Memoryless Property of the Geometric Distribution
You'll learn: How the geometric distribution "forgets" past failures when predicting the next success.
What Does "Memoryless" Mean?
The geometric distribution has a fascinating property: no matter how many times you've already failed, the probability of success on your next trial stays exactly the same. Past failures don't affect future probabilities.
Mathematically, this means:
P(X > n + k | X > n) = P(X > k)
In plain English: "Given that you've already waited n trials without success, the probability you'll wait at least k more trials is the same as if you were starting fresh."
Why This Happens
Remember, each trial in a geometric process is an independent Bernoulli trial with fixed probability p. Independence means each coin flip, each customer arrival, or each product test doesn't "know" or "care" what happened before. The process has no memory.
Real-World Analogy
Imagine you're flipping a fair coin, waiting for heads:
- You flip 5 tails in a row. What's the probability the next flip is heads? Still 0.5.
- You flip 100 tails in a row (!). What's the probability the next flip is heads? Still 0.5.
The coin doesn't "remember" those 100 tails. It doesn't become "due" for heads. Each flip is independent.
This contrasts sharply with real-world situations like waiting in line (where each person served gets you closer to the front) or drawing cards without replacement (where each card drawn changes the deck).
Key Takeaway: The geometric distribution is memoryless because each trial is independent—past failures provide zero information about when the first success will occur, making the "fresh start" probability identical no matter how long you've already waited.