Key idea: statistical processing

Although random behavior is a priori subject to unpredictable variations, we are nevertheless able to provide information on this type of phenomenon. The main idea is that this information will be provided by repeating the experiment. We can study the frequency of appearance of each result, the average value of these results and the oscillations around this average value. All this data is processed statistically.

Experience shows that, when we observe a large number of random phenomena, we generally detect laws governing the results, which are completely determined and stable. For example, regardless of where you toss an unrigged coin, if you toss it 1,000 times, you'll get about 50% heads and 50% tails. Similarly, if we study the size distribution of a group of individuals, whatever the sample taken from that group, we'll always have a distribution curve of the...