The Twenty Questions Strategy: Asking Smarter Yes/No Questions

The classic game of twenty questions hides a beautiful piece of math: twenty perfect yes/no questions can distinguish between 2²⁰ — over a million — possibilities. Most players waste that power.

Think in halves, not hunches

Every yes/no question is a coin you spend. A great question splits the remaining possibilities roughly 50/50, so the answer is informative either way. "Is it a giraffe?" only helps if the answer is yes; "Is it bigger than a person?" helps no matter what. Computer scientists call this binary search — guessing games taught it to humans first.

A practical question tree

1. Alive or not? (splits almost any deck in half)
2. If alive: animal vs. plant → wild vs. domestic → bigger than a dog?
3. If not alive: can you hold it? → is it electronic? → is it in this room?
4. Last 25% of the timer: stop splitting, start naming your top candidates.

The mistakes that cost rounds

Asking questions whose answer you can predict (wasted coin), repeating a cousin of an earlier question ("Is it big?" after "Is it small?"), and guessing specific words too early. In Guess It!, a wrong guess costs seconds — three bad hunches can eat a third of your clock.

This kind of structured thinking is also genuinely good exercise — see what word games do to your brain. Then go practice your tree against the AI opponent.