Critical Thought Experiments

Nope, 'neer is right. This is a rewording of the Monty Hall problem that fools a lot of mathematicians. It is always correct to change. It works out to 1/3 of the time you lose, 2/3 you win.

Wow, I'll bet you're right, but I still can't see how it works. Can someone explain it to me like I'm six?
 
Wow, I'll bet you're right, but I still can't see how it works. Can someone tell explain it to me like I'm six?

This one is conceptually very tough, because it violates a deep-seated bit of common sense that muddles our ability to correctly assign probabilities in a perturbed system. The hinge of it is, if you hold, you're gaining nothing from the discard. If you switch, you gain something. The discard is not random and is perturbing the outcome.

If you switch:
your first choice has one chance in three if being the winner.
If it is, you'll lose.

If however your first pick is a loser (chances two out of three), the remaining two boxes are one good, one dry. The leprechaun WILL discard the dry one. That's the kicker.
If your initial choice was dry, and you switch, you have a unity chance of getting the prize.
Thus, switching as a matter of routine raises your win chances to 2 out of 3.

The Wikipedia entry is informative imo. cn

https://en.wikipedia.org/wiki/Monty_Hall_problem
 
Ah fuck, i was thinking along the same longs as tyler. It makes enough sense. Because i made my choice when three options were available, and sticking with that choice even though one was removed still leaves me with the same odds.

But like I said, that fucker is supposed to pay up just for catching him.
 
This one is conceptually very tough, because it violates a deep-seated bit of common sense that muddles our ability to correctly assign probabilities in a perturbed system. The hinge of it is, if you hold, you're gaining nothing from the discard. If you switch, you gain something. The discard is not random and is perturbing the outcome.

If you switch:
your first choice has one chance in three if being the winner.
If it is, you'll lose.

If however your first pick is a loser (chances two out of three), the remaining two boxes are one good, one dry. The leprechaun WILL discard the dry one. That's the kicker.
If your initial choice was dry, and you switch, you have a unity chance of getting the prize.
Thus, switching as a matter of routine raises your win chances to 2 out of 3.

The Wikipedia entry is informative imo. cn

https://en.wikipedia.org/wiki/Monty_Hall_problem

Thank you, gentlemen, my mind is blown! This is the graph from your link, Neer, that finally simplified it enough for me to comprehend:

[TABLE="class: wikitable"]
[TR]
[TH]behind Door 1[/TH]
[TH]behind Door 2[/TH]
[TH]behind Door 3[/TH]
[TH]result if staying at door #1[/TH]
[TH]result if switching to the door offered[/TH]
[/TR]
[TR]
[TD]Car[/TD]
[TD]Goat[/TD]
[TD]Goat[/TD]
[TD]Car[/TD]
[TD]Goat[/TD]
[/TR]
[TR]
[TD]Goat[/TD]
[TD]Car[/TD]
[TD]Goat[/TD]
[TD]Goat[/TD]
[TD]Car[/TD]
[/TR]
[TR]
[TD]Goat[/TD]
[TD]Goat[/TD]
[TD]Car[/TD]
[TD]Goat[/TD]
[TD]Car
[/TD]
[/TR]
[/TABLE]
It's amazing that thousand of Ph.Ds could not see this. I guess I was in somewhat good company. Good, stupid company ;)
 
Thank you, gentlemen, my mind is blown! This is the graph from your link, Neer, that finally simplified it enough for me to comprehend:

[TABLE="class: wikitable"]
[TR]
[TH]behind Door 1[/TH]
[TH]behind Door 2[/TH]
[TH]behind Door 3[/TH]
[TH]result if staying at door #1[/TH]
[TH]result if switching to the door offered[/TH]
[/TR]
[TR]
[TD]Car[/TD]
[TD]Goat[/TD]
[TD]Goat[/TD]
[TD]Car[/TD]
[TD]Goat[/TD]
[/TR]
[TR]
[TD]Goat[/TD]
[TD]Car[/TD]
[TD]Goat[/TD]
[TD]Goat[/TD]
[TD]Car[/TD]
[/TR]
[TR]
[TD]Goat[/TD]
[TD]Goat[/TD]
[TD]Car[/TD]
[TD]Goat[/TD]
[TD]Car
[/TD]
[/TR]
[/TABLE]
It's amazing that thousand of Ph.Ds could not see this. I guess I was in somewhat good company. Good, stupid company ;)

