• Hi Guest Just in case you were not aware I wanted to highlight that you can now get a free 7 day trial of Horseracebase here.
    We have a lot of members who are existing users of Horseracebase so help is always available if needed, as well as dedicated section of the fourm here.
    Best Wishes
    AR

What are the best ratings for soccer ?

This just using elo ? 42% strike rate ? This I take it's home draw away markets pre ko ?

There are many many more opportunities inplay and getting higher prices for teams that were odds on. Do you program ?? Python and so on ?

I do some programming but I don't know python - the one I showed you is java.

Those years then I was using a so called "IQ method" that is some clever way of averaging the experts opinions (n.b. I do that for horses if you have seen my "press synthesis" threads). But for fixed odds footie I can't do that really - fixed odds tips come in different format, one "expert" wants me to put my money on some French team, the other on some German team and so on. How to compare ?
But with ELO also Greece will come out high, it's more predictable than PL that is.
It does n't means you get paid though, the bookie price lists are full of 1.10s and 1.20s.
Turkey statistics were similar to Greek, but I think Turkey has changed - they play very hard nowadays.

Roughly the scales of predictability from high to low used to be Greece - Italy - England - Germany.
 
Last edited:
.... recently I got pointed towards Cgmbet which has Elo ratings along with a host of other stats. They have a one off charge but also a trial period, you could have a look there it is fairly impressive as I have current membership. ;)
 
Last edited:
I found huge database of teams and results in calendar order from 2007 and converted it to text document.
It does n't have the ratings next to each team but I guess I can recompute those using Arpad's formula -or whatever other formula- and see if I can produce a better figure.
I can share those.
 
Last edited:
For my earlier work I used some recent elo charts from clubelo.com.
Now I can reconstruct the formula of Arpad Elo. It's certain that a recalibrated formula will give better results, with or without the retrospective corrections I mentioned previously. Also the sample is big so it should be ok w.r. stability.
The question now is who makes all this into a website and compute ratings every day for all of Europe ?
 
cosmicsports cosmicsports don't want to dampen your enthusiasm for elo if they were that good we would all be using them. It's a starting point I guess..

No I 'm not overenthusiastic.
But if some really thorough work is to be done, it can only be done the way I describe.
I also know it's a starting point and I explained the reasons in the other thread about Steel ratings.

Also a common advise from many old hands is "don't bet foreign matches - what do you know about those countries ?".
Now the PL is not really foreign to me and Italy is also not foreign to the "Italia fans", but you know what I mean.
 
No I 'm not overenthusiastic.
But if some really thorough work is to be done, it can only be done the way I describe.
I also know it's a starting point and I explained the reasons in the other thread about Steel ratings.

Also a common advise from many old hands is "don't bet foreign matches - what do you know about those countries ?".
Now the PL is not really foreign to me and Italy is also not foreign to the "Italia fans", but you know what I mean.
We are poles apart I just use numbers and make good money and in 2% who make it from football.. all my stuff is inplay these days.
 
..... have a look at the site I mentioned above cmgbet.com and see if there is anything there which interests you. There are lots of stats just waiting for the likes of you, nothing to lose by looking! :D
 
..... have a look at the site I mentioned above cmgbet.com and see if there is anything there which interests you. There are lots of stats just waiting for the likes of you, nothing to lose by looking! :D

I will.
One of my historical misfortunes was in 2001 when Panathinaikos beat Hamburg away (2-0). It was for me a certainty and the price was 5 to 1. But I could n't place the bet in one of the internet companies. No credit card, no Paypal in Greece yet, no paysafe cards and the like, did n't know any of the illegal bookies either. The home company required a treble. So I choose a Scottish match and perform dutching. I needed one more and I did n't want to do more dutching and lower the price. So I choose Liverpool v. Boavista. The Liverpool squad included Owen and Fowler and it was the first leg as well. Normally a 10-0 result is expected. But they drew 0-0 ! And Liverpool went on to qualify in Portugal !
 
