Announcements
Double Prize Money for NAPS Comp & Poker League in May! AGAIN!
giraldi

CgmBet System Software - (Exclusive PL Registered Forum Users 50% Discount Code)

Recommended Posts

Looks like an impressive piece of sosftware.
What is the predictivity value it scores though ?

Read about predictivity values and benchmarks here:
 



I don't mean to be critical but I am critical of everything, because our task (to beat the bookies) is difficult, like the legendary hunting for the snark.
It depends on this:
Who are our opponents ?
Let me tell you a story.
Summer of 1986, Mexico world cup, early group stage. I go to a large department store for shopping, in Athens-Greece. At the exit they were giving the customers a piece of paper saying "predict todays match and enter our great competition". The prize was a motorbike and the match to guess was Costa Rica v. Scotland. So I thought "everyone will go for Scotland here, they are respected as a European force and they recently beat Greece in the preliminaries, but they have some problems in the scoring department and the Costa Ricans will defend heroically". So I ticked the draw box. 
The match ended in a goalless draw so next morning I went to the shop to see what happened. What happened was that I was one of only eight who guessed correctly.
Did n't win the bike but one among eight was impressive - the number of customers the previous day must have been 1000 and more, it was summer sales period as well.
But these people were housewives, girls, grannies and so on. They did n't know about football.
So my opponents were weak. If I was to bet against them everyday I would surely make lots of easy money !
In the real world of betting it can never be like that. And indeed after I left the shop when I entered the competition, before going home I went to some football agent shops out of curiosity to see what they thought about Costa Rica versus Scotland. There it was different. The predictions were evenly matched between Scotland and the draw, with some 15% going for Costa Rica to win.
So this proves to you then that it is all about the opposition. We want to outclass the bookies like I did with the customers of that department store !
But this is an uphill struggle and although the efforts of the statisticians and software makers are commendable, they should be able to beat some quite difficult benchamrks in order to have any impact.

Edited by zabadac

Share this post


Link to post
Share on other sites

Thank you Froment for your intervention. You are right, my software is just a tool not a prediction software. 

It is just a paint brush. Not even an excellent quality paint brush can quarantee that if you will be the next Van Gogh but for sure will help those that have talent and work hard to sell more paintings than those that paint with low quality colored pencils:))

Hi Zabadac, interesting comments. You contradict yourself a little bit when said "to beat the bookies". After that, you are right in my opinion, we need to be more informed, equiped ...etc than the other players not than the bookies (this will be never possible - this would be the end of the betting industry anyway). 

Bookies just want as many customers as possible  and really  don-t care how you bet or how much you win as long as your won bets are paid from the other users account, users that are not using a god quality paint brush  :))

 

 

Edited by giraldi

Share this post


Link to post
Share on other sites
48 minutes ago, giraldi said:

Thank you Froment for your intervention. You are right, my software is just a tool not a prediction software. 

It is just a paint brush. Not even an excellent quality paint brush can quarantee that if you will be the next Van Gogh but for sure will help those that have talent and work hard to sell more paintings than those that paint with low quality colored pencils:))

Hi Zabadac, interesting comments. You contradict yourself a little bit when said "to beat the bookies". After that, you are right in my opinion, we need to be more informed, equiped ...etc than the other players not than the bookies (this will be never possible - this would be the end of the betting industry anyway). 

Bookies just want as many customers as possible  and really  don-t care how you bet or how much you win as long as your won bets are paid from the other users account, users that are not using a god quality paint brush  :))

 

 

 

Re. my story from 1986, there were no bookies then -at least we did n't have them.
And when I said "I went to the betting agents to see ..." I really mean the pools agents, a Greek tote game.
But from the various pools systems they had for sale, I could of course see what their thoughts were with respect to that particular match, Scotland v. Costa Rica.

Really I would prefer tote rather than bookies.
The exchanges are closer to tote, but in any case we are stuck with the bookies.
Those fellows then have engaged the services of some famous universities to help them with prediction theories, in case it has escaped you.
Nottingham-Belfast-Stanford-Catagna to name but a few.
Oxcam too maybe, only I have n't come across any Oxcam papers on footie prediction theory.

