Compare experimental probability, based on collected data, to theoretical probability, calculated from equally likely outcomes.
Theoretical probability is calculated ahead of time from equally likely outcomes, like P(heads) = 1/2 for a fair coin. Experimental probability instead comes from actually trying something and counting: (number of times the event happened) / (total number of trials).
The two usually do not match exactly, especially with few trials -- that difference is normal random variation, not a mistake. As the number of trials grows very large, experimental probability tends to drift closer and closer to the theoretical probability.
Mastery looks like: They correctly compute an experimental probability from trial data and explain how it relates to the theoretical probability.
Common stumbles: Assuming experimental and theoretical probability must match exactly; treating a small-sample mismatch as a sign of a broken or unfair device.