We’ve all heard the advice: if you can’t fall asleep, count sheep.
But if you’re a Six Sigma Black Belt or someone deep into Continuous Improvement, you might find yourself counting something else—Lambdas (λ).
That’s the inspiration behind this cartoon I drew early in my Six Sigma Black Belt training. Instead of fluffy sheep hopping fences, it’s “One Lambda… Two Lambda…” marching through the mind of a problem-solver at 2:00 a.m. Truth was, I was trying to understand why I (and anyone serious about improvement) needed to care about λ?
What is Lambda?
In statistics, λ (lambda) often represents the average rate of occurrence in a Poisson distribution. (See below for a reminder of what exactly that is.)
-
Think of customer arrivals at a hotel check in service desk.
-
Defects showing up in a production line.
-
Calls coming into a support center.
When events happen randomly but at a known average rate, λ is the key that helps us model and predict.
Why It Matters for CI Practitioners
Understanding λ allows leaders to:
-
Predict demand and workload – Are we staffed for the rate at which calls or requests arrive?
-
Quantify rare events – How often should we expect defects, errors, or delays?
-
Improve process design – By modeling arrival or defect rates, we can redesign systems for flow, balance, and resilience.
In other words, λ helps us move from guessing about variation to managing it.
A Practical Example
Imagine your call center averages 100 calls per hour. That’s your λ. Using the Poisson distribution, you can answer practical questions like:
-
What’s the probability of getting 150 calls in the next hour?
-
How often will we see zero calls in a 10-minute span?
-
What’s the chance of getting swamped with 20+ calls at once?
With λ in hand, we can make data-driven staffing and scheduling decisions instead of reacting in the moment.
Bringing It Back to the Cartoon
The cartoon is funny because it’s true—Black Belts often can’t turn off the problem-solving brain, even at night. Instead of counting sheep, we count the things that matter to processes: arrivals, defects, and opportunities for improvement.
So the next time you’re lying awake, try it:
“One Lambda… Two Lambda…”
And if it doesn’t put you to sleep, at least it might give you an idea for your next project.
Question for Readers:
Where have you used λ in your work? Was it for arrivals, defects, or something more creative?
A Quick Recap: What’s the Poisson Distribution?
The Poisson distribution is a statistical tool used to model the probability of a certain number of events happening in a fixed period of time (or space), given a known average rate (λ).
It’s especially useful when events:
-
Happen independently (one event doesn’t affect the next),
-
Occur at a constant average rate, and
-
Can happen any number of times (including zero).
Examples:
-
Calls coming into a call center per hour,
-
Cars arriving at a toll booth per minute,
-
Defects found on a length of fabric,
-
Patients arriving at an emergency room.
The beauty of the Poisson is that it answers practical questions like:
-
“What’s the chance we’ll get 0 calls in the next 10 minutes?”
-
“How likely is it we’ll get 12 patients in an hour?”
-
“What’s the probability of finding 3 or more defects in the next roll of material?”
In short: Poisson + λ = predictability in the unpredictable.
Drop a note in the comments on how you feel about using Lambda in your prediction modeling.