Recently I have some experiences in mistakes in the others' papers. One instance was that the formula was simply wrong, and another instance was that the method was numerically impossible to implement. Since both works have been officially published, I am wondering how often do this occur in published papers.
Do people really believe that published works are error-free?