In past decades, many empirical studies revealed return anomalies in many different asset classes and markets. Very recent publications have, however, even found evidence that stock markets react to the results of soccer matches. In this article, we argue that such empirical studies should be analysed carefully; we thus endorse the use of replication studies to verify results. Consequently, by rebuilding the study of Ashton et al. (2003), we are able to detect mistakes in the empirical set-up. Based on these findings, we demonstrate how even minor flaws can have a crucial influence on the results of such studies and point out pitfalls that are frequently encountered. We furthermore emphasize the importance of robustness checks to validate the results of empirical studies.