Distributions.

Tune μ and σ; watch the bell shift and stretch against samples drawn live.

Count rare events with λ — and watch the bars converge to a Normal as λ grows.

n coins, probability p — see the discrete count distribution shift and stretch toward Gaussian.

Waiting times between Poisson events — and the only continuous memoryless distribution.

A distribution over probabilities. The Bayesian conjugate prior for coin flips.

Looks like a Normal but has no mean — the running average never settles.

Count flips until the first head. The only discrete memoryless distribution.

Multiply independent shocks and the product goes log-normal. Income, wealth, viral spreads.

Every value equally likely — the flattest distribution. Yet sums of uniforms still go Gaussian.

Waiting time for k Poisson events. Shape k controls skew; rate λ controls scale.

One flip, two outcomes. The atom every other discrete distribution is built from.

Flip until the r-th success. Dispersion that Poisson can’t match.

Sampling without replacement. Binomial in the limit as the population grows.

Heavier tails than Normal at small ν. Shrinks to a Gaussian as degrees of freedom grow.

Sum of k squared standard normals. The engine of variance tests.

Ratio of two chi-squareds. The ANOVA distribution.

Failure times. Shape k flips the hazard rate from falling to rising.

Power-law tails. The 80/20 rule lived on a log-log axis.

Two exponentials back-to-back. The prior behind L1 regularization.

The CDF everyone secretly uses. Lighter centre, heavier tails than Normal.

The distribution of maxima. Extreme-value theory in one curve.

Min, mode, max — a cheap stand-in when that is all you know.

Two bells, one weight. Unmixes into components you can steer by hand.