Named after a New York delicatessen where comedians reportedly observed that a performer's future career length was proportional to how long they had already lasted.

The generalized claim: for non-perishable things — ideas, books, technologies, institutions — every additional period of survival is evidence of a longer remaining life. A play performed for four hundred years is a better bet to be performed in fifty more than one that opened last year.

The restriction is the important part. It applies only to things that do not have a natural lifespan. A person aged eighty has a shorter expectancy than one aged forty; a book in print for eighty years has a longer one than a book in print for forty.

Why it holds

It is a selection argument rather than a mystical one.

Things that survive a long time have been exposed to many conditions that could have killed them and were not killed. They have already passed tests the new thing has not faced. Survival is evidence of some robustness that is hard to observe directly, which is why it works as a proxy.

There is also a network mechanism: the longer something has been around, the more other things depend on it, which makes removing it progressively harder. That is path dependence contributing to the same result, and it is why the effect is stronger for infrastructure than for fashion.

The uses

Choosing tools. A twenty-year-old technology has demonstrated something a two-year-old one has not, and the difference is not primarily technical quality but maintenance risk — who will still be supporting it, how much accumulated knowledge exists, how many others depend on it.

Choosing what to read. A book still being read after a century has survived a selection process no new book has been through. This is a good default for someone with limited reading time and it will make you miss things, which is the trade.

Estimating how long a problem will persist. A complaint customers have made for a decade is unlikely to resolve itself next year.

The limits, which are real

It says nothing about quality. Durable and good are different properties, and plenty of durable things persist because of lock-in rather than merit. QWERTY is extremely Lindy.

It is wrong at regime changes. Something that survived for decades under stable conditions carries no evidence about a world where the conditions changed, which is exactly what creative destruction describes.

It is a prior, not a verdict. Where you have direct evidence about a specific case, use the evidence. Lindy is what you fall back on when you have nothing better, and most arguments that invoke it have something better available and would rather not look.