Why is the “central limit” a normal distribution?

Why is the "central limit" a normal distribution?

A visual trick to compute the sum of two normally-distributed variables.
3b1b mailing list: https://3blue1brown.substack.com/
Help fund future projects: https://www.patreon.com/3blue1brown
Special thanks to these supporters: https://www.3blue1brown.com/lessons/gaussian-convolution#thanks

For the technically curious who want to go deeper, here’s a proof of the central limit theorem using Moment generating functions:

And here’s a nice discussion of methods using entropy:

Relevant previous videos

Central limit theorem

Why π is there, and the Herschel-Maxwell derivation

Convolutions and adding random variables

Time stamps
0:00 – Recap on where we are
2:10 – What direct calculation would look like
3:38 – The visual trick
8:27 – How this fits into the Central Limit Theorem
12:30 – Mailing list


These animations are largely made using a custom Python library, manim. See the FAQ comments here:

You can find code for specific videos and projects here:

Music by Vincent Rubinetti.

Download the music on Bandcamp:

Stream the music on Spotify:


3blue1brown is a channel about animating math, in all senses of the word animate. And you know the drill with YouTube, if you want to stay posted on new videos, subscribe: http://3b1b.co/subscribe

Various social media stuffs:
Website: https://www.3blue1brown.com
Twitter: https://twitter.com/3blue1brown
Reddit: https://www.reddit.com/r/3blue1brown
Instagram: https://www.instagram.com/3blue1brown
Patreon: https://patreon.com/3blue1brown
Facebook: https://www.facebook.com/3blue1brown


Related post