What are the mean and variance of a Poisson distribution?

For X following a Poisson distribution with parameter lambda, both the mean and variance are lambda, and the standard deviation is the square root of lambda.

Is Poisson distribution the same as binomial distribution?

No. Binomial distribution counts successes in a fixed number of trials, while Poisson distribution counts occurrences in a fixed interval. Poisson can also approximate a binomial distribution when the number of trials is large and the success probability is small.

Mathematics • Statistics and Probability

What Is Poisson Distribution?

Q: What is Poisson distribution?

Poisson distribution is a discrete probability distribution used to model the number of times an event occurs in a fixed interval of time, space, area or volume when the event rate is known and the occurrences are suitably independent.

Q: What is the Poisson distribution formula?

P(X=k)=e^(-lambda) lambda^k/k!, where lambda is the average number of occurrences in the interval and k is a non-negative integer.

Q: Where is Poisson distribution used in real life?

It can be used for suitable count data such as customer arrivals, phone calls, website requests, defects, accidents, insurance claims, equipment failures and other events counted in a fixed interval.

Poisson distribution is a powerful way to study how many times an event occurs in a fixed interval. It helps us answer practical questions such as: How many customers may arrive in the next hour? How many calls might a call centre receive? How many defects may appear in a batch?

In simple words: If we know the average number of times an event happens in a certain interval, the Poisson distribution helps us estimate the probability of getting 0, 1, 2, 3, ... occurrences in that interval.

1. What Is Poisson Distribution?

Poisson distribution is a discrete probability distribution. It is used when we are counting the number of occurrences of an event in a specified interval of time, length, area, volume or another suitable region.

For example, suppose a website receives an average of 12 support requests every hour. We may want to know the probability that it receives exactly 15 requests in the next hour. This is a typical Poisson-type question.

The distribution is named after the French mathematician Siméon Denis Poisson. In standard notation, we write:

X\sim\operatorname{Poisson}(\lambda)

Here, (lambda) is the average number of occurrences in the interval.

2. The Main Idea Behind Poisson Distribution

Think about events that occur one at a time and are counted over a fixed interval. Examples include:

number of customers entering a shop in 10 minutes,
number of calls received by a help desk in one hour,
number of typing errors on a page,
number of defects in a metre of manufactured material,
number of website requests arriving in one minute,
number of accidents at an intersection in a month.

The important question is not when exactly an event occurs, but how many occurrences happen in the chosen interval.

Key idea: Poisson distribution is a model for counts of events within a specified interval when the average rate is known and the assumptions of the model are reasonably appropriate. It is not a model that should automatically be applied to every count.

3. Poisson Distribution Formula

The probability of exactly occurrences is:

P(X=k)=\frac{e^{-\lambda}\lambda^k}{k!},\qquad k=0,1,2,\ldots

where:

Symbol	Meaning
	Random variable representing the number of occurrences
	The particular number of occurrences we want
	Average number of occurrences in the chosen interval
	Euler's number, approximately 2.71828
	Factorial of

4. Conditions for Using the Poisson Distribution

A Poisson model is appropriate when the following assumptions are reasonably satisfied:

We count occurrences. The random variable is a count: 0, 1, 2, 3, ...
The interval is fixed. We specify a period of time, area, length, volume or another relevant interval.
The average rate is stable. The expected number of events per interval is approximately constant for the situation being modelled.
Occurrences are suitably independent. One occurrence does not substantially determine another occurrence in the model.
Events are not occurring in large simultaneous clusters. The ordinary Poisson process is intended for individual occurrences rather than strongly clustered events.

These are modelling assumptions. Real-world data can violate them, so the Poisson distribution should be checked against the context rather than used mechanically.

5. Mean, Variance and Standard Deviation

If:

X\sim\operatorname{Poisson}(\lambda)

then a particularly useful property is that the mean and variance are both equal to :

E(X)=\lambda

\operatorname{Var}(X)=\lambda

Therefore, the standard deviation is:

\sigma=\sqrt{\lambda}

Example: If a system receives an average of 16 requests per minute, then . The mean is 16, the variance is 16 and the standard deviation is .

6. A Concrete Example: Customers Arriving at a Shop

Suppose a small shop receives an average of 6 customers in 10 minutes. We want the probability that exactly 4 customers arrive in a randomly selected 10-minute interval.

