The pragmatic approach to the problem of induction associated with Reichenbach and Herbert Feigl (1902–88), that tries to show not that inductively based conclusions will be true, nor even that they will probably be true, but instead that there can be no other better strategy for predicting the future, or generalizing from evidence: induction will do well if anything at all will. An analogy is being stranded on a desert island with the opportunity to float off a bottle with a note requesting rescue: it may not be knowable how probable it is that this action will be successful, but it may be known that it will be successful if anything is, and hence the strategy is rational.