Not stupid. I think this is a brilliant puzzle (and I'd +rep Heis for it if I could, but I need to suck off a few more unicorns first ... horny bastards) because it "catches" on a glitch in our ability to think probabilistically. cn
 
Not stupid. I think this is a brilliant puzzle (and I'd +rep Heis for it if I could, but I need to suck off a few more unicorns first ... horny bastards) because it "catches" on a glitch in our ability to think probabilistically. cn

Yeah, I can't rep the guy anymore, either. I haven't been this challenged in years, thanks for the lesson. I love this sub-forum...
 
It's just counter intuitive. Damn preconceived notions! You guys would love this YouTube guy, his channels is called Veritasium. He covers tons of misconceptions that people have about all sorts of things, like where trees get most of their mass from. This guy wrote his phd on how to educate people with videos. He found that just telling someone "trees get most of their mass from air" they just to oh yeah I knew that, and the misconception strengthens. Instead he asks them, and most everyone forgets that plants inhale co2. Good videos if anyone is interested.
http://m.youtube.com/watch?v=RQaW2bFieo8
 
Neer not only gave the right answer, he also saw the lesson behind it. Humans are great at pattern recognition, and terrible at probability assessment. The problem is, because of intuition, we feel we are skilled at judging probability. I am glad this sparked some conversation. It's nice to see people handle disagreement without taking it personally or feeling insulted.

I have not viewed the wiki page, but this is how my mind breaks it down.

When you pick the first box your chances are 1/3. The leprechaun takes one choice away. At this point we are more or less given a new choice, it's a new game. If you stay, you still have 1/3 chance because you are still playing the old game. If you switch, your new choice is governed by 1/2 odds. If during my new choice I pick the same box (essentially stay with my original) I am also staying with the 1/3 odds.

Another way to see it is, the leprechaun is essentially giving you a choice between a single box or a pair. He is saying you can have box 1 (1/3 chance) or you can have box 2 AND box 3. (2/3 chance) The fact that he showed one of the pair is empty offers no new information, you already knew that.
 
Lets see if you learned anything. You meet an old high school friend on the bus. You ask if He has any children, and he says he has two. You ask if he has any daughters, and he says yes. You now know that he has two children and one is female. What are the chances that he has two daughters?

I expect not all will agree on this one.
 
Lets see if you learned anything. You meet an old high school friend on the bus. You ask if He has any children, and he says he has two. You ask if he has any daughters, and he says yes. You now know that he has two children and one is female. What are the chances that he has two daughters?

I expect not all will agree on this one.

Fuck you got me again. Intuition tells me 50/50 boy girl. But clearly that didn't work before for me.

The only thing that makes me think he has TWO females is that you said "has any daughters, and he said yes." Daughters implies more than one, so that means 100% two girls?
 
The variety is such that I can not have inside information about all areas, and the volume is such that it can not be pure chance. At the end of the 6 months do you believe I am a psychic? If I am not physic and it wasn't left to chance, how did I do it?

the part in bold seems to try to semantically preclude the next two parts.

but you could have a variety that seems to dabble in inside information from all areas, and the volume can be limited enough to preclude pure chance.

you are trying to assert an 'a priori' that can be explained by 'a posteriori'.

take that bullshit and go elsewhere.

i said good day sir.

I SAID GOOD DAY!
 
Fuck you got me again. Intuition tells me 50/50 boy girl. But clearly that didn't work before for me.

The only thing that makes me think he has TWO females is that you said "has any daughters, and he said yes." Daughters implies more than one, so that means 100% two girls?

