Negative Binomial Distribution
Negative Binomial Distribution: Trials Until the r-th Success
The negative binomial distribution describes the number of independent Bernoulli trials, X, that must be performed to obtain a fixed number r of successes, where each trial has the same probability of success p. Unlike the binomial distribution, the number of trials here is random, it is the count of successes that is fixed in advance.
Negative Binomial PMF for r = 3 successes, p = 0.4
Probability Mass Function
If X is the trial number on which the r-th success occurs, then
A possible outcome sequence while waiting for the 2nd success (r = 2, p = 0.5). The 2nd success occurs on trial x = 7, so seven trials in total were needed.
The factor (r−1x−1) counts the ways to arrange the first r−1 successes among the first x−1 trials. The very last trial, trial x, must be a success, so its probability p is written outside the combination, never inside it.
Mean and Variance
Worked Example
A basketball player makes free throws independently with probability p=0.7. What is the probability that her 3rd made shot happens on her 5th attempt?
Here r=3, x=5:
A dart player hits the bullseye independently on each throw with probability p=0.25. Let X be the number of throws required to hit the bullseye for the 4th time. Find E(X), the expected number of throws.
Give your answer in throws.
The figure shows a possible outcome sequence while waiting for the 2nd success (r=2) when trials succeed independently with probability p=0.5. Using the negative binomial formula (not the specific sequence shown), what is P(X=7), the probability that exactly 7 trials are needed to obtain the 2nd success under these conditions?
A possible outcome sequence while waiting for the 2nd success (r = 2, p = 0.5). The 2nd success occurs on trial x = 7, so seven trials in total were needed.
Negative Binomial PMF for r = 3 successes, p = 0.4