Abstract This paper compares Fractional, Geometric, Arithmetic, Harmonic, and Network-Based schemes for allocating coauthorship credits. Each scheme is operationalized to be flexible in producing credit distribution by changing parameters, and to incorporate a special situation where the first and corresponding authors are assigned equal credits. For testing each scheme, empirical datasets from economics, marketing, psychology, chemistry, and medicine, were collected and errors in how each scheme approximates empirical data was measured. Results show that Harmonic scheme performs best overall, contrary to some claims of preceding studies in support of Harmonic or Network-Based models. The performance of a scheme, however, seems to heavily depend on empirical datasets and flexibility of the scheme, not on its innate feature. This study suggests that the comparison of coauthorship credit allocation schemes should be taken with care.