The Most Controversial Problem in Philosophy

A controversial mathematical and philosophical problem with no consensus answer that has generated hundreds of papers over 22 years.

• Sleeping Beauty is put to sleep on Sunday night • A fair coin is flipped • If heads: she wakes Monday once • If tails: she wakes Monday and Tuesday • Each time she's put back to sleep, she forgets previous awakenings

When Sleeping Beauty wakes up, she is asked: 'What do you believe is the probability that the coin came up heads?'

• Intuitive answer: one-third (Monday heads, Monday tails, or Tuesday tails) • Alternative reasoning: one-half (the coin is fair, no new information was received)

Sleeping Beauty should answer one-half because she knows the coin is fair and receives no new information when she wakes up.

• Nothing changes between when the coin is flipped and when she wakes up • She knew she would be woken up regardless of the outcome • If asked the probability before sleep, she would say one-half

Not all three outcomes are equally likely just because they exist. The Tails probability (50%) should be split across two days (25% each), giving 50% heads and 50% tails combined.

When the experiment is repeated many times through coin flipping simulations, the results show one-third heads, one-third Monday tails, and one-third Tuesday tails—contradicting the 50-25-25 prediction.

When Sleeping Beauty wakes up, she learns something important: she has transitioned from a two-state reality to a three-state reality.

• Two possible states before waking: heads or tails • Three possible states when awake: Monday heads, Monday tails, or Tuesday tails • Each of the three states should be assigned equal probability

The implied question is not 'what is the probability the coin showed heads' but rather 'given you're awake, what's the probability the coin came up heads?' The answer to this is one-third.

Repeated simulations consistently show Sleeping Beauty wakes one-third of the time for heads and one-third each for Monday and Tuesday tails, supporting the thirder position.

Instead of two wakeups for tails, Sleeping Beauty is woken a million times if the coin lands tails, but only once if it lands heads.

The thirder position becomes more compelling when the disparity is extreme: with a million more wakeups in the tails case, it seems absurd to assign equal probability to heads.

Drawing from a bag containing one white marble and a million black marbles illustrates the point: it would be absurd to say there's an equal chance of drawing the white marble.

This extreme version uses the same logic to argue we are likely living in a simulation, since a future civilization could create unlimited copies of simulated worlds.

A soccer game between Brazil (80% likely to win) and Canada (20% likely to win). If Brazil wins, you're woken once. If Canada wins, you're woken 30 times.

The thirder position would predict Canada won, since there are many more wakeups in that scenario.

• If betting correctly on individual games, you should say Brazil each time • If betting on total correct answers across many wakeups, Canada wins despite being unlikely

The disagreement between halfers and thirders hinges on what question is being answered: correctly identifying outcomes (halfer view) versus correctly answering more questioning instances (thirder view).

Before the universe began, a coin was flipped: heads creates one universe, tails creates a quasi-infinite multiverse with every possible variation of Earth and its inhabitants.

Becoming conscious is like Sleeping Beauty waking up: there's no way to tell if you're in the single universe or one of the multiverse universes.

Given that far more universes exist in the tails scenario, should you conclude you're definitely in the multiverse, or are the odds 50/50?

This thought experiment demonstrates how the sleeping beauty problem connects to fundamental questions about the nature of reality and our place in it.

The best way to develop probability intuition is by working through scenarios and running simulations, as demonstrated with the Sleeping Beauty problem.

Both the halfer and thirder positions seem equally obvious to their respective proponents, which explains why this problem remains controversial despite decades of analysis.

Understanding these probability concepts has real implications for questions about simulation theory, multiverse existence, and how we assess unlikely outcomes.

Whether we're living in a simulation or a multiverse remains unanswered, but probability frameworks like those in the Sleeping Beauty problem are essential tools for exploring such questions.