How to settle a group dinner decision without an argument

Group dinner arguments are often less about food than about process: nobody wants to be the person who decides, and whoever pushes hardest tends to win. A good process addresses that before anyone names a restaurant. This guide walks through one that can work for a table of friends, a team lunch or a family weekend, and shows where LocalSelections fits at each step.

Start with the category, not the restaurant

"Where should we eat?" has hundreds of answers, which is why it stalls. "What kind of food?" has maybe a dozen. Settle the cuisine first and the restaurant question shrinks to a handful of nearby places.

In LocalSelections the wheel and the other games work on categories first. You pick Food, spin or roll over cuisines such as Italian, Mexican, sushi or burgers, and only then does it look up real places near you in that category. If your group has a favourite shortlist, you can build a custom wheel with your own options and weights instead.

Set the limits before the vote

Two limits remove most of the friction:

• Hard no's. Anything a person cannot eat should be removed from the list before voting, not argued about afterwards. Allergies and dietary rules are not preferences.
• Distance and time. Agree how far you will travel. LocalSelections searches around the location you share, so the results already stay close; the useful conversation is whether anyone has a ride constraint.

Let everyone rank, not just shout

When people say their favourite out loud, the loudest option wins and the quiet person quietly resents it. Ranking gives every person a way to express a full ordering of their preferences, including their second and third choices.

LocalSelections' group mode uses that idea. Each person taps the options in the order they want them. Behind the scenes it uses a Borda count: with a list of n ranked options, the first choice gets n−1 points, the next n−2, and so on down to zero. An option someone left unranked gets zero from that person, which counts as abstaining, not as a veto. Every option also gets a base weight of one, so nothing disappears from the wheel entirely.

The result is a wheel whose slices are sized by the group's combined rankings: options with more support get bigger slices, and options nobody ranked get small ones. We cover the mechanics in how to run a live group vote.

Then let chance finish the job

A weighted spin does something a straight count cannot: it gives the decision an outcome nobody personally chose. Once the slices reflect the group, nobody has to concede to a specific person. The favourite is still the most likely result, and everyone agreed to the process in advance.

If you would rather decide by elimination, the Decision Bracket shows nearby picks one at a time and you swipe through until one survives. It may suit a pair or a small group better than a large one.

When the choice is close, score it

Some dinners deserve more than a coin flip: a birthday, a client meeting, a place you will only get to try once. For those, the weighted decision matrix lets you name the things that matter (price, distance, noise, whether it takes bookings), give each a weight, and score each candidate from one to five. The total tells you which option fits your own priorities, and the table shows why.

Make the final step easy

Once you have a result, the next friction is logistics. After you pick, LocalSelections offers a "what next" menu for the chosen place: directions in Google Maps and, for food places, links that search food-delivery services for the restaurant if you would rather order in. Those are links out; LocalSelections does not take bookings or orders itself.

A short checklist you can reuse

1. Agree the category (cuisine or type of outing) before naming venues.
2. Remove hard no's and agree a travel limit.
3. Let each person rank the options privately, or tap in the app.
4. Spin the weighted wheel and treat the result as final.
5. If it is a special occasion, score the top three in a matrix first.
6. Check the place is open before you set off. Listings can be out of date, which is why place results carry a "Report closed / wrong" button, and why we explain how places are chosen in how LocalSelections chooses and ranks nearby places.

A note on fairness

No method is perfectly fair, and a weighted random pick can still land on an option somebody ranked low. That is the trade you are making: you accept a small chance of a less-loved result in exchange for a process nobody can accuse of favouring someone. If a result truly will not work for a person (an allergy, a closed venue), skip it and spin again. The goal is a decision, not a ritual.

Ready to try it? Open the games, pick Food, and let the table spin.