Lindy Effect, explained.
The Lindy Effect is a longevity heuristic: for some non-perishable ideas and practices, a longer surviving history can be evidence of a longer remaining life.
Why it happens
Survival is informative when repeated challenges test enduring usefulness. A proof does not wear out like a battery. But age alone cannot tell you whether persistence came from usefulness, institutional protection, or a lack of alternatives.
A book that has been read for a century may have a different outlook from one published last week. It has already endured changes in taste, distribution, and the world around it. That survival can be informative—but only if the process that kept it relevant continues.
The Lindy heuristic concerns some non-perishable things: ideas, texts, practices, and institutions. It does not say that old things cannot disappear. It suggests that, under suitable assumptions, greater age can be associated with greater remaining longevity.
Read the result
Change observed age while holding the other controls fixed, then change Lindy strength. If age matters only when strength is positive, you are seeing an assumed relationship—not an independent discovery about old things.
A worked example
Choosing a reference book
Two books explain the same topic: one is new, and one has remained in use for decades.
Ask what kept the older book useful and whether the subject has changed. A timeless proof and an obsolete software manual face different conditions.
Use longevity as one piece of evidence alongside accuracy, relevance, and the reason the work survived.
OPTIONAL DEEPER DETAILGo deeper: inside the model
Age, strength, and remaining life
a is observed age, s is Lindy strength, and m is the model’s median remaining life in years. At strength 0, the median is fixed at 50 years. At strength 1, it equals observed age.
Remaining lifetimes are sampled from a lognormal distribution with log-median ln(m) and log-scale spread 0.3, 0.7, or 1.2. Sample statistics vary slightly with the seed. This convenient teaching model imposes an age relationship; it is not a fitted survival distribution.
Where this idea is useful
Age can be one input when evaluating a method, a standard, or a body of knowledge. It invites a useful question: what has this already survived? Pair that question with another: has the environment changed enough to make its history irrelevant?
A mathematical idea
A useful proof can remain useful without physically wearing out. Continued use may reveal enduring relevance.
A shared standard
A long-lived protocol can accumulate tools, familiarity, and adoption. Those mechanisms may support persistence.
A book still in print
Repeated discovery across generations can suggest durability, while saying little about any particular title’s future.
A common misconception
“An old idea must be correct.”
Longevity is not a truth test. Persistent misconceptions can survive too; the mechanism behind survival matters.
What this explanation leaves out
- Physical organisms and components age and wear out. Their survival mechanisms generally differ from the non-perishable examples above.
- Disruption, regulation, changing tastes, and dependence on a fragile institution can break a historical pattern abruptly.
- Looking only at survivors hides failures. A sound empirical claim needs a defined population, information about those that disappeared, and an appropriate survival model.
- This experiment builds the age–longevity relationship into its assumptions. It does not estimate a real survival process or establish that the heuristic is true.
Does the Lindy Effect apply to people?
Not in the same straightforward way. Biological aging involves wear, disease, and changing hazards. A claim about non-perishable ideas cannot be transferred to a person's remaining lifespan.
Name an old practice you trust. What has it survived, and what could make that history stop being informative?
Associated thinkers
Associations marked provisional are awaiting source review.
Further reading
Explore Taleb's writing on uncertainty for the broader context of the Lindy heuristic. The link is contextual reading, not validation of this simulation. Thinker relationships marked provisional remain editorial associations, not claims of sole invention.