This study aimed to clarify the cause of rugby head and spinal cord injuries through a network centrality analysis of 14-year (2004-2018) longitudinal data in Japan. The study hypothesis is that understanding the causal relationship among the occurrence of serious injuries, the quality of player experience and play situation as a network structure could be possible to obtain practical knowledge on injury prevention. In this study, bipartite graphs are used to make it easier to understand the situation of players and injuries. This would also help to elucidate more characteristic subgroup. A network bipartite graph and subgroup (cluster) analyses were performed to clarify the injured players' experience and the cause of injury. We used the algorithm of R program, IGRAPH, clustering edge betweenness. For subgroup extraction, the modularity Q value was used to determine which step to cut. The Japanese rugby population was 93,873 (2014-2018 average), and 27% were high school students. The data showed that careful attention would be particularly needed for groups of inexperienced Japanese high school players. Our study suggests that we should consider introducing rules that prohibit "head-on collisions" in youth rugby.