.... recently I got pointed towards Cgmbet which has Elo ratings along with a host of other stats. They have a one off charge but also a trial period, you could have a look there it is fairly impressive as I have current membership. ;)

cgmbet is the tool developed by the punterslounge people.
Now that I saw it I remembered.
Impressive, stylish, meticulous programming.
But again those guys do not provide us with data re. success rates.
 
I found a very big bookie database.
It's from opap.gr and contains 389911 matches, the odds and the results.
Someone was updating this file from April 2000 to September 2019. Then he stopped doing it but the number Is big enough.

I converted the odds to probabilities using the known formula and I obtained the following results:

overall predictivity = 36.9%
percentage of homes when the first chance is a home = 52.2%
percentage of aways when the first chance is an away = 48.9%
percentage of draws when the first chance is a draw = 42.4% (* only 609 cases)

My figure of 38.5% predictivity is considerably higher.
I cannot possibly expect to have any effect by playing against OPAP because their percentage take is too big (currently 11-12% as opposed to 5% of the internet companies), but it looks like I have beaten them in terms of predictivity.

The OPAP organization odds makers used to be Ladbrokes and nowadays they are Ladbrokes trained people, according to my information.

Predictivity has nothing to do with the company's take percentage, as some people erroneously believe, because the whatever odds are converted to probabilities (a bookmaker when seen as prognosticator is as good as any other - a big take percentage means that he just is n't paying you fair).

I asked someone and he said he has a similar file from pinnacle and pinnacle's predictivity computes somewhat higher than OPAP.
 
Last edited:
..... really looks impressive and you look deeply interested. I wonder how many on here can plunder the depths of your thinking and come up with suitable answers. Patience and dedication seems to be the order of the day here. Must admit I like something lighter to deal with, but each to his own. ;)
 
..... really looks impressive and you look deeply interested. I wonder how many on here can plunder the depths of your thinking and come up with suitable answers. Patience and dedication seems to be the order of the day here. Must admit I like something lighter to deal with, but each to his own. ;)


Oh well, those results files I 'm talking about have been lurking about for some time really.
The answer may be that there is no answer or that we can never go far enough.
But my conclusion is that to handle things the way we want (maybe with your favourite ratings or some adaptation) what we need is our own ratings website that updates every hour.
I 'm going to talk to some programming gurus to see how this can be done.
 
Oh well, those results files I 'm talking about have been lurking about for some time really.
The answer may be that there is no answer or that we can never go far enough.
But my conclusion is that to handle things the way we want (maybe with your favourite ratings or some adaptation) what we need is our own ratings website that updates every hour.
I 'm going to talk to some programming gurus to see how this can be done.
.... keep in touch and let me know how thing are going. I'm always interested in systems and their workings! ;)

ok.
company working for opap:

 
..... just had a look there and something wrong somewhere I think as the league stats are not updated. They have Sunderland in the Premier League for example. Pity because it looks useful otherwise. ;)
 
I 'm not familiar with this area of automation programming (and cannot go to a library either these days - fiddlesticks).
But there is this guy who does the races: turfpronos.fr.
It's a press synthesis site for the French races and it works by itself.
All his tables are created by a spider program gradually from 4 pm in the afternoon till 10 pm. At 10 pm sharpish they are completed.
Also this happens: Sometimes the other sites he is watching display their selections in some subpage and not in the page where the selections appear normally. So in those cases he makes no entry, it is left empty.
So a computer program could be constructed to take all the livescores in as they come and a second computer program update the ratings.
 
I 'm not familiar with this area of automation programming (and cannot go to a library either these days - fiddlesticks).
But there is this guy who does the races: turfpronos.fr.
It's a press synthesis site for the French races and it works by itself.
All his tables are created by a spider program gradually from 4 pm in the afternoon till 10 pm. At 10 pm sharpish they are completed.
Also this happens: Sometimes the other sites he is watching display their selections in some subpage and not in the page where the selections appear normally. So in those cases he makes no entry, it is left empty.
So a computer program could be constructed to take all the livescores in as they come and a second computer program update the ratings.
That would be a nice little program live up to date rating..
 
Back
Top