Bayes theorem, the geometry of changing beliefs
Why Hacker News recommends it
HN people recommend it as the most intuitive introduction to Bayes' theorem and cite it in base-rate arguments.
What it is
3Blue1Brown presents Bayes' theorem through a geometric diagram, showing how evidence should update beliefs. It works through the Kahneman and Tversky style questions and base rates.
Recommended by 2 HN commenters across 2 different months, 2022 to 2024.
What Hacker News said
Strong rec godelski · 2024-07 · on “Instrumenting Python GIL with eBPF”
Probably, but more likely a lack of knowledge and an overestimation of their knowledge. People are famously bad with statistics but people are also famously over confident in their understanding of numbers and data. I mean cloud servers really try to get some extra 9s on the end of their uptime. (People are also __terrible__ with rates of change. e.g. 90% increase as from last year). But I think an interesting one is the also literally the most common example that's in many stats textbooks, at least the ones that cover Bayesian stats. That being that if a test is 95% accurate, that it does not mean that that scoring positive on a test means that there's a 95% chance you have whatever the tes
Recommends logshipper · 2022-01 · on “Visualizing Bayes Theorem (2009)”
There's a very cool 3Blue1Brown video [0] that helps give you an intuitive feel for Bayes Theorem. [0] [this video]
Cites adhesive_wombat · 2022-03 · on “Manim – Python library for creating mathematical animations”
Wow that's cool. I've always wondered how to do something those swish YouTube videos like 3Blue1Brown[1]. [1] [this video]
Cites aeroevan · 2023-11 · on “Home schooling's rise from fringe to fastest-growing form of education”
This is the classic problem solved by Bayes: [this video] there just aren't that many homeschooled people to provide a defense. I'm another successful homeschooled-until-college adult, but homeschooling isn't that common now and it was even less popular when I was growing up. EDIT: unless you were mentioning defense of homeschooling given they were homeschooled. I would assume most people on HN who were homeschooled had a decent education/homeschooling experience.
Cites ernst_klim · 2024-04 · on “This is a teenager”
> This metric is also a proxy for living in a violent environment. Probably, probably not. The probabilities of witnessing someone being shot is extremely low in both environments. If amount of people who are living in violent environment is much lower, it may be that a person who witnessed someone being shot is more probable from a good environment. [LINK] [this video]