Decision Maker

Random selection from a list of options is mathematically equivalent to flipping a weighted coin — useful when choices are subjectively equivalent and the cost of further deliberation exceeds the cost of picking arbitrarily. Behavioral research suggests randomized choice often outperforms over-analysis on low-stakes decisions because it avoids analysis paralysis.

Enter options separated by lines or commas; the tool returns one selected uniformly at random. Each invocation is independent — re-running can yield the same result twice. For weighted selection, list the heavier options multiple times.

The underlying mechanism is simple. A pseudorandom number generator produces a value uniformly distributed in [0, 1). Multiply by the option count, floor it, and the result is an array index. Each option has equal probability of being selected. For weighted selection, list each option with a weight equal to its desired probability share — listing 'pizza' three times and 'sushi' once gives pizza a 75 percent chance. The PRNG (Math.random in browsers, typically xorshift or PCG-class) is statistically uniform but not cryptographically secure; the difference doesn't matter for decision-making but matters for use cases like raffles where guessability matters.

A worked example. Five lunch options entered: pizza, sushi, salad, burger, sandwich. Each invocation picks one with 20 percent probability. Run it 10 times and you might see (purely as illustration): sushi, burger, pizza, pizza, salad, sandwich, sushi, burger, sandwich, pizza. Repetition is normal — independent trials produce clusters and gaps that look non-random to humans. To pick without replacement (no repeat across selections in a single sitting), shuffle the list and read off the order. To weight pizza twice as likely as the others, list it twice: pizza, pizza, sushi, salad, burger, sandwich. Each entry is now 1/6 = 16.7 percent, but pizza appears twice for 33.3 percent total.

Limitations and the philosophical bit. Randomness genuinely helps when the options are roughly equivalent on the criteria that matter, and additional deliberation has diminishing returns. Restaurant picks, weekend activities, default-to-A-or-B project tiebreakers — all good fits. Anything with significant downside, ethical weight, or long-term consequence deserves explicit reasoning, not a coin flip. The Tim Harford observation, often quoted, captures something real: when the random pick feels disappointing, that's information — your preference was hidden by the appearance of equivalence. Use the random result as a forcing function to clarify what you actually wanted. People often re-roll until they get the answer they secretly preferred, which is fine — the act of generating and rejecting is the deliberation. Don't pretend that's the same as random selection working.