Scientists uncover Feynman’s formula for finding best holiday restaurant
Researchers decoded a hidden Feynman equation for choosing when to stop searching for restaurants while traveling.
Intelligence analysis by GPT-5.4 Mini
A team says Richard Feynman turned a dinner-choice puzzle into a stopping problem: keep exploring until a place clears a moving quality threshold. Their tests suggest people use a simpler version of that strategy.
A scientist named Richard Feynman helped turn a dinner problem into a math puzzle. The question was: when should a person stop trying new restaurants and just keep going back to one good place?
The answer is like picking a campsite on a long trip. If many nights are left, it makes sense to keep looking for a better spot. If almost no time is left, it is smarter to stop searching and enjoy what is already good.
Researchers found that real people do something close to this, but in a simpler way. They lower their standard little by little as the trip gets shorter, kind of like being pickier at the start of a treasure hunt and less picky near the end.
Analysis
What the paper says
Researchers reporting in the Proceedings of the National Academy of Sciences say they reconstructed a handwritten Feynman solution to a problem about when to stop searching and start choosing. The original spark, according to the article, was a lunch conversation in the 1970s about whether to keep trying new dishes or stay with a favorite meal.
The team reframed the puzzle as a traveler deciding where to eat over a fixed number of nights in a city. In Feynman’s approach, the traveler samples different restaurants until one beats a threshold for quality. That threshold is not static: it drops faster as fewer nights remain, because there is less time left to benefit from a great discovery.
How the model changes
The article says the researchers also tested versions where restaurant quality is not evenly distributed. If most places are mediocre but a few are excellent, the strategy is to keep searching longer. If restaurants tend to be broadly similar and already decent, the threshold should be lower and the search should end sooner.
To compare theory with behavior, the team recruited 2,520 participants for an online task. People were shown a city stay of varying length and a set of restaurants with revealed quality values after selection. The results suggested participants did not follow Feynman’s exact curve. Instead, their threshold fell roughly linearly with the share of nights remaining.
The researchers say that simpler rule still worked well. The article frames the finding as evidence that people use intuitive stopping strategies that resemble the math, even if they do not mirror the original equation exactly.
Key points
- Researchers say they decoded a Feynman note about when to stop searching for a better restaurant.
- The problem is treated as a stopping problem: balancing exploration against using the time left.
- Feynman’s model uses a quality threshold that falls faster as the trip nears its end.
- In a test with 2,520 participants, people used a simpler, roughly linear drop in threshold.
- The article says the strategy changes when restaurant quality is uneven rather than evenly spread.
The study could help explain how people make better choices when time is limited and options keep changing. The article suggests the strategy may be useful beyond food, in any situation where searching has a cost and the chance to revisit a good option matters.
The paper also shows that the elegant Feynman solution is not exactly how people behave in practice. If the real world is messier than the model, the best rule may depend heavily on how options are distributed, making one universal strategy hard to rely on.
