Skip to content

How likely is a result between two values?

Binomial, Poisson and normal distributions: drag the bounds and read P(X ≤ k) or P(a ≤ X ≤ b) with the mean and variance.

Distribution

Probability of
Key values as a table
kP(X = k)P(X ≤ k)
09.537 × 10⁻⁷0.000095%
10.000019070.002%
20.00018120.02%
30.0010870.13%
40.0046210.59%
50.014792.07%
60.036965.77%
70.0739313.16%
80.120125.17%
90.160241.19%
100.176258.81%
110.160274.83%
120.120186.84%
130.0739394.23%
140.0369697.93%
150.0147999.41%
160.00462199.87%
170.00108799.98%
180.000181299.998%
190.0000190799.999905%
209.537 × 10⁻⁷100%

P(8 ≤ X ≤ 12)

73.68%

Under the binomial distribution with n = 20, p = 0.5, the chance is 73.68%; the chance of anything else is 26.32%.

Mean
10
Variance
5
Standard deviation
2.23607
Probability as a decimal
0.736824

Binomial, Poisson and normal distributions: drag the bounds and read P(X ≤ k) or P(a ≤ X ≤ b) with the mean and variance.

How to use

  1. Choose a distribution: binomial for successes in a number of tries, Poisson for counts of rare events, normal for measurements.
  2. Set its parameters, then choose the event: between two values, at most b or at least a.
  3. Type the bounds or drag the handles under the graph; the shaded bars or area are the probability shown above.

Chance of at most k heads in 10 fair coin flips

Heads (k)Probability
25.5%
317.2%
437.7%
562.3%
682.8%
794.5%
898.9%

How it works

A probability distribution says how likely each possible result is. The binomial distribution counts successes in n independent tries that each succeed with probability p, such as heads in 20 coin flips: P(X = k) = C(n, k)·pᵏ·(1 − p)ⁿ⁻ᵏ, with mean np and variance np(1 − p). The Poisson distribution counts events that happen at an average rate λ, such as calls per hour: P(X = k) = e^−λ·λᵏ / k!, with mean and variance λ.

The normal distribution is the bell curve of measurements with mean μ and standard deviation σ; for it, the probability of a range is the area under the curve. For the counts the probability of a range is the sum of its bars. In 20 fair coin flips, P(8 ≤ X ≤ 12) is the sum of the bars from 8 to 12, about 73.7 %; a normal curve with the same mean 10 and σ = √5 gives a similar answer, which is why the normal approximation works for large n.

The answers assume the model fits: independent tries with a fixed p, events at a steady rate, or data that really is bell-shaped. Definitions follow the NIST/SEMATECH e-Handbook of Statistical Methods, sections 1.3.6.6.1, 1.3.6.6.18 and 1.3.6.6.19.

Frequently asked questions

When should I use the Poisson distribution instead of the binomial?

When you count events over time or space with no fixed number of tries, such as emails per day. For many tries with a small p, the binomial is close to a Poisson with λ = np.

Why do the bounds count as included?

For counts, P(8 ≤ X ≤ 12) includes 8 and 12. Use 9 and 11 to leave them out. For the normal distribution a single value has zero probability, so it makes no difference.

What is the probability of exactly k?

Choose a ≤ X ≤ b and set both bounds to k: the one shaded bar is P(X = k). The table under the graph lists P(X = k) for the values around the mean.

How do I get P(X > k) or P(X < k)?

For counts, P(X > k) is P(X ≥ k + 1) and P(X < k) is P(X ≤ k − 1).

Do I enter 50 or 0.5 for a 50% chance?

Enter 0.5 in the binomial chance p field. It accepts probabilities from 0 to 1, not percentages from 0 to 100.

Send feedback