In sharp contrast all the -rich in other respects- Greek organization in the eighties had in terms of  "prediction theory" as I recall, was a table they were obliged to prepare every week and use it to draw draw balls, in case some coupon matches were rained off. This table had the numbers 40-30-30 or 40-40-20 everywhere and people were laughing at it. Rumour had it they employed cabaret actresses loyal to the government party (PASOK) to write down those tables,
But they had no reason to become involved, so they did n't care (in those days, now that they offer fixed odds, they do care).

So in real life we make our bets with the bookies, we are trying to beat the bookies. What else ?

Anyway, you say it can go some of the distance, but you don't call yourself a "Van Gogh".
Fair enough, but we need benchmarks in order to work. 
Everyone needs probability theory benchmarks.
Archimedes said "dos moi pa sto kai tan gan kinato", in order to burn the Roman boats.


* In my story the moral is that the draw in the Costa Rica v. Scotland match was going to pay me some 10 to 1, 20 to 1 among the shoppers and
some 2 to 1 with the agents. So I 'd still be a mediocrity but the prices would  be 20 to 1 ! Impossible of course, but in the particular setup it worked like that,

Edited by zabadac

Share this post


Link to post
Share on other sites

Not that I want to remember it, being Scottish and all, but we actually famously lost the match v Costa Rica 1-0, and it was at Italia 90 not Mexico 86.

 

Kinda calls into question the rest of your points, if you can't get those basics right, no?

 

 

Share this post


Link to post
Share on other sites
55 minutes ago, only1woz said:

Not that I want to remember it, being Scottish and all, but we actually famously lost the match v Costa Rica 1-0, and it was at Italia 90 not Mexico 86.

 

Kinda calls into question the rest of your points, if you can't get those basics right, no?

 

 

 

Alright, the department stores "Le Mignon" closed in 1998, so it could have been 1990 and I might have backed Costa Rica to win instead of what I said.
But the rest of the story is true.
As for the football agency, the Costa Rica win probability estimate maybe stood at 20%.
My guess was correct because otherwise why did I go and wait for the motorbike draw ?

It's 30 years now and I don't keep a diary but i have another experience with department store competitions which is even more weird but unrelated with the present conversation.

 

Edited by zabadac

Share this post


Link to post
Share on other sites

Anyway, as I said in the other thread I have reached a predictivity of 38.5% for the major European leagues.
The overall maximum ever observed was 42%, for easy leagues where the home wins were over 50% (Greece of old, Turkey of old).
In what sense can we say "easy leagues" and "difficult leagues" ?
My experience as a football fun is this:
In the 70s-80s I lived sometimes in Greece and sometimes in the UK and I was vsiting the football grounds in both countries.
The Greek matches were easy. When the stronger team played at home and scored the opener, the opposition crumbled. Final result 2-0, 3-0, 4-1 was the typical pattern with few exceptions.
The English matches were different, to my astonishment as a newbie. Small teams like Carlisle-Sunderland could give a hard time to giants like Liverpool-Leeds. Liverpool's magic mainly prevailed ok, but they had to play for their money.
So that's the difference.

But 38.5% predictivity is not bad.
The question is "do I get my money back from this ?".

My present theory is based entirely on elos, which you also have.
Your last image says PSG - Bayern Munich 2126-2135.
This for me translates into 44.2% for the home win, 26.3% for the draw and 29.5% for the away win.
It also says "buy PSG if the price is 2.26 decimal or over", "buy the draw if 3.8 decimal or over", "buy Bayern if 3.39 or over".
Suppose I do this, or declare the match a "no bet" if none of the three prices offered meets my criterion.
What will be the balance after some trials ?
If it is negative it means I 'm not good enough, if it is positive it means I 'm doing well.

I have n't started yet.
I had some other computer program before to the one I have now and it proved unsatisfactory. 
But right now I can't do nothing, because of the postponements of course.

The thing is can there be any further enhancements and does your software provide any clues ?
 


 

 

 

Share this post


Link to post
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.