Convolutions | Why X+Y in probability is a beautiful mess

Item: Convolutions | Why X+Y in probability is a beautiful mess
Rating: 5
Author: PostSingularity Institute

by 3Blue1Brown

Formal & Physical Sciences Mathematics PBMathematics PBKCalculus and mathematical analysis

🎙 3Blue1Brown 👥 8.6M 📅 June 27, 2023 ⏱ 27 min 👁 987K 📄 science communication 🧭 2026-08-28

Available in: English (current) Français

Keywords

convolutionprobability densitycentral limit theoremvisualizationrandom variables

Summary

The video begins with a quiz about the distribution of the sum of two independent normal random variables, setting the stage for a deep dive into convolution. It first revisits the discrete case using weighted dice, presenting two visualizations: summing probabilities along diagonal slices of a joint probability grid, and the ‘flip-and-slide’ method where one distribution is reversed and slid across the other, computing dot products. These lead to the discrete convolution formula. The video then transitions to continuous distributions, explaining probability density functions and deriving the continuous convolution integral by analogy. An interactive demo illustrates the mechanics of the integral, showing how the product of f(x) and g(s-x) is integrated to yield the convolution value. The example of summing uniform distributions demonstrates how convolution produces a triangular distribution, and repeated convolutions lead to a bell-shaped curve, illustrating the central limit theorem. The video concludes with a second continuous visualization using diagonal slices of the product surface f(x)g(y), which provides an elegant proof of why the normal distribution is the fixed point of convolution, answering the opening quiz.

178 words

Critical Evaluation

Value of the Information & Strength of the Argument

The video excels in providing deep intuition and multiple perspectives on convolution. It builds the concept from the ground up, using concrete examples and interactive visualizations to make abstract mathematical operations tangible. The argumentation is clear and logical, progressing from discrete to continuous cases and from specific examples to general principles. The connection to the central limit theorem is particularly well-motivated, showing how repeated convolution naturally leads to the normal distribution. The value lies in its ability to transform a potentially dry topic into an engaging and insightful exploration, making the mathematics feel intuitive and beautiful.

Scientific Rigor, Source Quality, Title Accuracy

The video demonstrates high scientific rigor. The mathematical explanations are accurate and well-structured, with careful attention to details such as the independence assumption and the scaling factor in the continuous case. The sources cited are primarily the channel’s own resources (manim, GitHub, website) and the music used, which are relevant but not academic references. The title accurately reflects the content, focusing on convolution and its role in probability. The video’s strength lies in its pedagogical approach rather than in citing external literature, which is appropriate for its educational purpose.

200 words

Title / Content Match

The title accurately reflects the content, focusing on the concept of convolution in probability and its connection to the central limit theorem.

Quality & Reliability

9/10

The video is produced by a renowned mathematics educator with a strong track record of accurate and insightful explanations. The content is mathematically rigorous, with clear definitions and derivations, and the visualizations are carefully constructed to illustrate the concepts. The channel's reputation and the positive reception from the community support high reliability.

Chapters

Cited Sources

3Blue1Brown website — Official website with additional resources and information about the channel.
Manim (3Blue1Brown's animation library) — The custom Python library used to create the animations in the video.
Manim Community Edition — Community-maintained version of the Manim library.
3Blue1Brown video code repository — Repository containing code for specific videos and projects.
Music by Vincent Rubinetti — Composer of the music used in the video.
The Music of 3Blue1Brown on Bandcamp — Album page for the music used in the video.
The Music of 3Blue1Brown on Spotify — Streaming page for the music.
3Blue1Brown FAQ — FAQ section explaining the use of Manim.
3Blue1Brown subreddit — Community discussion forum.

Concurring Sources

Convolution — The video's explanation aligns with the standard mathematical definition of convolution.
Central limit theorem — The video's demonstration of repeated convolutions leading to a normal distribution is a direct illustration of the central limit theorem.

External References

3b1b.co

Contribution & Novelties

The video’s original contribution lies in its pedagogical approach to convolution, offering two distinct visualizations (diagonal slices and flip-and-slide) that build intuition for both discrete and continuous cases. It elegantly connects these visualizations to the central limit theorem, providing a satisfying explanation for why the normal distribution emerges from repeated convolution. The interactive demo and the final diagonal-slice proof offer fresh perspectives that are not typically found in standard textbooks.

Pour aller plus loin :

Convolution — Wikipedia article providing a comprehensive overview of convolution in various contexts.
Central limit theorem — Wikipedia article detailing the theorem and its implications.
Probability density function — Wikipedia article explaining the concept of PDFs.
Characteristic function (probability theory) — A key tool for proving the central limit theorem, related to the Fourier transform.

130 words

Radar Profile

The radar chart shows a very strong profile with high scores across all dimensions, particularly in quality of information and fiabilité. The slightly lower score in niveau technique reflects the fact that the video is aimed at a general audience, but it still maintains a high level of mathematical depth.

Reliability 9/10

💬 Très positif. Sur les 30 commentaires analysés, les spectateurs expriment une admiration unanime pour la clarté des explications, la beauté des animations et la capacité de la vidéo à rendre des concepts mathématiques complexes intuitifs et accessibles.