You are right, the question is ambiguous, and it's meant to be. I am attempting show that numbers have their own internal logical consistency which sometimes conflicts with our logic. He may have meant daughters plural, but easily could have meant single. Say your life depended on it, would you think it's safer to assume he meant daughters plural, or forgo that assumption in favor of odds based on what you can know for sure?
 
Ok help me with this guys. I told cannabineers situation to a friend, the island that catches on fire one that I solved. After I tell her that I start another fire on the island and take cover behind the new burn line, she points out that the size of the island was never specified and she assumed that it was small and thus not large enough to do such a task. I tell her the island size was not specified and that if you limited the size of the island to the point where setting a backfire is not safe, then you fail the puzzle.

Is this correct, or was she safe in assuming that the island was too small for that task? I think she failed because she assumed the size of the island. This is where over thinking comes into play, and I'm sure that's what stopped her from solving the problem.
 
You are right, the question is ambiguous, and it's meant to be. I am attempting show that numbers have their own internal logical consistency which sometimes conflicts with our logic. He may have meant daughters plural, but easily could have meant single. Say your life depended on it, would you think it's safer to assume he meant daughters plural, or forgo that assumption in favor of odds based on what you can know for sure?

I assume that he would say daughters if he had two. But we all know what assumptions make...
Also I'm making this assumption off very little evidence, so it is risky anyways.
 
Ok help me with this guys. I told cannabineers situation to a friend, the island that catches on fire one that I solved. After I tell her that I start another fire on the island and take cover behind the new burn line, she points out that the size of the island was never specified and she assumed that it was small and thus not large enough to do such a task. I tell her the island size was not specified and that if you limited the size of the island to the point where setting a backfire is not safe, then you fail the puzzle.

Is this correct, or was she safe in assuming that the island was too small for that task? I think she failed because she assumed the size of the island. This is where over thinking comes into play, and I'm sure that's what stopped her from solving the problem.

I had the same thoughts. The size of the island does seem to come into play. At some point it would be too small for a burn line to work, but then we wouldn't need to ask the question. The problem is, we the listener do not know that a small island moots the question.
 
I choose only situations where there is a 50/50 outcome. In sports, a team either wins or it doesn't. Stocks either rise or fall. I choose races with no more than 6 contestants and which give prizes for 1st 2nd and 3rd place, so the chance of winning a prize is still 50/50. (unrealistic to find races which award half the contestants perhaps) I wait till the apprentice is down to 2 people, choose elections with only two candidates, ect. Knowing my outcome will be either-or with each prediction, I start out with millions of letters and send half the either and half the or. After the event I predicted, I drop the half that missed from my mailing list, and send the next prediction to only the hits. I do this for 6 months until 1 person remains on the list, and then I hit them with my car.

i was right. this scenario, however unlikely, is not ruled out a priori.

just semantical fapping.

your puzzle was bad and you should feel bad.

just kidding, i still love you.
 
I assume that he would say daughters if he had two. But we all know what assumptions make...
Also I'm making this assumption off very little evidence, so it is risky anyways.

Well he just answered 'yes'. The question was, do you have any daughters. The only thing you can safely assume from that is he has at least one. You could infer he has two, and the chances of you being right would be...
 
I had the same thoughts. The size of the island does seem to come into play. At some point it would be too small for a burn line to work, but then we wouldn't need to ask the question. The problem is, we the listener do not know that a small island moots the question.

I think you nailed it, I was unable to point it out. If the island was too small it's a baited question. So we must assume its large enough for any situation. That's why I said she failed; because she failed to take into account that the island me be large. Seems she forgot Australia is an island/continant.
 
Well he just answered 'yes'. The question was, do you have any daughters. The only thing you can safely infer from that is he has at least one. You could infer he has two, and the chances of you being right would be...

Always 50/50. Unless he wears a fur cod peice that raises his average testicle temperature, making him more likely to have females.
 
Back
Top