What are control and treatment in a model experiment?
PICTURE THIS: DATA SPLIT
Fit on train, tune on val, report on test once.
Simple meaning
Control is usually the current production model.
WHY — A/B Testing instead of guessing?
Why interviewers care about A/B Testing:
people who only read docs from people who shipped.
and tied to MLOps work.
Name the idea, why it exists, then one short example.
End with when you use it and one common pitfall.
STEPS — What happens step by step?
Before you speak the answer, walk the interviewer through these steps:
- 1Control is usually the
current production model.
- 2Treatment is the new
candidate serving a share of users or requests.
- 3Random assignment and a
pre-registered metric keep the comparison honest.
- 4Give an example
One tiny concrete case you can say aloud.
- 5Common mistake
What juniors usually get wrong.
- 6Close
When you pick this over the alternative.
EXAMPLE — See it in action
Here's a short line you can speak, broken into clear beats:
Note: Adapt this scaffold to your own project — keep it under 60–90 seconds.
Key takeaway
Control is usually the current production model. Treatment is the new candidate serving a share of users or requests.