/sci/ - Science & Math

Probably around 12C11x0.3^11x0.7

Depends how many times teams draw really, also can he predict draws?

100%

Since he is psychic, 0%.

0.0029296875 or 0.3 % about 1 in 341


asumming there are 2 teams in each game and that they each have a 50% chance of winning.

where the fuck is this Paul, and where are you getting this shit?

>>1357306
>>1357311
Faggots that apply math before common sense.
Inb4 math is common sense

>>1357315

http://en.wikipedia.org/wiki/Paul_the_Octopus

Ohai u oldfags, newfags, whateverfags. check out this new ca/m/wh0/re site full of russians: http://u5l.de/jb Best is, it is all 4free

>>1573

>>1357316
Math is most definitely not common sense.

>>1357316
what the fuck does that mean? do you have some reason why probability doesnt work?

Ohai u oldfags, newfags, whateverfags. check out this new ca/m/wh0/re site full of russians: http://u5l.de/jb Best is, it is all 4free

>>5658

>>1357324
>how likely it is that it was all just coincidence
It's a squid, it does not know football. It's a coincidence, regardless of any probability.
IE. probability it's a coincidence <> probability of him being correct 11 out of 12 times.

>>1357324

If he picks the better team of the two then he is more likely of being right.

Also, some games can be draws and some cannot.

>>1357330
>squid

Ohai u oldfags, newfags, whateverfags. check out this new ca/m/wh0/re site full of russians: http://u5l.de/jb Best is, it is all 4free

>>7220

>>1357330
>It's a coincidence, regardless of any probability.

Retard level: maximum

>>1357333
yeah i assume he cant process information about which team is better.

therefore his picks are basically coin flips.

therefore it is correct.

>>1357330

this anon is correct

I always thought squids were more basketball fans than football fans

>>1357346
>squid

Ohai u oldfags, newfags, whateverfags. check out this new ca/m/wh0/re site full of russians: http://u5l.de/jb Best is, it is all 4free

>>3422

>>1357343

only 6 of the 17 games could be draws, and he can't pick a draw. You didn't factor this into your calculation.

>>1357333
Bullshit ! No draws in World Cup

Ohai u oldfags, newfags, whateverfags. check out this new ca/m/wh0/re site full of russians: http://u5l.de/jb Best is, it is all 4free

>>5717

CoolOctopus.jpeg

>>1357350

>6 of 12

>>1357359

CONGRATULATIONS.

You found a typo on the internet. Well done you.

Ohai u oldfags, newfags, whateverfags. check out this new ca/m/wh0/re site full of russians: http://u5l.de/jb Best is, it is all 4free

>>5584

Germany is a good team, therefore it wins almost always
Paul almost always predicts Germany wins (probably likes the color)
Therefore Pauls predictions are right most of the time
Tl;dr: everything is alright

Ohai u oldfags, newfags, whateverfags. check out this new ca/m/wh0/re site full of russians: http://u5l.de/jb Best is, it is all 4free

>>4988

>>1357299
1/4096, 1/8192 to guess the 12th. and so on.

>>1357380

Not fancy reading other posts before making an idiot out of yourself then namefag?

Ohai u oldfags, newfags, whateverfags. check out this new ca/m/wh0/re site full of russians: http://u5l.de/jb Best is, it is all 4free

>>3689

>>1357382
i'm just pointing out the probability of him guessing 12 games out of sheer luck.

>>1357387

6 of the 12 games could have been draws.
as has been stated in this thread more than once.

Ohai u oldfags, newfags, whateverfags. check out this new ca/m/wh0/re site full of russians: http://u5l.de/jb Best is, it is all 4free

>>4965

>>1357391
this justs modifies the calculus from 1/2^n to 1/3^n for these matches.

that changes however if the octopus could eventually choose not to osition itself on either basket (thus marking a draw, probably).

>>1357399
I fucking hate tripfags.

Ohai u oldfags, newfags, whateverfags. check out this new ca/m/wh0/re site full of russians: http://u5l.de/jb Best is, it is all 4free

>>2628

>>1357371

This is almost certainly the answer.

Too bad for Paul that Germany got fuck up the ass yesterday. He's headed for the pot now.

>>1357402
1/46656 btw

>>1357405
i heard he actually pointed out spain would win the match, and he guessed right?

