5 ideas
Francesco, Marquis of Saluzzo, was Francis I's lieutenant in Italy and owed the French king his rank. In 1536, frightened by prophecies then circulating that Charles V would win and France would fall, he went over to the Emperor. Montaigne uses him to show that a forecast can wreck a man who believes it: Saluzzo gave up his loyalty and his position because of predictions, not because of the forces actually facing him.
On Samothrace, people showed Diagoras the painted votive tablets left by sailors the gods had saved from shipwreck, as proof that prayer works. He answered that the tablets of those who drowned were not there, because the dead do not dedicate pictures. The evidence for a prediction or providence is gathered only from its survivors, so it proves nothing.
People who study almanacs cite them because the almanacs are sometimes right. Montaigne answers that anyone who predicts enough things will hit some by chance. The hits get remembered and repeated as marvels, while the far more numerous misses are quietly forgotten. A forecaster's reputation therefore comes from volume and selective memory, not from any real knowledge of the future.
Oracles and prophetic writings are deliberately vague, full of riddles and figures. That vagueness works in their favor, because the clever reader reads their own meaning into the text afterward and credits the oracle with it. Treat a prophecy's apparent accuracy as a measure of how ambiguous it is and how inventive its interpreters are, not as a measure of its foresight.
Montaigne quotes Lucan: let the future come suddenly, let men be blind to fate, and let the fearful keep their hope. Knowing what is coming would add dread to misfortunes we cannot avoid and destroy the hope that lets us act. Since the future cannot be known, the sensible course is to stop consulting signs and act on present judgment, accepting that the outcome is not ours to read in advance.