Signal Versus Noise with the Lindy Filter
Judge what's true and lasting by how long it has already survived and how often you look
- Difficulty
- Moderate
- Time to result
- ~ongoing to results
- Steps
- 4
- Confidence
- 83%
Signal versus noise is the discipline of separating the small fraction of information that actually matters from the vast majority that will be irrelevant tomorrow. MacGill pairs it with the Lindy effect: the future life expectancy of a non-perishable thing is proportional to its current age, so a book that has lasted 200 years will likely last 200 more, while a day-old BuzzFeed post will likely vanish by tomorrow. A second lever is frequency — using the trading analogy, at year-one intervals roughly half of market movement is signal, but check daily and it is almost all noise. Together they tell you to weight long-surviving sources heavily, resist recency and closeness biases, and reduce how often you consume volatile information so you act on signal rather than tomorrow's fish-and-chips wrapper.
Origin
MacGill combines Nassim Taleb's Lindy effect with a Shane Parrish trading analogy about signal and noise. He notes the classics endure — War and Peace, 1984 — because something foundational underpins them, unlike content made in the last 24 hours.
Core principles
- 01The longer something has survived, the longer it will likely last
- 02Higher observation frequency means more noise and less signal
- 03New is not better; recency is a bias, not a virtue
- 04Classics endure because something foundational sits underneath them
How to run it
- 1
Apply the Lindy test
Ask how long the book, idea, or source has already existed, and use that age to estimate its remaining lifespan.
Pro tip A 200-year-old book will likely be read for 200 more; a day-old post likely dies tomorrow.
- 2
Check your observation frequency
Recognise that the more often you look at volatile information, the higher the noise-to-signal ratio.
Pro tip Lengthen the interval between checks to raise the signal fraction.
- 3
Guard against recency and closeness bias
Notice the pull toward the newest, most visible content and treat novelty as a bias rather than a merit.
Watch out The brand-new self-help book is not automatically better than the enduring classic.
- 4
Weight enduring sources
Prioritise material that has survived decades or centuries, and study the historical figures behind ideas, not only current voices.
In the wild
Put a long-term trade on a stock. Compare it at the start and end of year one and roughly half the movement is signal worth heeding. Shorten the interval to every six months and there is more noise, less signal. Check it daily and it is almost all noise and near-zero signal, because the only question that matters is whether the information helps you make an effective decision.
→ A clear rule that frequent checking degrades decision quality.
MacGill notes people binge modern entrepreneur content from Gary Vee or Grant Cardone yet have no idea who John Rockefeller, Benjamin Franklin, or JP Morgan were — the figures who actually reshaped society. Studying what survived history yields far more durable lessons than recency-biased new content.
→ A shift from recency-driven consumption to enduring, foundational sources.
Common mistakes
Confusing new with better
Recency bias makes fresh content feel more relevant when its short survival predicts short usefulness.
Checking volatile data too often
Increasing observation frequency drowns the signal in noise and provokes bad decisions.
Is it for you?
Best for
Building an information diet, choosing what to read, and filtering market or media movements.
Not ideal for
Genuinely fast-moving domains where the newest data really is the relevant data.
From the transcript
“if it's been around for a day it's probably not going to be around there tomorrow”
“as you begin to increase the frequency at which you look at the market movement... there's more noise and less signal”
From the episode
Mental Models 101 - How To Make Better Decisions - George MacGill - #069
George MacGill