Any attempt to measure or quantify the human tendency to err must fail, because fallibility cannot exist in degrees. Fallibility is a binary concept: a fallible entity is limitlessly error prone, whereas an infallible entity can never err. Every attempt to measure the verisimilitude or credence or confidence of a theory is revealed, by this binary logic of fallibility, to be fundamentally misconceived. For such an attempt to succeed, it would need to establish a method not only for measuring error proneness in the sciences—the usual context for such attempts—but also for measuring the verisimilitude or the credence or the confidence of the method itself, which introduces an infinite regress. Ironically, these vain attempts merely testify—in both their explicit aims and their inability to achieve those aims—to the human fallibility that they implicitly deny.