Fun Ways to Decide What to Eat

Cook the constraint, not the craving

Most food debates start at the wrong end. Asking what you fancy invites an unbounded answer, and an unbounded answer is exactly what nobody can produce at seven in the evening. Start with the things that are already fixed.

Three constraints do most of the work: what will spoil if you ignore it again, how long you have before people stop being reasonable, and who is actually eating. Answer those first and the field narrows from everything to about four things. The craving question then becomes easy, because you are choosing between real options rather than inventing one from nothing.

That last constraint is the one worth being careful with. This site gives no dietary or nutritional advice, and allergies, intolerances and ingredient labels are yours to check.

• What needs using up: the half bag of spinach, the chicken with two days left
• Time available: twenty minutes, forty, or a slow Sunday afternoon
• Who is eating: how many, and anything someone will genuinely not touch

Cuisine first, dish second

A long list of dishes is hard to choose from because the options are not comparable. Pad thai and shepherd's pie are not close enough to weigh against each other. Cuisine is a smaller, cleaner list — six or seven categories at most — and once you have picked one, the dish inside it almost picks itself.

Say Italian, then decide between pasta and something on bread. Say Indian, then decide between the dal you have made forty times and the thing you keep meaning to try. Two small decisions in sequence are far less tiring than one large one, which is the practical core of what people mean by decision fatigue.

Choosing a cuisine and choosing a restaurant are different problems, and mixing them is why takeaway nights drag. A cuisine is a direction. A restaurant is a specific place with a real menu, a delivery radius and an opening time. Agree the direction together, then let one person handle finding a place that fits it. Do both at once and you renegotiate the direction every time somewhere turns out to be closed.

The two-person veto

The veto method suits households of two, and it works by splitting the labour of choosing. One person proposes three options. The other removes one, with no justification required. The proposer picks from the two that survive.

Whoever did not want to think only has to react, which is much easier than generating ideas from a blank start. The proposer still gets a real choice at the end, so it does not feel like losing. And nobody has to defend a preference out loud, which is where these conversations usually go wrong.

Put a limit on it if the veto starts getting used every night. One per person per week keeps it honest and stops it becoming a way of winning by attrition.

Rotate who chooses by day

Some evenings do not need a method. They need a rota. Assign days: Monday and Wednesday are yours, Tuesday and Thursday are theirs, Friday belongs to whoever is cooking. The chooser picks, nobody negotiates, and the other person eats what arrives.

This works because it removes the discussion rather than making it more efficient. It also spreads the boredom fairly. Nobody ends up on a permanent loop of someone else's five favourite meals, and the person who always caved gets evenings that are genuinely theirs.

Rotas scale past two people — flatmates, families, a group who eat together on Thursdays. Write the order somewhere visible, because a rota nobody can remember is just another argument with extra steps.

Keep a shortlist of five you actually rotate through

Most households cook the same handful of meals and pretend otherwise. Naming that handful out loud is more useful than it sounds. Write down five meals you can make on a weekday without shopping, checking a recipe or resenting it.

Five is deliberately small. It fits in your head, it covers a working week, and it is short enough that adding something new feels like an event rather than a chore. When a sixth earns its place, drop whichever one you keep skipping.

Keep a second list of five for the nights you are ordering: places you actually like, not the twenty the app puts in front of you. A delivery app is built to keep you browsing. A list of five is built to get you fed.

Randomise last, once the list is honest

Randomising is the final step, not the first. A wheel full of things you do not want is worse than no wheel at all: you spin it, dislike the answer, and start negotiating again, having now spent the spin as well as the evening. Put five real options on it and the result is one you can live with whichever way it lands.

That is the whole trick. Randomness does not choose well, it chooses fast, and its only real value is that it ends the debate. So it helps exactly when every option on the list is one you would accept — and it has no place in medical, legal, financial, safety or employment decisions, where one outcome is meaningfully worse than another.

A few small features make the habit stick. Entries are editable, so you build the list once and keep it. No-repeat mode stops the same dinner coming up three nights running. Weighted odds let you make the twenty-minute meal likelier on a weeknight than the one that takes an hour. You can save the list locally or share it by link so everyone is spinning the same five. Results come from the browser's cryptographic random number generator and are settled before the animation plays. There is a separate article here on the food wheel itself if you want the mechanics in detail.