Here:

\lambda=6,\qquad k=4

Using the Poisson formula:

P(X=4)=\frac{e^{-6}6^4}{4!}

Now:

P(X=4)=\frac{e^{-6}\times1296}{24}\approx0.1339

So the probability is approximately 0.134, or about 13.4%.

This does not mean exactly 4 customers must arrive every time. It means that, under the model assumptions, an interval with exactly 4 customers has a probability of about 13.4%.

7. Example: Website Traffic

Suppose a website receives an average of 20 requests per minute. We want the probability of exactly 18 requests in the next minute.

\lambda=20,\qquad k=18

P(X=18)=\frac{e^{-20}20^{18}}{18!}\approx0.0847

So the model gives a probability of about 8.47%.

This idea can help technology teams think about expected request loads and capacity. In real systems, however, traffic may be bursty or correlated, so engineers may need more advanced models when the Poisson assumptions do not fit.

8. Example: Telephone Calls

A customer-support centre receives an average of 3 calls every 10 minutes. What is the probability of receiving exactly 5 calls in the next 10 minutes?

\lambda=3,\qquad k=5

P(X=5)=\frac{e^{-3}3^5}{5!}\approx0.1008

Therefore, the probability is approximately 10.08%.

A company could use this type of analysis when studying call volumes, staffing requirements and queueing behaviour.

9. Example: Manufacturing Defects

Imagine that quality-control records show an average of 2 defects per 100 metres of a particular material. If the production conditions remain comparable, the Poisson model can be used to estimate the probability of finding a particular number of defects in a 100-metre section.

For exactly 3 defects:

\lambda=2,\qquad k=3

P(X=3)=\frac{e^{-2}2^3}{3!}\approx0.1804

So the probability is approximately 18.04%.

Manufacturers can use count-based probability models as one part of quality-control analysis, especially when defects are relatively rare and the assumptions are reasonable.

10. Example: Road Accidents

Suppose historical records for a particular road segment show an average of 2 accidents per month. If the situation is sufficiently stable and the events are reasonably independent, a Poisson model can describe the number of accidents in a month.

The probability of no accidents in a month is:

P(X=0)=\frac{e^{-2}2^0}{0!}=e^{-2}\approx0.1353

So the model gives about a 13.53% probability of zero accidents in one month.

This type of analysis can support risk assessment, although real accident data may involve weather, traffic volume, road conditions and other factors that can make a simple Poisson model inadequate.

11. Example: Emails Received by a Teacher

Suppose a teacher receives an average of 8 student emails per day. If we treat the daily count as a suitable Poisson-type process, we can estimate the probability of receiving exactly 10 emails on a particular day.

P(X=10)=\frac{e^{-8}8^{10}}{10!}\approx0.0993

The probability is about 9.93%. This is a useful classroom example because it shows that Poisson distribution is about counting occurrences, not about the size or content of the emails.

12. What If the Time Interval Changes?

The value of must be adjusted when the interval changes.

Suppose a help desk receives an average of 12 calls per hour. Then the average number in 30 minutes is:

\lambda_{30\,\text{min}}=12\times\frac{30}{60}=6

For 15 minutes:

\lambda_{15\,\text{min}}=12\times\frac{15}{60}=3

So a common mistake is to use 12 as even when the question asks about a different time interval.

Rule: Convert the average rate to the same interval mentioned in the question before applying the Poisson formula.

13. Probability of Zero, At Least and At Most

The Poisson formula gives the probability of an exact count. Other questions can be handled by adding probabilities or using the complement.

Exactly k

P(X=k)=\frac{e^{-\lambda}\lambda^k}{k!}

Zero occurrences

P(X=0)=e^{-\lambda}

At most k

P(X\le k)=\sum_{r=0}^{k}\frac{e^{-\lambda}\lambda^r}{r!}

More than k

P(X>k)=1-P(X\le k)

At least k

P(X\ge k)=1-P(X\le k-1)

For example, “more than 3” means 4, 5, 6, ... and is often easiest to calculate using the complement.

