Memoryless Property — Definition, Formula & Example
The memoryless property is the feature of a random waiting time where the time you have already waited tells you nothing about how much longer you will wait. Each moment, the process starts fresh.
A nonnegative random variable is memoryless if for all . For discrete variables the condition applies to all nonnegative integers . Equivalently, the survival function satisfies . The exponential distribution is the only continuous distribution with this property. The geometric distribution on is the only such distribution on the positive integers.
Key Formula
Where:
- = A nonnegative random waiting time, such as time until the next event
- = Time already waited, with no event yet
- = Additional time you ask about
- = Rate parameter of an exponential distribution (events per unit time)
Worked Example
Problem: The wait for a bus is exponentially distributed with a mean of 10 minutes, so . You have already waited 5 minutes. What is the probability you wait at least 10 more minutes?
Step 1: Waiting at least 10 more minutes means the total wait exceeds 15 minutes. Find that probability.
Step 2: Divide by the probability of the condition you already know, .
Step 3: Compare with a fresh start. The chance a new arrival waits more than 10 minutes is the same value.
Answer: The probability is , about 36.8%. This equals the chance a person who just arrived waits more than 10 minutes. The 5 minutes you already waited did not change the outlook.
Why It Matters
The memoryless property is the reason Poisson processes and continuous-time Markov chains are tractable. When arrival and service times are exponential, the time until the next arrival or service completion does not depend on how long you have already waited, which makes waiting-line (queueing) models tractable. It comes up in actuarial exam preparation (SOA Exam P covers the exponential distribution) and in reliability engineering, for example to find the remaining lifetime of a component that has already survived for some time.
Common Mistakes
Mistake: Computing when the problem says the item has already lasted units, and so ignoring the condition.
Correction: The memoryless property concerns the conditional probability . In the bus example, . The correct conditional answer is . Also, apply the property only to exponential or geometric variables. Uniform and normal waiting times do depend on time already elapsed.
