The Mathematical Theory of Probability has, as is well known, two main branches:
1°. Direct or Deductive Probability, in which the behaviour of a system is deduced from a knowledge of its constitution.
Thus the problem “What is the probability of throwing a double-six in a single cast of two dice” belongs to deductive probability, since we know the constitution of the dice–that they are symmetrical six-sided bodies–and from this knowledge we deduce the probability of a double-six.
And 2°. Inverse or Inductive Probability, in which the constitution of a system is inferred from its observed behaviour.
As an example of Inductive Probability, suppose we visit a rifle-range and find a man firing at the target. At first we know nothing regarding his skill as a marksman ; but after watching a few shots we are able to form a rough estimate of it, and this estimate becomes more sharply defined and more confident the longer we observe his performance , so that ultimately we attain to a trustworthy knowledge of his “constitution” as a marksman, i.e. we can assign the probability that his next shot will hit or miss.