>>1357371

Actually paul picked serbia over germany (and was right) and he also picked spain over germany and was right again. Also, whoever said that the octopus couldn't determine which team is better and so each pick is essentially 50/50 is absolutely correct

Ohai u oldfags, newfags, whateverfags. check out this new ca/m/wh0/re site full of russians: http://u5l.de/jb Best is, it is all 4free

>>2284

Ohai u oldfags, newfags, whateverfags. check out this new ca/m/wh0/re site full of russians: http://u5l.de/jb Best is, it is all 4free

>>5806

>>1357413
i can think of the fact the colors on these two flags were warmer than those on germany's flag. that might be a reason for the octopus to choose them over germany.

The teams must be aware of Paul's predictions. This most likely affects their morale for the coming match.

Ohai u oldfags, newfags, whateverfags. check out this new ca/m/wh0/re site full of russians: http://u5l.de/jb Best is, it is all 4free

>>5560

Ohai u oldfags, newfags, whateverfags. check out this new ca/m/wh0/re site full of russians: http://u5l.de/jb Best is, it is all 4free

>>2312

>>1357424
>>1357424

Serbias flag is red white and blue which is definitely not warmer than black yellow and red

>>1357449
red and white is actually warmer than red and black.

what other flags had blue in there?

>>1357424
I doubt he has even seen the flags.

>>1357413
he choose germany 10 out of 12 times
obviously it's not a 50/50 pick

>>1357478

he picked germany 9 out of 12 times

ftfy

FFS HOW HARD IT IS TO UNDERSTAND!? PAUL DOESNT PREDICT GAME RESULTS HE CAUSES HIS FAV TEAM TO WIN.

>>1357456

No it isn't stop spreading your shit theories in this thread and the other

>>1357494

Yes, that's because he's predicting the future, and most of the time Germany wins.

>>1357507
>Yes, that's because he's predicting the future
yes, that's the obvious choice why he just likes the German flag more than the others (notice he picked Spain twice also)

>implying they are not doing it over and over until the octopus picks whatever they want and then show it

get a brain

>>1357590
And? What is the point?

The prediction still happens before the match.

>>1357590
the prediction are done publicly in an aquarium in front of everyone. Not in a secret room.

Who the fuck cares.

>>1357590

His predictions are broadcast live.

Is the octopus monitored 24/7 by an independent team of researchers? If not then the simplest explanation is that it's taught to choose the team that is most likely to win based on the predictions of the people training it.

Call me when it's capable of actually predicting the accurate results.

>>1357675
Or more simply, they put his favorite food in the chosen box.

But it only displace the problem, anyway. How does the one who put the food would be able to guess the outcome of the match with a 11/12 accuracy?

>>1357675

The odds were heavily on Germany to beat Serbia in the group stage and slightly in Argentina's favour to beat Germany in the quarterfinals, yet Paul chose the underdog in both of these matches. No person would have chosen Serbia to beat Germany in that second match. Obviously only a psychic octopus was capable of doing so.

>>1357817
>It is unlikely that a specific person wins the lottery, therefore nobody has ever won the lottery

>>1357831
horrible analogy is horrible.

there's only one Paul

>>1357426
This definitely has an effect on their playing.
Self fulfilling prophecy.

>>1357675
>>1357675

>implying servia was very likely to win germany

>>1357864
see
>>1357857

it's a 50% chance, either he guesses them all correctly or he doesn't

>>1357892
i like the way you think

>You either win the lottery or you don't, therefore there's a 50% chance you'll win the lottery

>>1357892
You either win the lottery or you don't, therefore there's a 50% chance you'll win the lottery.

How does your precious science explains that, faggots?

Paul is my hero

>>1357892
then it is either 100% or 0%
how can it be 50%?

>>1358556

0/10

  ▲
▲ ▲

You guys are missing the point
Paul is psychic
He's the next step in octopus evolution

>>1357295
Ohai u oldfags, newfags, whateverfags. check out this new ca/m/wh0/re site full of russians: http://u5l.de/jb Best is, it is all 4free>>3422
Ｗｗ_Ｗ.Ａ_Ｎo_n[iｇnＯＲｅＴｈIｓHＥRE]ＴＡ_ＬＫ .Se
xtrre yzvj jv e g r fsbis qussnz b qy q

Across the world there will be hundreds of "psychic" animals, thousands of human "psychics", people guessing and pulling names out of hats and looking at goats entrails etc. How many of them were wrong compared to the few that were right? But all the ones who were wrong get forgotten, while the one here that was (mostly) right gets shown on national news.

