Course contentsShow
← Data Science
Lesson 146 of 2,1454. Discrete Probability DistributionsFree lesson

When to Use Poisson vs Other Distributions

Compare conditions for Poisson versus binomial, geometric, and negative binomial in count modeling.

When to Use Poisson vs Other Distributions

You'll learn: How to choose between Poisson and other discrete distributions (binomial, geometric, negative binomial) based on the structure of your counting problem.

The Decision Framework

You've now learned four key discrete distributions for counting events. The trick is knowing which one fits your situation. Here's how to decide:

Use Poisson when:

  • You're counting events in a continuous interval (time, space, volume)
  • You know the average rate (λ), not the number of trials
  • Events occur independently
  • Example: Number of emails arriving per hour, defects per square meter of fabric

Use Binomial when:

  • You have a fixed number of trials (n)
  • Each trial has exactly two outcomes (success/failure)
  • You're counting successes out of those n trials
  • Example: Number of heads in 10 coin flips, defective items in a batch of 50

Use Geometric when:

  • You're counting trials until the FIRST success
  • You don't know how many trials it will take
  • Example: Number of job applications until first interview

Use Negative Binomial when:

  • You're counting trials until the rth success (r > 1)
  • Like geometric, but waiting for multiple successes
  • Example: Number of sales calls until closing 5 deals

The Key Question

Ask yourself: "Do I know how many times something will happen, or am I measuring a rate over an interval?" If it's a rate over time/space with no fixed trial count, choose Poisson. If you're counting within a fixed number of attempts, choose binomial or its relatives.

Key Takeaway: Use Poisson for rate-based counting over continuous intervals; use binomial for fixed-trial success counting; use geometric/negative binomial for waiting-time problems with trial-by-trial structure.