“What do you feel like for dinner?”
If you live with someone, you probably say and hear that pretty often. And while it may not be a source of true conflict, dumb arguments or even just short-term frustration arise often enough to give hack standup comedians fodder for their sets.
Even though I don’t live with anyone, I get meals with friends quite a bit, so the question still comes up frequently. Years ago, I started using this method for choosing a place.
Steps
- Person 1 comes up with a list of five places that sound good enough.
- Person 2 eliminates two places from the list, narrowing it down to three.
- Person 1 chooses the final spot from the remaining three.
Guidelines
- This process should only be started after both people have said they don’t have a specific place in mind. If someone wants to go somewhere, go there.
- At any point, either person is free to stop everything and say, “Nevermind, I want to go here.” The process is not sacrosanct: its purpose is to produce a decision, so if one is arrived at before the last step, that’s great.
- No vetos outside of the described narrowing process. If the other person doesn’t like a place, they can eliminate it during their turn. If they dislike literally every option chosen by the other person so much that they refuse to eat at any of them, an open conversation about food preferences needs to happen.
- Switch up who goes first.
- If coming up with five options in the first step is too cumbersome or doesn’t provide enough value, you can shorten the process: Person 1 comes up with three options, then Person 2 chooses the final spot from those three.
Comments
This is nothing revolutionary or incredibly clever; if there’s any magic, it’s the boundary definition in the first step. The “problem” with the question in a privileged first-world nation often comes down to option paralysis: if neither person can think of anything that sounds significantly better than anything else, that leaves dozens or even hundreds of restaurants that are all effectively equal.
For people like me, it’s pretty easy to find some combination of static metrics to use for differentiation: distance, overall quality of food, speed of service, cost, etc. But that leads to repetitive choices, and it’s usually not good for one person to decide all (or even most) of the time. Based on conversations with partners and friends, many people don’t naturally dig for differentiation that formally, so this process requires that behavior from Person 1 by forcing them to name five restaurants from the sea of available options.
I’ve found that after doing this a few times, people tend to be more aware of what they don’t want on a given day. “What do you feel like for dinner?” “Nothing sounds especially good, but definitely not burritos or anything fried.” Which is great! The entire goal is to narrow down a large number of options to a manageable number, so anything that contributes to that process is positive.
On the other side, it’s also quite common for someone to hear the name of a place and realize it’s exactly what they want that night, especially if Person 1 thinks out loud while building their initial list. Again: they are encourage to announce their choice and stop the process.
This obviously only works well if both people are at least minimally considerate of the other person’s preferences. If one person consistently presents initial lists filled with options that the other person has told them they don’t like, there may be deeper issues to address.

Leave a comment