The geometric distribution

The geometric distribution

Spec 3.1–2 · Edexcel A-Level Further Maths: Revision notes

Studying this course? Keep your progress in one place.

Key facts

  • X∼Geo(p)X\sim\text{Geo}(p) is the number of trials up to and including the first success, so x=1,2,3,…x=1,2,3,\dots
  • P(X=x)=p(1−p)x−1P(X=x)=p(1-p)^{x-1}.
  • P(X>x)=(1−p)xP(X>x)=(1-p)^x, P(X≥x)=(1−p)x−1P(X\geq x)=(1-p)^{x-1} and P(X≤x)=1−(1−p)xP(X\leq x)=1-(1-p)^x.
  • E(X)=1pE(X)=\frac1p and Var(X)=1−pp2\text{Var}(X)=\frac{1-p}{p^2}.
  • The model needs two outcomes, a constant pp and independent trials.

The geometric model

The number of independent trials needed to get the first success.

Carry out independent trials, each with the same probability pp of success, and count how many trials it takes to get the first success. This is a geometric distribution, X∼Geo(p)X\sim\text{Geo}(p).

It needs three conditions: each trial has two outcomes; pp is the same every time; and the trials are independent.

XX counts the trial on which the success happens, so x=1,2,3,…x=1,2,3,\dots with no upper limit. The number of rolls of a fair die up to and including the first six is Geo(16)\text{Geo}\left(\frac16\right).

  1. 1

    Trial 1

    failure, probability 1 − p

  2. 2

    Trials 2 to x − 1

    failures, each 1 − p

  3. 3

    Trial x

    success, probability p

The first success on trial x.

A basketball player scores with probability 0.4 on each independent shot. Which variable is geometric?

The probability function

The first x−1x-1 trials fail and trial xx succeeds, so P(X=x)=p(1−p)x−1P(X=x)=p(1-p)^{x-1}.

The first success is on trial xx when the first x−1x-1 trials all fail and trial xx succeeds, so P(X=x)=p(1−p)x−1P(X=x)=p(1-p)^{x-1}. The power is x−1x-1, not xx.

Tail probabilities are simple, because X>xX>x just means the first xx trials all fail: P(X>x)=(1−p)xP(X>x)=(1-p)^x.

Learn P(X>x)=(1−p)xP(X>x)=(1-p)^x: it avoids summing a series and turns "more than" and "at least" questions into one power.

00.050.10.15123456789101112xP(X = x)
Geo(1/6): rolls needed to get the first six.
  • P(X=x)P(X=x)p(1−p)x−1p(1-p)^{x-1}
  • P(X>x)P(X>x)(1−p)x(1-p)^x
  • P(X≤x)P(X\leq x)1−(1−p)x1-(1-p)^x

Worked example

A component is defective with probability 0.080.08. Let XX be the number tested up to and including the first defective one. Find P(X=4)P(X=4), P(X>5)P(X>5) and the smallest nn with P(X≤n)>0.5P(X\leq n)>0.5.

X∼Geo(0.25)X\sim\text{Geo}(0.25). Find P(X=3)P(X=3).

Mean and variance

The mean is 1p\frac1p and the variance is 1−pp2\frac{1-p}{p^2}.

For X∼Geo(p)X\sim\text{Geo}(p): μ=E(X)=1p\mu=E(X)=\frac1p and σ2=Var(X)=1−pp2\sigma^2=\text{Var}(X)=\frac{1-p}{p^2}.

Both are in the formulae booklet, and you do not have to prove them.

A smaller pp makes success rarer, so the mean wait and the spread both grow. The standard deviation is the square root of the variance. A common slip is using 1−pp\frac{1-p}{p}: the denominator is p2p^2.

0.10.20.30.40.50.60.70.80.9124681012xyp = 1: mean 1p = 0.5: mean 2p = 0.25: mean 4p = 0.1: mean 10E(X) = 1/p
The mean wait E(X) = 1/p grows as p falls
  • Mean E(X)E(X)1p\dfrac1p
  • Variance Var(X)\text{Var}(X)1−pp2\dfrac{1-p}{p^2}

Worked example

XX is the number of rolls of a fair die up to and including the first six. Find the mean, variance and standard deviation.

X∼Geo(p)X\sim\text{Geo}(p) and E(X)=4E(X)=4. What is Var(X)\text{Var}(X)?

Modelling and applying

State the trial and success, then use X∼Geo(p)X\sim\text{Geo}(p); a series gives P(X odd)=12−pP(X\text{ odd})=\frac{1}{2-p}.

In context questions, say what counts as a trial and a success, then state X∼Geo(p)X\sim\text{Geo}(p) with its value of pp.

For P(X odd)P(X\text{ odd}) the terms p, p(1−p)2, p(1−p)4,…p,\ p(1-p)^2,\ p(1-p)^4,\dots form a geometric series with ratio (1−p)2(1-p)^2. The sum to infinity a1−r\frac{a}{1-r} gives P(X odd)=p1−(1−p)2=12−pP(X\text{ odd})=\frac{p}{1-(1-p)^2}=\frac{1}{2-p}.

When two players alternate and the first to succeed wins, the first player wins when XX is odd.

Comment on assumptions: trials must be independent and pp constant; a player who improves with practice breaks the model.

0.10.20.30.40.50.60.70.80.910.20.40.60.811.2xyp = 0.5: 2/3p = 1: 1P(X odd)
P(X odd) = 1/(2 − p): the first player wins more often than the second, and always at least half the time
  • P(X odd)P(X\text{ odd})12−p\dfrac{1}{2-p}

Worked example

Two players take turns to roll a fair die, player A first. The first to roll a six wins. Find the probability that A wins.

Two players take turns to toss a fair coin, and the first to get heads wins. What is the probability that the first player wins?

Try an exam question

Each attempt at a task succeeds with probability 0.20.2, independently. Let XX be the number of attempts up to and including the first success. Find P(X=3)P(X=3), P(X>4)P(X>4), E(X)E(X) and Var(X)\text{Var}(X).

[5 marks]

That's the notes covered.

Carry on to the next subtopic.

Try a question

Find P(X=4)\mathrm{P}(X=4).

1 mark, marked against the mark scheme

Revision notes