In 30 seconds
This 3Blue1Brown short shows how to calculate the sum of two independent weighted dice by multiplying outcome probabilities and adding along diagonals: a discrete convolution.
- 1GridA grid lists all outcome pairs.
- 2Pair probabilityFor independent dice, multiply face probabilities.
- 3Equal sumsEqual sums lie on diagonals.
- 4Add diagonalsAdd pair probabilities on each diagonal.
- 5ConvolutionThe resulting distribution is a convolution.
Takeaways
- A grid lists all outcome pairs.
- For independent dice, multiply face probabilities.
- Equal sums lie on diagonals.
- Add pair probabilities on each diagonal.
- The resulting distribution is a convolution.
Read the edited text by chapter
Weighted dice
List and weight outcome pairs.
Suppose I roll two weighted dice and want the probability of each possible sum. This short illustrates the computation for independent rolls; the full video linked in its description extends the idea to continuous distributions.
I start with a six-by-six grid whose cells represent all pairs of outcomes. The dice need not be fair, so each face can have its own probability.
For independent dice, the probability of a pair is the probability of the first face multiplied by the probability of the second. If the rolls are dependent, a joint distribution is needed instead.
Diagonal sums
Combine pairs with equal totals.
The probabilities can be shown as heights in a three-dimensional plot. Pairs with the same sum lie on the same diagonal of the grid.
For example, one diagonal collects outcomes summing to six and another collects outcomes summing to seven. Adding the heights along each diagonal gives the probability of that sum.
Collapsing the plot along the diagonals produces the full distribution of the sum. This operation is discrete convolution of the two independent dice distributions.
More from this channel
40:44EN
3Blue1Brown· Oct 16, 2026Grant Sanderson: There's never been a better time to go into mathGrant Sanderson, Constantin Kogler · 3Blue1BrownStudents fear AI has solved mathematics. Grant Sanderson argues the opposite: even if machines prove theorems better than we do, the institution of math stays valuable—and this is a rare moment when newcomers can reshape it.
51:55EN3Blue1Brown· Sep 18, 2026The last IMO problem AI could not solveGrant Sanderson · 3Blue1BrownGrant Sanderson walks through IMO 2025 Problem 6 — a tiling puzzle that frontier models missed — and the visual patience it takes to prove a construction is optimal.
0:55EN3Blue1Brown· Jul 24, 2026The 64 Sugar Cubes Puzzle: Six Equal DistancesA 4×4×4 block has three colors; prove that six same-colored pairs share a distance.
33:51EN3Blue1Brown· Jul 16, 2026What Cross-Entropy Connects: Compression, Language Models and KL DivergenceGrant Sanderson · 3Blue1BrownGrant Sanderson builds cross-entropy from code lengths and uses it to explain next-token training and distillation.
2:25EN3Blue1Brown· Jun 12, 2026How Shannon Measured the Entropy of EnglishA guessing game helped Claude Shannon estimate how much information English text contains, and why prediction is central to compression.
1:02EN3Blue1Brown· Jun 10, 2026A Robot Walk and the Puzzle of Optimal Binary EncodingHow few bits does it take to send directions that occur with unequal frequencies?