View previous topic :: View next topic 
Author 
Message 
Rach Site Administrator
Joined: 16 Jul 2005 Posts: 1056 Location: Middle of nowhere in England

Posted: Fri Jul 09, 2010 3:52 pm Post subject: The Best World Cup Pundit? 


Personally one of my favourite stories about the World Cup is Paul the psychic octopus. Originally from Weymouth, but now living in Germany the octopus has correctly guessed the outcome of all of Germany's world cup fixtures so far this world cup (causing panic when he predicted Spain to beat Germany in the semifinal). For the final he has predicted Spain to beat Holland.
http://www.telegraph.co.uk/sport/football/worldcup2010/teams/spain/7881110/WorldCupfinalPaulthepsychicoctopuspredictsSpainwillbeatHolland.html
So far this world cup he has got 6/6 predictions correct (all of Germany's fixtures). How does he do it?!?! Surely its highly unlikely to get 6/6 guesses right? They should feature him on the TV coverage, as I am sure he gueses the outcome correctly more often than the actual pundits. _________________ ~~~I don't suffer from insanity, I enjoy every minuite of it!~~~


Back to top 


Lamiaceae Site Administrator
Joined: 05 Jul 2006 Posts: 7651 Location: To the right of my computer


Back to top 


area51newmexico Goddess
Joined: 08 Jul 2005 Posts: 10598 Location: East Yorkshire, England

Posted: Sun Jul 11, 2010 8:31 am Post subject: 


How clever! It's pretty cool, I wonder how he does it?!
I'm not interested in football but I am interested to see who wins just to see how littlemultiarmedPaul does _________________ Helen, the Administratrix of www.area51newmexico.com


Back to top 


Rach Site Administrator
Joined: 16 Jul 2005 Posts: 1056 Location: Middle of nowhere in England

Posted: Mon Jul 12, 2010 2:18 pm Post subject: 


Surely the probability of him getting all the predictions right are miniscule (its not just a case of 50% for this match and 50% for that  the probability of him getting both right are less than 50%, only 1/4 or 25% surely). _________________ ~~~I don't suffer from insanity, I enjoy every minuite of it!~~~


Back to top 


Specter Forum Champion
Joined: 30 Apr 2006 Posts: 1698

Posted: Sat Jul 17, 2010 8:13 am Post subject: 


Gamblers fallacy, typically used in roulette. The probability of any event is invariable based on the outcome of previous events, if those events are independent.
If you were to go back to the start of the season and attempt to predict the probability of him picking all the winners it would be .5 times the number of games. But each game has a 5050 chance.
ie. it doesn't matter how many times red has come up, its still a 5050 on black. _________________


Back to top 


Rach Site Administrator
Joined: 16 Jul 2005 Posts: 1056 Location: Middle of nowhere in England

Posted: Sat Jul 17, 2010 4:30 pm Post subject: 


But surely if you flip a coin, its 50:50 heads or tails, but if you flip the same coin twice the probability of 2 heads is 25%, 2 tails 25% and one of each 50%  doesn't it work the same way in football  when predicting 2 matches, 25% get both right, 25% get both wrong, 25% get 1st right and 2nd wrong, 25% get 1st wrong and 2nd right? _________________ ~~~I don't suffer from insanity, I enjoy every minuite of it!~~~


Back to top 


Specter Forum Champion
Joined: 30 Apr 2006 Posts: 1698

Posted: Sun Jul 18, 2010 3:04 am Post subject: 


Before you flip the coin at all the probability of getting 2 heads in a row is .25
If you flip the coin and get a heads, the probability of getting heads on the next flip is still .5 _________________


Back to top 


Lamiaceae Site Administrator
Joined: 05 Jul 2006 Posts: 7651 Location: To the right of my computer

Posted: Mon Jul 19, 2010 11:50 am Post subject: 


Specter wrote: 
Gamblers fallacy, typically used in roulette. The probability of any event is invariable based on the outcome of previous events, if those events are independent.

Oh how true!  I hate people at roulette who *still* believe this! this guy I go gambling with stuck to this for ages. After months of telling him what an idiot he was ( complete with long email dissertations explaining why he was stupid to believe in such a thing ) he finally came round to understanding me.
I still give him an accusing look before we hit the roulette tables though just in case he should slip up and make some comment.
A website I used to like reading was Wizard Of Odds, i'm glad to see it's still about.
http://wizardofodds.com/gambling/bettingsystems.html _________________ I should update my sig. What to put here for $CurrentYear ? 

Back to top 


AVARiCE Lowering the Tone Since 2005
Joined: 23 Sep 2005 Posts: 5780 Location: London, England

Posted: Thu Jul 22, 2010 1:25 am Post subject: 


Specter wrote:  Gamblers fallacy, typically used in roulette. The probability of any event is invariable based on the outcome of previous events, if those events are independent.
If you were to go back to the start of the season and attempt to predict the probability of him picking all the winners it would be .5 times the number of games. But each game has a 5050 chance.
ie. it doesn't matter how many times red has come up, its still a 5050 on black. 
lrn2statistics.
You actually highlighted the bit where you went wrong yourself...
These events are not independant of each other. The only way a team can win the overall tournament relies on them winning every round (ie counting all group matches as a "win" if they get through) in order to progress.
Or, to analogise your way, every time it lands on black a red gets taken off the board.
If each match is 0.5 probability per team (assuming it was a fair probability  which it isn't but I digress) then the odds of winning a game are as much but the overall tournament winners will have a 5 x 0.5 chance of winning.
This, however, assumes that the octopus called the tournament winners in advance, which you seem to be talking about I think? I didn't read the article tbh. Regardless though, the events were neither independant nor fair which makes this slightly impressive.
~ AVARiCE _________________


Back to top 


Specter Forum Champion
Joined: 30 Apr 2006 Posts: 1698

Posted: Tue Jul 27, 2010 1:49 am Post subject: 


The event of picking the winner, assuming the winner of the match is independent of the choice (which is not strictly true, because the choice might effect team moral etc etc but is a convenient simplification), is independent of all the choices which have occurred previously.
We are talking about the octopus picking a winner, not the chance of the team winning. So the situation we analyse is, the octopus picks a box, and if you assume that he doesn't know anything about the teams or the boxes, then its 5050 that he picks the winner (since the winner isn't dependant on his choice as stated above)
I do lrn2statistics in my 3rd year uni Data analysis course. _________________


Back to top 