It's conformation bias. Add to the fact that he seems to choose germany more often, and germany win more often due to being a good team, and it isn't really that amazing even considered in isolation

well, predicting 12 out of 12 is 1 in 144.

so, predicting 11 out of 12 is 1 in 132.

not that impressive of odds, once you think about all the other people that are probably doing crazy things to predict the world cup.

here's a capitalist idea: given that this is probably a victim of the law of large numbers, but some people believe it.

you should bet them, say 20 that the team opposite the team that the octopus predicted would win, and 10 if the octopus is correct. you can then cover your ass by betting 20 on a match with somebody else. this way, if the match is considered to be 50/50, you have 5 profit average

only problem would someone defaulting on the bet. of course the match could not be 50/50

>>1358618
>Paul resides in Sea Life Oberhausen, in the town of Oberhausen, Germany. Paul's name was taken from the title of a poem by the German children's writer Boy Lornsen: Der Tintenfisch Paul Oktopus.[9] According to statements from Sea Life's entertainment director, Daniel Fey, Paul demonstrated his intelligence early in life:
>"There was something about the way he looked at our visitors when they came close to the tank. It was so unusual, so we tried to find out what his special talents were."

>>1358629
maybe he chooses germany because of his german blood?

>>1358641

12 out of 12 with a two possibilities choice is 2^12 = 4096.

I work out the chances of getting 10 or more out of 12 50-50 chances as 4096/145 = 28 to one

So with 28 octopuses, on average one would do as well as paul has done.

Hardly impressive considering the thousands of attempts at prediction that must have been going on around the world.

Still, I think paul is a real cool guy

I think Paul has predicting accurate scores as well. Crunch the numbers on that.

Just think out loud here.
Octopii are incredibly intelligent animals.

This all relies on the assumption that Paul enjoys this experience of being placed in front of boxes, and that no one is maliciously tampering with the experiment.

Now, Paul only undergoes the guessing game when Germany is playing. If he's guessing for a tournament, and Germany loses, he's done with the guessing game.

So all he has to do is pick germany most of the time, and he'll be right most of the time, not because of the skills of any of the teams, but because every trial except the last one depends on a german participation and win anyway.

none of the formula presented thus far have incorporated the fact that N number of games is dependent on X number of german wins, have they?

>>1358686

correct me if i'm wrong

>>1358750

That's not even wrong

>>1358750

oh, sorry. forgot to explain that i think that 12 is too long to do manually to prove, so i just did 4

you can do 3 for your own enjoyment.

it's (number of games)^(2)

>>1358768

Do you think the odds of guessing a coin flip correctly 12 times in a row is 1/144?

somebody should make Paul predict the stock market

>>1358775

good point, i concede

>>1358721
wrong here, I think its actually 4096/79 = a 1 in 52 chance.

The same as flipping a coin 12 times and getting 11 heads and 1 tails.
Getting 11 heads and then 1 tails: 0.5^(11) * 0.5^(1)
Getting 10 heads, 1 tails, then 1 heads: 0.5^(10) * 0.5^(1) * 0.5^(1)
etc.
Because there are 12 different possible positions for the tails flip, the actual probability would be:

0.5^(11) * 0.5^(1) * 12 = 0.5^(12) * 12 = 0.29297%

Very lucky octopus

NOT a single explanation here is adequate. Let me explain.

Say one person believed that Paul was really a psychic. How would you disprove his belief and show him he is wrong

Answering with "oh, its just a coincidence" will not convince anyone.

Bear in mind, from the perspective of the octopus, assuming no manipulation, the choice is 50/50 or 0.5. Since he got a sequence of 11 correct, it becomes 0.5^11. A VERY low number. Even if it wasnt a sequence of 11 its still very impressive.

Neither will "He was manipulated to choose team A over team B by putting diff food in diff boxes" etc as that would mean whoever manipulated the octopus got the predictions correct on very low odds. Including likes of Serbia winning over Germany. No person would have bet on that. ON TOP OF THAT, surely as "scientists" you realise that you actually have to provide proof that manipulation through placing of diff food items occurred.

