Let’s play a game. You’ll flip a coin, and if it comes up heads I’ll give you $1. If you flip heads again I’ll give you $2, then $4, then $8, and so on. When the coin comes up tails, the game is over and you can keep your winnings.

Because I’m taking a risk, I ought to charge you an entrance fee. What’s a fair fee? Surprisingly, it seems I should charge you an infinite amount of money. With each new flip your chance of success is 1/2 but your prospective earnings double, so your total expected earnings — the earnings times their chance of being realized — is infinite:

E = (1/2 × 1) + (1/4 × 2) + (1/8 × 4) + … = ∞

Nicholas Bernoulli first described this problem in 1713. One proposed resolution is that it ignores psychology — we’re considering the monetary value of the prize rather than its value to us. Gold shines more brightly for a beggar than for a billionaire; once we’ve amassed a certain sum, the appeal of greater riches begins to diminish. “The mathematicians estimate money in proportion to its quantity,” wrote Gabriel Cramer, “and men of good sense in proportion to the usage that they may make of it.”

(Thanks, Ross.)