A method for identifying faults contributing to a data transfer error in a tape drive. The method includes calculating and comparing a long-term error rate and a short-term error sample. Based upon such calculation and comparison, a pre-selected error recovery procedure may be initialized. The long-term error rate equals a total number of bytes of data transfer error divided by a total number of bytes processed. A short-term error sample includes a number of bytes of data transfer error in a predetermined number of sequential blocks of data divided by the total number of bytes transferred in the predetermined number of sequential blocks of data. The method calculates and monitors a predetermined number of short-term error samples for a predetermined number of sequential blocks of data defining a window. Under normal operational conditions where head/media interface is free of debris, short-term error samples would have values slightly and randomly larger or smaller than long-term error rate. As debris accumulates gradually at head/media interface, the value of succeeding short-term error samples will be larger than the long-term error rate.