You cant just say "Oh, probs different food was placed in boxes" and settle for that explanation without showing that diff foods were placed in there. But again, that brings up the issue of whoever placed the food there gettign the predictions right.


tl;dr, no explanation provided thus far could pass as an adequate scientific explanation. They're are just dismissive arguments from the default position of "lol psychics dont exist". Whi I agree with, I dont believe in them, but saying "lol psychics dont exist" is as valid of a statement as saying "lol, psychics do exist" if you cant prove your statement,.

>>1358835

lol you are fucking mad son

you are also wrong.

you disregarded the possibility of draws in you calculation.

your explanation was inadequate.

Also,

>implying anyone thinks he's psychic

>>1358835

To add. If you did a hypothesis test at 1% S.L. The null hypothesis being Paul isnt psychic, alt hypothesis being he is. If you crunched the numbers as >>1358827 did for example you would actually have to reject the null and accept the alt hypothesis that Paul is psychic.

Heck, even at 0.5% S.L. you would have to accept the alt hypothesis

>>1358835
>NOT a single explanation here is adequate.

Except for >>1358629 and >>1358721, that is. It is simple - around the world there are hundreds of things like this going on. A few like paul, are correct, and become national news. The many that are incorrect are forgotten.

And btw he got 2 wrong out of 12, not one

>>1358842

Germany didnt draw though. If he predicted a win or less and Germany drew it would have counted as Paul being wrong. But he wasnt, he was right 11 times.

I offered no explanation, I said that none of the explanations in here are particularly solid. Yes, there are people who think he is psychic.

This is why science is shit. Dumb people use it and get rid of all common sense.

>>1358871

A believer would retort that because most of these people/predictors are not psychic. So of course they will be incorrect. Paul is psychic thats why he is correct.


Plus he was in the media from 2 years ago. And he got lot of the predictions for the Euro right there. He was being observed from the beginning of this world cup, because people thought of him as "psychic". Its not that he was starting to get preidctions right as the WC progresses and suddenly media took interest in him.

>>1358874

He doesn't pick Germany to win or less you dick.
He picks the winning team. As in when he chose Serbia to win and they did he was correct.

Your calculation is wrong and if you cannot see this there is no hope for you.

lrn2statistics, /sci/

p-value < 0.0005

Holy shit.

At least, it would be amazing if he were not picking flags. The pickings were not independent. If flags were reused, he could've preferred one flag over others. Even if flags were not reused, perhaps those countries with certain patterns or colors on their flags, for some reason, were more likely to have better football teams; say, certain cultures may produce flags octopuses like more AND better football teams.

>>1358926
In other words, a fair experiment would be to use random colors for any country, and randomize them again if a country is reused.

>>1358901

If you look at the wikipedia page the euro results were him picking germany all the time, even when they lost. I am sure there are many other similar things in the media of countries round the world that seemed ok at first, then turned out to be wrong.

As for the beliver point, well, lets say you got a hundred people to flip coins to see who would win. One or two might get it right every time, with the others getting it wrong. A believer could say that the one or two were psychic, while the other 90+ were not. Would this convince you?

The Wikipedia page says he got it right 10 out of 12 times. Is it wrong?

ITT: People not realizing that picking any team every time in a tournament will result in N or N-1 correct picks.

Some random neuron firing probably made him responsible for not picking Germany.

>>1358959
>wikipedia wrong?

>>1358926
Or you can just accept that we will soon have starships with the flightdecks submerged and an octopussy who navigates through subspace with the power of it's awesome mind.

>>1358968
If it's wrong, I'm hoping someone here will know which entry is wrong and will be able to correct it.

>>1358961
>>1358961

>implying Paul (PBUH) picked Germany every time
>implying if Paul picked Germany every time and Germany lost all three group games Paul (PBUH) would have guessed correctly twice

>>1358926
Oops I messed that up.

p-value = ~0.005

>>1358912

They used him only for games that Germany played. The choice for winning team was either Germany or the other team. In other words the choice was;

Germany wins. Germany loses. Draw wasnt a choice.

Again, youre not explaining anything. Youre saying LOL DICK UR RONG!.

This was my point. No one thus far has provided a solid and convincing explanation.

IF my calculations are wrong then demonstrate how they are wrong.

tbh I dont even remember making a calculation.

>>1359043

His chances of guessing correctly in the group games was 1 in 3.

>>1358939

See, this Anon is good. Yeah I dont need convincing myself. The 100 people flipping coins analogy was one of the first things that I thought of myself.

But again, that would not convince a believer, who would retort as you say, the ones that got it right were psychic, unlike the others.

