Only people with an IQ of 115 or higher can get this right

Page 11 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

Ryland

Platinum Member
Aug 9, 2001
2,810
13
81
50% since you know you are either in the box that had 2 gold balls or the one that had one of each. Thus there is equal chance that the remaining ball is gold.
 
  • Like
Reactions: Cerb

Carson Dyle

Diamond Member
Jul 2, 2012
8,173
524
126
Obviously the question is worded wrong because it doesn't indicate what box you're in and for it to be 2/3 you'd then have to be in a 2nd box which isn't implied.

Therefore there are 2 acceptable answers

That is incorrect.

That's why they call it a probability question.
 

local

Golden Member
Jun 28, 2011
1,852
517
136
Ok, after reading 250 some odd posts and rereading the original question several times I can see where the 2/3 chance is coming from... I am going to change my answer. 2/3
 

Darwin333

Lifer
Dec 11, 2006
19,946
2,329
126
Who is they? Mayne is the only one who can make that call and he is balls deep in a broken AC unit, and/or drunk as fuck right now.

Are any ATOTer's missing? If so they could have taken him up on his "getting out of rear choke challenge" and he is currently balls deep in him.
 
  • Like
Reactions: Mayne

Mayne

Diamond Member
Apr 13, 2014
8,849
1,380
126
you guys are terrible at reading comprehension and statistics.
 

Hitman928

Diamond Member
Apr 15, 2012
6,695
12,370
136
Obviously the question is worded wrong because it doesn't indicate what box you're in and for it to be 2/3 you'd then have to be in a 2nd box which isn't implied.

Therefore there are 2 acceptable answers ;)

50% since you know you are either in the box that had 2 gold balls or the one that had one of each. Thus there is equal chance that the remaining ball is gold.

https://www.youtube.com/watch?v=CGMc8B60ZpU - Has an explanation with visuals.
https://github.com/yourfriendaaron/box-paradox - simulation that auto runs or that you can play to demonstrate how the probability works out. Note, must have python installed.
 

Hitman928

Diamond Member
Apr 15, 2012
6,695
12,370
136
you guys are terrible at reading comprehension and statistics.

keep-trollin-trollin-trollin.jpg
 
  • Like
Reactions: Cerb

Mayne

Diamond Member
Apr 13, 2014
8,849
1,380
126
you have never dealt with a 52 year old woman that is scottish and has giant fake books. ..and she is 6'2"
 

Mayne

Diamond Member
Apr 13, 2014
8,849
1,380
126
I'm going to work tomorrow and spend the rest of the week getting this apartmnet in order before she shows up..so tired right now...but I don't have animals living with me anymore :)
 

Mayne

Diamond Member
Apr 13, 2014
8,849
1,380
126
the best part is guys, you have never met my older sister. It's a whole new game from me.
 

Mayne

Diamond Member
Apr 13, 2014
8,849
1,380
126
the best is her son will be sleeping in my bedroom. it never ends at my place.
 

mindless1

Diamond Member
Aug 11, 2001
8,749
1,759
136

It's not that we don't understand how they came to the 2/3 conclusion, rather it's that Bertrand's, and your, logic is wrong. It's that you, and Bertrand, don't understand how to apply statistics to this problem, in the context of how it is posed in the question.

Once the stipulation is made that the first coin is gold, there are only two boxes to consider, but also only two coin colors. Because it was worded that way, there is no 3 in the calculation.

Look at the video at 1:57 where it was stated "there are 3 possible outcomes". This is in error. The first coin could have been either G1 or G2, that does not matter because it was only established that a gold coin was chosen, not a specific gold coin. Next probability was stated as choosing a 2nd gold coin, but again not a specific gold coin, so the choices G1 and G2 in the video are the same outcome because they are both gold coins.

The only valid answer is 50% (1/2) because G1 and G2 cannot be distinguised as different choices based only on the parameter of choosing a gold coin. There is no way to differentiate G1 and G2. The variable of gold is equal for both, so G1=G2, or they are both G1 and there is no G2 if you prefer to state it that way. G1/G2 and G2/G1 are both the same choice, cannot be enumerated separately.
 
Last edited:
  • Like
Reactions: paperfist

Hitman928

Diamond Member
Apr 15, 2012
6,695
12,370
136
It's not that we don't understand how they came to the 2/3 conclusion, rather it's that Bertrand's, and your, logic is wrong. It's that you, and Bertrand, don't understand how to apply statistics to this problem, in the context of how it is posed in the question.

Once the stipulation is made that the first coin is gold, there are only two boxes to consider, but also only two coin colors. Because it was worded that way, there is no 3 in the calculation.

Look at the video at 1:57 where it was stated "there are 3 possible outcomes". This is in error. The first coin could have been either G1 or G2, that does not matter because it was only established that a gold coin was chosen, not a specific gold coin. Next probability was stated as choosing a 2nd gold coin, but again not a specific gold coin, so the choices G1 and G2 in the video are the same outcome because they are both gold coins.

The only valid answer is 50% (1/2) because G1 and G2 cannot be distinguised as different choices based only on the parameter of choosing a gold coin. There is no way to differentiate G1 and G2. The variable of gold is equal for both, so G1=G2, or they are both G1 and there is no G2 if you prefer to state it that way. G1/G2 and G2/G1 are both the same choice, cannot be enumerated separately.

If this were true, why do experimental results prove out the probability of 2/3 and not 50%?
 

mindless1

Diamond Member
Aug 11, 2001
8,749
1,759
136
If this were true, why do experimental results prove out the probability of 2/3 and not 50%?
Care to elaborate? It's not necessary to disprove something that is already logically impossible and in error. An infinite # of reasons why something is wrong is not needed, only ONE reason.

It is simple fact that the only variable was gold vs silver, and thus G1 must equal G2 until another variable is introduced, something like heads up or down, different denomination, etc.
 

Hitman928

Diamond Member
Apr 15, 2012
6,695
12,370
136
Care to elaborate? It's not necessary to disprove something that is already logically impossible and in error. An infinite # of reasons why something is wrong is not needed, only ONE reason.

It is simple fact that the only variable was gold vs silver, and thus G1 must equal G2 until another variable is introduced, something like heads up or down, different denomination, etc.

Or, your reasoning is wrong so your disproof is invalid.

If you actually try the situation yourself, just as described in the problem, 2/3 of the time the second ball will be gold and 1/3 of the time the second ball will be silver. If it were a 50% chance, the second ball you pull should be silver half the time and gold half the time. If the real life (perfectly repeatable) result of experimentation doesn't match your logical conclusion but matches someone else's, yours is the one that is incorrect.