Uploaded November 2026 | Updated November 2026, 4 days ago
Why on earth do we divide by n−1 instead of n when we estimate the variance of a distribution?
In this 2-minute visual explanation, you’ll see where Bessel’s correction comes from. It doesn’t start from a formula to memorize. It comes from a picture you can see.
Here’s what we cover:
- Why variance is the “width” of a distribution
- Why the sample mean always makes the variance come out too small
- How to measure variance without knowing the true mean, using squared distances between pairs of points
- The 4×4 table of distances, with its pesky zeros on the diagonal
- Why removing those zeros turns 4×4 into 4×3, and where the n−1 comes from
Spoiler: the correction was hiding in the table all along.
If this clicked for you, share it with a friend who’s studying statistics, and subscribe for more math and machine learning explained visually.
Check out my book, Grokking Machine Learning: manning.com/books/grokking-machine-learning-second-edition
More from Serrano Academy: youtube.com/c/LuisSerrano
Animations made with Manim.
#statistics #variance #besselscorrection #math #machinelearning #datascience #serranoacademy
Why on earth do we divide by n−1 instead of n when we estimate the variance of a distribution?
In this 2-minute visual explanation, you’ll see where Bessel’s correction comes from. It doesn’t start from a formula to memorize. It comes from a picture you can see.
Here’s what we cover:
- Why variance is the “width” of a distribution
- Why the sample mean always makes the variance come out too small
- How to measure variance without knowing the true mean, using squared distances between pairs of points
- The 4×4 table of distances, with its pesky zeros on the diagonal
- Why removing those zeros turns 4×4 into 4×3, and where the n−1 comes from
Spoiler: the correction was hiding in the table all along.
If this clicked for you, share it with a friend who’s studying statistics, and subscribe for more math and machine learning explained visually.
Check out my book, Grokking Machine Learning: manning.com/books/grokking-machine-learning-second-edition
More from Serrano Academy: youtube.com/c/LuisSerrano
Animations made with Manim.
#statistics #variance #besselscorrection #math #machinelearning #datascience #serranoacademy