An extensive analyses would be better. The fact that the octopus picks Germany predominantly over others indicates something (perhaps about the flag).

The two predictions that Germany would lose though, once against Serbia and once against Spain is worth exploring I reckon. (The flag theory isnt as solid on this one since last time it predicted Germany would beat spain but it didnt)

>>1359050

So..? That just shows that his chances of getting a correct answer were even LOWER, but he still got it right....

>>1359084

It means your original calculation was wrong.
>the choice is 50/50 or 0.5. Since he got a sequence of 11 correct, it becomes 0.5^11. A VERY low number.

I'm going to assume the Wikipedia page is right since it gives references for its data, and that the OP is wrong, and thus that Paul got it right 10 out of 12 times.

Hypothesis 1:

Paul picks randomly. The probability of the observed picks is <span class="math">1/2^{12} \approx 2.44 * 10^{-4}[/spoiler].

Hypothesis 2:

Paul is psychic, but has limited accuracy. He picks the right team with probability p, with <span class="math">p \in [0.5, 1][/spoiler]. Assume a uniform prior on p. The probability of the observed picks is <span class="math">2 \int_{0.5}^1 p^{10} (1-p)^2 dp = 675/292864 \approx 2.30 * 10^{-3}[/spoiler].

Hypothesis 3:

Paul picks Germany with probability p. Assume <span class="math">p \in [0, 1][/spoiler] with a uniform prior on p. The probability of the observed picks is <span class="math">\int_0^1 p^{10} (1-p)^2 dp = 1/858 \approx 1.17 * 10^{-3}[/spoiler].

If we give the same prior probability to each of these three hypotheses, then the probability Paul is psychic is about 62%. However, that would be retarded. It ignores the evidence we have previously accumulated that psychics aren't real, and that octopi don't know anything about football. I'm not going to try to quantify that evidence, but the ratio between the probabilities of hypotheses 2 and 3 is going to have to get a lot larger than 2 before I start believing.

>>1359108

i loled

>>1359108
Hypothesis 4: FIFA is refball.
Probability = 1.

>>1359108
Hypothesis 1 should be multiplied by 24. There are 12 different ways that one could get 1 of the guesses wrong. And since we are testing a hypothesis that the guesses are random, we should include the probability that he got 11/12 WRONG by multiplying our value by two.

More importantly, your conclusions are philosophically problematic. There are also an infinite number of alternative hypotheses that would account for any given random result better than the null hypothesis (it's like a flip of a coin) would.

If he got 6/12 right, for example, one hypothesis could be that an extraterrestrial is controlling his results to make him guess the same number correct as incorrect. This hypothesis would predict 6/12 precisely, better than the null coincidence hypothesis would.

>>1359107

Ok, but thats irrelevant. With the new calculations the probability will be even lower. That doesnt really support the "its pure luck argument

▲
▲ ▲

>>1359142
>Hypothesis 1 should be multiplied by 24. There are 12 different ways that one could get 1 of the guesses wrong...

The analysis is using Bayesian methods. In Bayes' theorem, you use the probability of the observed outcome, not the probability of the observed outcome and all the other outcomes that would be just as unusual.

>More importantly, your conclusions are philosophically problematic. There are also an infinite number of alternative hypotheses that would account for any given random result better than the null hypothesis (it's like a flip of a coin) would.

That applies to Bayesian methods in general, and my response would be that you should handle this by giving low prior probabilities to complex hypotheses (e.g. alien control leading to a 6/12 result exactly) by Occam's razor. Thus I would agree that giving equal weight to hypothesis 2 is wrong. This is somewhat subjective, but I think some degree of subjectivity is unavoidable.

To put it in terms /sp/ can understand:

Paul picked correctly 10 out of 12 times.
Paul picked Germany 10 out of 12 times.

There is no more evidence Paul is psychic than there is that Paul likes the German flag.

>>1359187
the flag faces the camera, not paul.

>>1358871
>>1358871

You are fucking retarded. No ones arguing that psychic shit is real, we're saying that this octopus is a fucking badass.

>>1359166

We have smart anons here?

>>1359166
It's far too subjective IMO. You're trying to give probabilities of truth, rather than giving exact probabilities of certain (ranges of) outcomes according to certain models like null hypotheses.

>>1357299
To find out, we need to know the probability that Paul picks Germany even when there's not a football game.