coupons every week back in those days (I am from Greece).
So I could do this pooling and separarte the good sources of information from the bad.
That registered an imporvement, but it did n't last long. The newspaper I was doing it for did n't really like it, also others things were happening and the football betting
company decided to gradually kill the pools and move towards fixed odds betting.

In racing however it is different.
The websites who offer a daily card service have to work out all the races without exception.
So info of this type can be gathered.
It is the French who go for this mainly. There are many websites from France with daily card service, as opposed to 7-8 in the UK.

Take a peek here:

http://www.turfpronos.fr

This is a website that comes out of Morocco and it is dedicated to races of the French PMU system.
The guy who runs it is a very knowledgeable and studious and for tomorrow you can take a peek for Turffontein.
You can also see how neatly he makes the tables, using saddle numbers always.
In every case he says at the bottom "synthese", meaning the result of the averaging process.

The idea of "la synthese" is that "the process of aggregation creates a composite tipster who is better than the best tipster of the company".
It is not a simple process to find the formula that does this averaging, if you want to be rigorous about it.
To calibrate the system first you have to look at the individual contributors. So what is the strength of Racing Post's second choice relative to the first,
of the third relative to the second and so on.
Once you did that then you have to pit them one against the other. There are some advanced computer algorithms that do that and they come in
various names such as "pool aggregation", "probability conflations" and others.
You have to effectively find a coefficient of strength for each of the contributors.
The top experts gain high coefficients of strength (or indices of strength rather as it is an exponent really), but some gain low coefficients.
Who are the "useless", those who give lousy tips that is ?
An obvious case of a useless tipster is one who is an idiot. His guesses are random numbers so he does n't count, or should n't count.
Another case -which the algorithms manage to identify- is when someone copies from the others and has no opinion of his own.

The results of the method in horse racing are not bad.
With betfair -which we don't have here- maybe it's in the black.

There are the critics also however.
They say to me "with what you did you managed to approximate the market or recreate the market so you have n't done nothing really".
Well maybe, but here is an example of my work (not the Frecnhman's) from some days ago, before racing stopped:

 

Quote

Saturday, 13 March 2020

WOLVERHAMPTON A.W., Friday 14-03-2020

The tipstersynthesis verdict page


Wolverhampton A.W. 5.25 (UK time)   11 run, scratches: 8
---------------------------------------------------------------------------------------------------------
racingpost.com:             7-1
skybet.com:                 1-2-4-7-8
timeform.com:               7-4-2
sportinglife.com:           1-9-2
attheraces.com:             9-2-10
attheraces.com (predictor): 9-7-1-8
oddschecker.com:            9-2-10

Verdict: 7(19.0%),1(18.0%),2(14.0%),9(13.0%),4(9.0%),10(7.0%),6(5.0%),3(5.0%),5(5.0%),11(5.0%)
         (difficulty index = 8.8)
---------------------------------------------------------------------------------------------------------


Wolverhampton A.W. 6.00 (UK time)   10 run, scratches: 
---------------------------------------------------------------------------------------------------------
racingpost.com:             2-7
skybet.com:                 7-2-1-6-5
timeform.com:               7-1-6
sportinglife.com:           6-4-3
attheraces.com:             7-5-2-9
attheraces.com (predictor): 9-6-1-5
oddschecker.com:            7-5-9

Verdict: 7(24.0%),6(13.0%),2(13.0%),1(10.0%),5(9.0%),9(8.0%),4(7.0%),3(6.0%),8(5.0%),10(5.0%)
         (difficulty index = 8.8)
---------------------------------------------------------------------------------------------------------


Wolverhampton A.W. 6.30 (UK time)   10 run, scratches: 
---------------------------------------------------------------------------------------------------------
racingpost.com:             7-2
skybet.com:                 2-9-6-4-3-7-5-8
timeform.com:               2-8-9
sportinglife.com:           2-4-6
attheraces.com:             9-7-6
attheraces.com (predictor): 9-7-8-1
oddschecker.com:            9-7-6

