The Bernoullis and Thomas Bayes: Probability and scientific method
Even so, not every conclusion yielded by probable reasoning is acceptable; only some have degrees of certainty or probability which are worth having. If, therefore, scientific method is to permit probable reasoning, we need some account of when and why such reasoning is acceptable. For this purpose we must find ways of measuring or 'weighing' degrees of probability which will show how a conclusion should be qualified. Huygens himself contributed to such an understanding when, in 1657, he published what is recognised as the first textbook treatment of quantitative probabilistic reasoning in games of chance. He was not attempting to quantify degrees of probability; indeed, words customarily used to signify the probability concept are absent from his tract. Nevertheless he, like us, made those connections between chance and probability that we find so natural.