Experimental probability is the probability that an event occurred in the duration of an experiment. It is calculated by dividing the number of event occurrences by the number of times the trial was conducted.
For example, if a coin is flipped 1,000 times, and the result is tails 530 times, the experimental probability of flipping tails is 530/1000, which is 0.53. Theoretical probability, on the other hand, is reached by taking the two possible outcomes and dividing them. The theoretical probability of flipping tails is then 1/2, which is 0.50.
Experimental, sometimes referred to as empirical, probability is often confused with theoretical probability. Theoretical probability is the probability reached by dividing the number of favorable outcomes by the number of possible outcomes. While theoretical probability is very useful, there is often not enough data to calculate. In such cases, experimental probability is the proper substitute if and when the sample size of the experiment is large enough.
Experimental probability is not very useful in experiments with a very small sample size. However, it has the benefit of not needing to rely on assumptions, as it is calculated directly from empirical results. This makes it especially useful in experiments wherein it is necessary to determine if events are statistically independent.