A two envelopes paradox calculator shows the arithmetic of a famous puzzle. You hold one of two envelopes, one with twice as much money as the other, and you may swap. Enter the amount you see: the calculator lists both possible pairs, the tempting but flawed 5X ÷ 4 result, and the correct average of 3x ÷ 2 for each envelope.
How to use the two envelopes paradox calculator
- Type the amount in your envelope.
- Choose whether your envelope is the larger or the smaller of the two.
- Press Calculate and read the pair, the average and the explanation.
The puzzle
Two sealed envelopes hold money. One holds twice as much as the other, and you are given one of them at random. You may keep it or swap for the other. Because you do not know which envelope is which, the choice looks like a coin toss, so most people expect that swapping changes nothing. The paradox is that a short calculation seems to show that swapping always pays.
The tempting argument
Call the amount you see X. The other envelope holds either half of that, X ÷ 2, or double, 2X, and each looks equally likely. The expected value of the other envelope is then ½ × (X ÷ 2) + ½ × 2X = 5X ÷ 4, which is 25 percent more than X [1]. With 20 in your envelope, that is ½ × 10 + ½ × 40 = 25. The calculator prints this line so you can see exactly what the argument claims.
Why it cannot be right
If the argument were sound, it would also apply to the other envelope, and you would want to swap back, then swap again, forever. Both envelopes cannot be worth more than each other. The encyclopaedia entry on the problem says the switching argument is not valid as stated because the symbol A is used inconsistently [1]. In this calculator's notation, the same inconsistency affects X.
Fixing the notation
The repair is to fix the pair of amounts first. Let the smaller envelope hold x and the larger hold 2x. Whichever envelope you were handed, it is the smaller one half of the time and the larger one half of the time, so its expected value is ½ × x + ½ × 2x = 3x ÷ 2 [1]. The other envelope has exactly the same expected value, so the average over many rounds is the same whether you swap or keep.
What X means in each branch
In the tempting argument, X is the smaller amount in the branch where the other envelope holds 2X, and it is the larger amount in the branch where the other holds X ÷ 2. That is two different quantities with the same name. In the first branch the pair is X and 2X. In the second the pair is X ÷ 2 and X. Averaging across these two pairs as if the pair stayed the same is the hidden step that produces the extra 25 percent. This explanation is our own working.
A worked case: the larger envelope
Suppose you see 20 and you know it is the larger one. Then the smaller one holds 10, so x = 10. The pair is 10 and 20, and the average of the two envelopes is 3x ÷ 2 = 15 [1]. Swapping would give you 10 instead of 20, a loss of 10 in this case. The page also shows the case that you hold the smaller envelope.
A worked case: the smaller envelope
Now suppose you see 20 and it is the smaller one. The pair is 20 and 40, so x = 20 and the average is 30 [1]. Swapping would give you 40, a gain of 20. In the game itself you never learn which case you are in. Over many rounds, the rounds in which you gain and the rounds in which you lose balance out, which is why the calculator shows both arrangements side by side.
What the average means
The average shown is the average of the two envelopes in the pair, 3x ÷ 2, and it is the same for the envelope you hold and the one you do not hold, before either is opened. It is not a prediction about the single envelope you see, because once you see 20, the pair has become one of two particular pairs. The calculator keeps these ideas apart by printing the pair for each case.
Switching over many rounds
If you played the game thousands of times with the same pair of amounts, keeping every time and switching every time would earn the same total on average, 3x ÷ 2 per round [1]. Switching only pays in one particular round, and it pays exactly when you hold the smaller envelope. Since you cannot tell, the rule that you gain 25 percent by switching is an illusion created by the flawed argument.
What this calculator does not model
The page works with the plain version of the puzzle, with a known pair of amounts. It does not draw the amounts from a probability distribution, and it does not run a simulation. Discussions of the problem go further and give the amounts a distribution, which changes what you may conclude about a particular amount, but that is beyond this tool. We state this so that you do not read the output as advice about real games or bets.
Using the calculator
Try several amounts and both cases. Notice that the naive figure is always 1.25 times what you see, while the correct average is 0.75 times what you see in the larger case and 1.5 times in the smaller case. The naive figure averages the other envelope's two possible contents, 0.5 X and 2 X, while the correct figures average the two envelopes of each possible pair, which is a different calculation. The tool puts the two side by side so that the mix-up is visible.
Common mistakes
- Treating X as a fixed amount. It changes meaning between the two branches.
- Reading the average as the value of the one you see. It is the average of the pair.
- Believing the 25 percent gain is real. It comes from the flawed step.
- Using it as betting advice. It is a thought experiment.
What next
Use this often? Add Two Envelope to your Home screen.
Frequently asked questions
What is a two envelopes paradox calculator?
A tool that works through the two envelopes puzzle for an amount you enter, showing both pairs and the correct average.
Does switching envelopes help on average?
No. Both envelopes have the same average of 3x ÷ 2, where x is the smaller amount, so keeping or switching gives the same average.
Why does the 5X ÷ 4 argument fail?
It uses the same symbol for two different amounts in its two branches, so the average it computes is not a valid average.
What if I know my envelope is the larger one?
Then the other holds half, so switching loses. If you know it is the smaller, the other holds double and switching gains.
Is this a betting strategy?
No. It is an educational calculation about a thought experiment.
Sources
Was this tool helpful?
What went wrong?
We only receive the reason you pick, never what you typed into the tool.
Thanks for your feedback.