News for hackers. Stuff that matters.
Re:Two envelopes problem by johnrob on Tuesday May 31, 2022 @11:07AM (#31570401)
Sounds like a misuse of relative return (growth rate). If you gain 100% today, then lose 50% tomorrow, you are back where you started. Absolute numbers give you the correct answer: x -> 2x is a gain of x, 2x -> x is a loss of x, so expected value is zero (x is a hard number not a ratio).
Re:Two envelopes problem by metacritic12 on Tuesday May 31, 2022 @11:11AM (#31570453)
Exactly! The logic of "the average of +100% and -50% is positive" only applies if the denominator is the same. In this case it isn't.

The relative percentage "rule of thumb" doesn't even feel like a primitive in math. It seems wrong to reason from that.

Re:Two envelopes problem by platz on Tuesday May 31, 2022 @11:15AM (#31570494)
why isn't this listed in wikipedia as an answer.
Re:Two envelopes problem by whatshisface on Tuesday May 31, 2022 @12:07PM (#31571237)
Because if the larger envelope is changed to contain four times as many dollars, then /2 vs. *4 points back towards swapping indefinitely.
Re:Two envelopes problem by MrBlueIncognito on Tuesday May 31, 2022 @12:53PM (#31571829)
If one envelope contains 4 times the money of the other envelope, then by swapping you either gain 3x or lose 3x, where x is the amount of money in the smaller envelope. Even in this case, the average return is +3x -3x = 0.
Re:Two envelopes problem by vehementi on Tuesday May 31, 2022 @11:53AM (#31571030)
It is, it's the one where you treat both as having the same expected value
Re:Two envelopes problem by platz on Tuesday May 31, 2022 @12:00PM (#31571165)
that the expected value is the same that is an outcome of all solutions, so just stating that is reductive.

What is at question is the reason for getting to the expected value is the same.