Decision Maker

A random pick is the right tool when (a) two or more options feel roughly equivalent and (b) you'd rather decide quickly than perfectly. "Where do we eat lunch?" is the canonical Singapore use case — a list of 4-6 hawker centres or food courts where any choice is fine but the act of choosing burns five minutes of group debate. A weighted pick is the right escalation when you do care more about one option but want to leave some chance for the others — e.g. weighting the closest hawker centre 3:1 against the further ones biases toward convenience without locking it in.

What weighting actually does

Behind the scenes the tool sums the weights, generates an unbiased random number in [0, total), and walks the list adding weights until the cumulative total exceeds the random number. That position is the pick. Two options with weights 1 and 3 will select the second option three times as often as the first, on average — but in a single pick you might still get the "less likely" choice. That's the point: weighted randomness is biased, not deterministic. If you want one specific option to always win, you don't want a decision maker, you want a choice.

Why crypto.getRandomValues over Math.random

JavaScript's standard Math.random() uses a pseudorandom generator seeded from the clock — fine for casual use but predictable if you can observe a few outputs. window.crypto.getRandomValues() draws from the OS entropy source (the same one TLS session keys use) and gives genuinely unguessable bits. For most users the difference is invisible; for anyone who cares about the principle of "actually fair" picks, it matters. The tool also applies rejection sampling to avoid modulo bias — a subtle effect where a naive random & n tilts toward smaller numbers by a few parts per million.

ASEAN decision-making culture

Random picks have local flavours across Southeast Asia. Indonesian and Malaysian groups often use suit (the local rock-paper-scissors gesture) for two-person ties rather than coins. Singaporean office groups frequently rotate the "decision burden" — the person who picked last gets to pick again. Filipino families use the "unahan" (first one to say it wins) for trivial things like the last piece of food. Thai office groups gravitate toward seniority-based decisions; using a random tool to override that hierarchy can feel novel. Vietnamese groups often blend democracy and randomness — vote first to narrow the list, then random-pick the final option from the top few.

When NOT to use a decision maker

Three contexts where random picks are the wrong tool: (1) decisions with reversibility cost — if you'd hate to undo the choice, that's a signal it isn't actually 50/50 and a random pick will frustrate you; (2) decisions involving other people's preferences you don't know — you'd be randomising on incomplete information; (3) decisions you've already made but don't want to admit — the tool will sometimes pick the "wrong" answer and you'll over-ride it, which means you knew the answer the whole time. Use the tool to escape mild deliberation paralysis. Don't use it to outsource real thinking.