14. Where Is Poisson Distribution Used in Real Life?

Field	Possible count being modelled	Why it is useful
Retail	Customers arriving per hour	Helps study demand and staffing
Call centres	Calls arriving per minute	Supports queue and staffing analysis
Internet & technology	Requests arriving at a server	Helps reason about workload and capacity
Manufacturing	Defects in a fixed length or area	Supports quality-control analysis
Transport	Accidents in a road segment or period	Helps quantify event counts and risk
Insurance	Claims in a time period	Useful for modelling claim counts under suitable assumptions
Healthcare	Occurrences of certain events in a defined population/time window	Can support statistical modelling and resource planning
Biology	Counts of events in a defined area or sample	Useful when event-count assumptions are appropriate
Education	Errors or requests counted over a fixed interval	Provides a practical way to study count variation
Telecommunications	Calls or messages arriving in a time interval	Helps analyse traffic and service capacity

15. Poisson Distribution in Modern Technology

Poisson-type models are particularly useful when technology systems generate large numbers of events that can be counted over short intervals.

Web servers: requests arriving at a server can be treated as count data over a chosen interval when the model assumptions are reasonable.
Cloud systems: service requests and other event counts can be analysed for capacity planning.
Network traffic: packet or connection arrivals can sometimes be approximated by Poisson processes over suitable intervals, although modern networks often show burstiness and dependence.
Reliability engineering: failures or incidents can be counted over operating time when a constant-rate model is appropriate.
Data science: Poisson regression extends the basic idea to situations where the expected event count depends on explanatory variables.

The important lesson is that Poisson is a model, not a statement that every technology event occurs randomly in exactly the same way.

16. Poisson Distribution and Binomial Distribution

Both distributions deal with counts, but they answer different kinds of questions.

Feature	Binomial	Poisson
What is counted?	Number of successes in a fixed number of trials	Number of occurrences in a fixed interval
Number of trials	Fixed	Not the central defining quantity
Parameters
Possible values
Typical example	Number of defective items among 100 tested items	Number of defects found along a fixed length of material

Poisson distribution can approximate a binomial distribution in suitable rare-event situations when the number of trials is large and the success probability is small. A common approximation uses:

\lambda=np

17. Why Is Poisson Distribution So Useful?

Its usefulness comes from a simple question that appears in many areas of life:

“If something happens at an average rate, what is the probability that it happens this many times in the next interval?”

The model turns an observed average rate into a complete probability distribution for possible counts. That makes it useful for planning, risk analysis, capacity decisions, quality control and statistical inference.

18. Common Mistakes Students Make

Using the wrong interval: If the average is per hour but the question asks about 20 minutes, first convert the mean to 20 minutes.
Confusing the mean with the requested count: is the average; is the particular count.
Forgetting the factorial: The denominator contains .
Using a continuous distribution for a count: Poisson is discrete.
Assuming every count is Poisson: The model assumptions and context matter.
Forgetting the complement: Questions such as “more than 4” are often easier as .

19. Quick Revision

Concept	Formula / idea
Distribution
Exact probability
Mean
Variance
Standard deviation
Zero occurrences
At most k
More than k
Binomial approximation	when rare-event conditions are appropriate

20. Final Takeaway

Poisson distribution is one of the most useful discrete probability models for studying the number of occurrences of an event in a fixed interval. Once we know the average rate , the formula

P(X=k)=\frac{e^{-\lambda}\lambda^k}{k!}

allows us to estimate the probability of different counts.

From customers entering a shop and calls arriving at a help desk to defects in manufacturing and requests reaching a server, the same mathematical idea appears in many fields. The real skill, however, is not just substituting numbers into a formula—it is recognising when the Poisson model is appropriate.

Frequently Asked Questions

What is Poisson distribution in simple words?

It tells us the probability of getting a certain number of events in a fixed interval when the average event rate is known and the model assumptions are suitable.

Why is lambda used?

represents the average number of occurrences in the interval being studied.

Why are mean and variance equal?

This is a distinctive mathematical property of the Poisson distribution: both are equal to its parameter .

Can Poisson distribution be used for website traffic?

It can be a useful starting model for event counts such as requests per minute when the assumptions are reasonably satisfied. Real web traffic can be bursty and dependent, so more detailed models may be necessary.

Is Poisson distribution continuous?

No. It is a discrete probability distribution because it counts whole-number occurrences: 0, 1, 2, 3, and so on.

Sources and Further Reading

This article is written as an educational explanation based on standard probability theory. For further study, see the OpenStax treatments of Poisson distribution and discrete probability distributions.