Recent types of train protection systems such as ATC require the large amounts of low-level configuration data compared to conventional types of them. Hence management of the configuration data is becoming more important than before. Because of this, the authors developed an error-proof system focusing on human operations in the configuration data management. This error-proof system has already been introduced to the Tokaido Shinkansen ATC data management system. However, as effectiveness of the system has not been presented objectively, its full perspective is not clear. To clarify the effectiveness, this paper analyses error-proofing cases introduced to the system, using the concept of QFD and the error-proofing principles. From this analysis, the following methods of evaluation for error-proof systems are proposed: metrics to review the rationality of required qualities are provided by arranging the required qualities according to hazard levels and work phases; metrics to evaluate error-proof systems are provided to improve their reliability effectively by mapping the error-proofing principles onto the error-proofing cases which are applied according to the required qualities and the corresponding hazard levels. In addition, these objectively-analysed error-proofing cases are available to be used as error-proofing-cases database or guidelines for safer HMI design especially for data management.