Richard Feynman's Lunch Order Sparks 50-Year Mathematical Mystery解決

How Richard Feynman's Lunch Order Became a 50-Year Mathematical Mystery Solved by Scientists

In the late 1970s, a seemingly mundane lunch at a Thai restaurant in Glendale, California, led to one of physics' most peculiar unsolved puzzles. Nobel laureate Richard Feynman, known for his groundbreaking work in quantum electrodynamics, turned his friend Ralph Leighton's indecision over ordering ginger chicken or trying a new dish into a mathematical problem that would remain unsolved for decades.

Feynman, a professor at Caltech and co-author of the bestselling memoir *Surely You're Joking, Mr. Feynman!* with Leighton, saw the dilemma as an example of what mathematicians call an "explore-exploit dilemma." This type of problem involves balancing the risk of trying something new (exploration) against the comfort of sticking to a known favorite (exploitation). During the meal, Feynman scribbled equations on a piece of paper, framing the decision as a formal mathematical optimization problem and solving it on the spot. However, his notoriously messy handwriting rendered the notes nearly indecipherable for decades.

The "Restaurant Problem" and Its Unique Challenge

The puzzle Feynman tackled falls under a category of mathematical questions called optimal stopping problems. These problems are used to model decisions like when to stop searching for a house or a job and commit to an option. However, Feynman's version introduced a twist: unlike a house or job search, where options may no longer be available if passed over, diners can always return to a restaurant or dish they already enjoy. This added complexity meant the goal was not just to find the single best option but to maximize overall enjoyment across multiple meals.

Deciphering Feynman's Handwritten Solution

Nearly 50 years after Feynman's lunchtime scribble, a team of researchers led by cognitive scientist Brian Christian (University of Oxford), Evan Russek (Hunter College), and Thomas Griffiths (Princeton University) finally cracked the mystery. They painstakingly reconstructed Feynman's notes, which had been preserved in his personal archive but were nearly illegible due to his rushed and idiosyncratic handwriting. The team enlisted help from Ralph Leighton and Michael Gottlieb, curator of the official Feynman Lectures website, to fill in the gaps.

Once deciphered, the notes revealed that Feynman had derived a "quality threshold"—a benchmark to determine whether trying a new dish was worth the risk compared to sticking with a known favorite. This threshold declined over time, reflecting the idea that diners should become less inclined to explore as they near the end of their dining experiences.

Confirming and Extending Feynman's Solution

The researchers confirmed that Feynman's solution was mathematically optimal for the specific problem he had framed. They then expanded his work to cover broader scenarios, such as choosing between multiple restaurants rather than just dishes at a single eatery. Testing these generalizations against various distributions of restaurant quality, they found closed-form solutions that extended Feynman's lunchtime insight into a comprehensive theoretical framework.

Real-World Implications for Decision-Making

To see how Feynman's model compared to real-world behavior, the researchers conducted experiments where participants made a series of dining decisions. They found that human decision-making closely mirrored the mathematically optimal strategy Feynman had derived. Participants' tendency to settle for "good enough" options shifted over time in a way that matched Feynman's declining threshold curve. This suggests that everyday intuition about when to stop exploring and stick with a favorite may be more mathematically sophisticated than previously thought.

Feynman's playful approach to a friend's lunch dilemma has not only stood the test of time but also provided new insights into human decision-making. What began as a casual scribble on a restaurant napkin has evolved into a robust mathematical framework, bridging the gap between theoretical physics and the psychology of choice.