Regression To The Mean

The Idea

Contributed by @philhagspiel |  Edited and curated by @philhagspiel

The best will always appear to get worse and the worst will always appear to get better over time.

icon

Logic & Reasoning

icon

Finding Truth

image

Sports stars who play a phenomenal season or ourselves being on an extraordinary winning streak at work — our performance will eventually get worse.

While the science of performance is complex and context-dependent, random chance always plays some role in basically everything we do. Regardless of our skills at playing basketball or picking effective business strategies, randomness will influence the outcome to some degree.

Whenever someone performs extraordinarily good or bad, luck or bad luck will have played some role. This means that outcomes are not perfectly correlated with skill (they almost never are).

As luck and bad luck themselves aren't stable over time, the following scenarios are highly relevant:

  • I perform extremely well which is partly based on luck. As luck isn't stable, but will itself even out over time, the "random" (lucky) part of my performance will get worse at some point — my performance declines overall.
  • I perform extremely bad which is partly based on bad luck. As bad luck isn't stable, but will itself even out over time, the "random

The principle of Regression To The Mean applies whenever we encounter strong deviations from averages (or means) or look at extreme groups.

It explains not only changes in observed performance but, for example, also why very intelligent people marry partners who are less intelligent than they are (it's way more likely to find someone who isn't as extraordinarily intelligent).

It further applies to business and science whenever we observe strong deviations from the mean values of a distribution (such as quarterly earnings or drug effectiveness values over time).

Explore

➞ This Veritasium Youtube video explains the concept beautifully in just over 7 minutes.

➞ In this Farnam Street blog post, the crucial relationship between Regression To The Mean and statistical correlation is explained nicely.

This article explores the urban legend of the "Sports Illustrated Cover Jinx" which can be explained by Regression To The Mean.

Resources

If this idea resonates with you, some of these resources might add value to your life.

LinkNAMEFormatAuthor
Predictably Irrational
Book
Dan Ariely
Thinking Fast And Slow
Book
Daniel Kahnemann
Factfulness
Book
Hans Rosling
The Sovereign Individual
Book
James Dale Davidson
VSI: Thinking & Reasoning
Book
Jonathan Evans
Antifragility
Book
Nassim Taleb
Skin In The Game
Book
Nassim Taleb
Fooled By Randomness
Book
Nassim Taleb
Principles
Book
Ray Dalio
59 Seconds - Think A Little Change A Lot
Book
Richard Wiseman
The Great Mental Models (vol. 2)
Book
Shane Parrish
The Great Mental Models (vol. 1)
Book
Shane Parrish
Enlightenment Now
Book
Steven Pinker
21 Lessons For The 21st Century
Book
Yuval Noah Harari
Paul Graham
Blog
Paul Graham
Farnam Street
Blog
Shane Parrish
Lesswrong
Blog
Edge.org
Blog
Untools.co
Blog
The Systems Thinker
Blog
Modern Wisdom
Podcast
Christ Williamson
You Are Not So Smart
Podcast
David McRaney
The Portal
Podcast
Eric Weinstein
Lex Fridman Podcast
Podcast
Lex Fridman
The Knowledge Project
Podcast
Shane Parrish
Philosophize This!
Podcast
Stephen West
Conversations With Tyler
Podcast
Tyler Cowen
Philosophy For Our Times
Podcast
Hidden Brain
Podcast
Kurzgesagt
YouTube Channel
TED-ed
YouTube Channel
Crash Course: Statistics
YouTube Channel
3Blue1Brown
YouTube Channel
Quanta Magazine
YouTube Channel
Primer
YouTube Channel
Veritasium
YouTube Channel
Talks at Google
YouTube Channel
Vsauce
YouTube Channel
Brilliant.org
Courses