Poisson Distribution Calculator screenshot

Poisson Distribution Calculator

Poisson point, lower-tail and upper-tail probabilities plus the mean and standard deviation.

Privacy First
0 Runs
Updated 2026-11-06
No upload
Runs in this tab
No signup

What this page computes

Type the rate lambda, the average number of events per interval, and the count k. The point probability is e^(−lambda)lambda^k/k!, computed as exp(k ln lambda − lambda − ln k!) so a large k does not overflow; the page adds P(X ≤ k) and P(X ≥ k), names each tail, and prints the mean and variance (both lambda) with the standard deviation. Figures print to six significant figures.

The Poisson model assumes events arrive independently at a constant rate, which is why it has a single parameter; the page does not check that assumption. A rate at or below zero is refused, and so is a rate above 1000, where the direct sum is outside the range this page evaluates accurately. A negative count is refused.

Everything runs in the browser: the same lambda and k give the same probabilities after any reload.

The page evaluates the Poisson distribution for a rate λ: P(X = k) = e^(−λ)λ^k/k!, with the lower tail P(X ≤ k), the upper tail P(X ≥ k), and the mean, variance and standard deviation all printed. At λ = 3 and k = 2 the point mass is 0.224042, the lower tail 0.42319 and the upper tail 0.800852. A rate of zero or below is refused, and so is a rate above 1000, past which the summation stops being accurate.

Get Poisson point and tail probabilities from a rate. The mean equals the variance, and both are printed.

How to use it

Type the rate λ, the average number of events in one interval, and the number of events k. λ must be above zero and at most 1000; k cannot be negative. Each tail is labelled, so a question about “at least k events” is answered by the upper-tail line directly.

One parameter, and the overlapping tails

The two tails overlap at k, so their sum is 1 + P(X = k) rather than 1 — the same overlap the binomial page has, because both are discrete. Poisson has one parameter, and the mean and variance are both λ, which is why the distribution is described as equidispersed: if real counts vary more than their mean, this is the wrong model. Nothing here is fitted to data; the rate is taken as given.

What to Expect

Best for

  • Counting events in a fixed interval with a known rate

What this tool keeps

  • Nothing leaves this tab: the arithmetic runs in the page you are reading, and a reload discards both the input and the result.

Sample output

P(X = 2) = 0.224042 P(X ≤ 2) = 0.42319 (lower tail) P(X ≥ 2) = 0.800852 (upper tail) mean = variance = λ = 3, sd = 1.73205 k = 2 events in one interval

Frequently asked questions

Do the tails add to 1?

No. They overlap at k, so P(X ≤ k) plus P(X ≥ k) is 1 + P(X = k) — 0.42319 and 0.800852 at λ = 3 and k = 2. Each line names its tail so neither is read as the other.

Why is a rate above 1000 refused?

Because past that point the summation behind the cumulative probabilities stops being accurate, and the page would rather refuse than print a number it cannot stand behind.

Is anything uploaded?

No. All three probabilities are evaluated in this page from the two numbers you typed, and nothing is posted to a server or stored.

Developer program

Build custom tools for your store

Join the SDK program to create useful ecommerce utilities for your store.