Consider the experiment of tossing a single coin. Two outcomes are possible, viz. obtaining a head or obtaining a tail. The probability that it is a tail is 1/2, i.e. 0.5. This probability is determined without an experiment based on the principle that each of the possible outcomes must be equally likely. In reality it may not be that, for every two tosses there will be one tail. But if the number of tosses is increased, however, the actual results will approximate more closely to the one-in-two pattern. Thus, in 2000 tosses there may be 998 tails. If the probability of a tail being tossed is one in two, it does not follow that, in order to maintain the probability ratio, the next toss will produce a head.
As per the Classical approach, the probability is the ratio of the number of equally likely possible outcomes favorable for an event to the total number of possible outcomes. If there are m possible outcomes that favor the occurrence of event A and there are n total possible outcomes, then the probability of the occurrence of event A is the ratio of m to n (m/n). The possible outcomes favorable for an event and total number of outcomes must be known without performing experiments.