Verdict: 2(22.0%),7(15.0%),9(15.0%),8(9.0%),6(9.0%),4(8.0%),3(6.0%),5(6.0%),1(5.0%),10(5.0%)
         (difficulty index = 8.7)
---------------------------------------------------------------------------------------------------------


Wolverhampton A.W. 7.00 (UK time)   8 run, scratches: 
---------------------------------------------------------------------------------------------------------
racingpost.com:             5
skybet.com:                 2-5-4-3-7-1
timeform.com:               2-8-5
sportinglife.com:           5-2-4
attheraces.com:             2-5-4-1
attheraces.com (predictor): 2-5-1-7
oddschecker.com:            2-5-4

Verdict: 5(30.0%),2(28.0%),4(9.0%),1(8.0%),8(7.0%),7(7.0%),3(6.0%),6(5.0%)
         (difficulty index = 6.3)
---------------------------------------------------------------------------------------------------------


Wolverhampton A.W. 7.30 (UK time)   7 run, scratches: 
---------------------------------------------------------------------------------------------------------
racingpost.com:             4
skybet.com:                 3-7-4-1-2
timeform.com:               3-5-1
sportinglife.com:           7-1-6
attheraces.com:             1-7-5
attheraces.com (predictor): 1-2-7-4
oddschecker.com:            1-7-5

Verdict: 1(23.0%),7(18.0%),3(15.0%),4(14.0%),5(11.0%),2(10.0%),6(9.0%)
         (difficulty index = 6.6)
---------------------------------------------------------------------------------------------------------


Wolverhampton A.W. 8.00 (UK time)   3 run, scratches: 
---------------------------------------------------------------------------------------------------------
racingpost.com:             2
skybet.com:                 1-3-2
timeform.com:               2-1-3
sportinglife.com:           2-3-1
attheraces.com:             2-1-3
attheraces.com (predictor): 2-3-1
oddschecker.com:            2-1-3

Verdict: 2(52.0%),1(26.0%),3(22.0%)
         (difficulty index = 2.8)
---------------------------------------------------------------------------------------------------------


Wolverhampton A.W. 8.30 (UK time)   12 run, scratches: 8
---------------------------------------------------------------------------------------------------------
racingpost.com:             6-4
skybet.com:                 6-2-4-1-5-8
timeform.com:               5-4-6
sportinglife.com:           6-4-1
attheraces.com:             6-4-2
attheraces.com (predictor): 4-6-2-5
oddschecker.com:            6-4-2

Verdict: 6(31.0%),4(23.0%),5(8.0%),2(8.0%),1(6.0%),3(4.0%),7(4.0%),9(4.0%),10(4.0%),11(4.0%),12(4.0%)
         (difficulty index = 7.9)
---------------------------------------------------------------------------------------------------------
thetipstersynthesis v 2.1, (c) CosmiC Software 2015-


(*) difficulty index = the average price in probability, based on unit stake, 100% payout percentage

In my opinion to make this method strong what we need is to find the best possible contributors.
It always pays of course if you know something to make further eliminations, but it's about IQ really.

Well, can you in some sort of way help in this ?
As you can see I have exhausted -or nearly exhausted- the UK sites that offer free tips.

Also does anyone remember a tipster site of a fellow named Geoffrey ?
That was many years ago, in 2002. I remember that because there was a local pub with SIS and it closed in October of that year.
I corresponded with him and he also gave me his phone number.
But I lost everything, I don't have a web adress or mail any longer and he does n't seem to come out in google searches.
The thing is he was doing the same thing and he had columns upon columns of tip contributors that were being averaged (in his free samples).
Does anyone remember him ?



 

 

Edited March 19, 2020 by zabadac

[zabadac]

Not a single one from Wton I see !
I did n't follow them from the start.
I made a bet in the penultimate on Moss Gill -as he was very hot fav.
In the last I lost.