Question: To what degree can we prove the "even handedness" of coin tossing?
The Experiment: Lets toss a coin 100 times. We get a certain number of heads, most likely between 40 and 60, but occasionally outside these limits.
We refer to these 100 tosses as "The Experiment".
Each time we do "The Experiment" (click the "Toss -->" button) we come up with another answer.
In the real world, we usually only do an experiment once.
Public opinion polls are taken once. These are expensive, and the news organizations that take them wait some time before taking another, and try to deduce the trend from the two results.
Medical experiments are usually done once, again because of the extensive time and expense involved. On rare occasions they are repeated, when the results have a crucial bearing on the efficacy of a particular medication.
And while we're all interested in gas prices, the precision of the result isn't important enough to justify doing them more than once every few days!
The advantage of doing experiments on a computer ("computer simulation") is that we can repeat the experiment many times, and determine the distribution (i.e., the range of values) rather than a single final result!