In this study, we verified a new method for evaluating rainfall events that trigger sediment disasters, based on disasters occurring in the Heavy Rain Event of July 2018. First, we calculated antecedent precipitation indices (APIs) using various half‐life times and draw two‐dimensional diagrams (snake line diagrams) of all combinations of APIs. Second, we set the line (critical line, CL) defining hazardous area on each snake‐line diagram based on maximum values recorded before the event. Third, we compared rainfall data from the July 2018 Heavy Rain Event with CLs to calculate the amount of precipitation exceeding past maximum levels. The method was applied to 1614 slope failure and 639 debris flow cases based on hourly precipitation data derived by analyzing radar and rain gauge observations. The results showed that 97% of the slope failures and all but one debris flow occurred where the plots on the snake line diagrams exceeded CLs during the event. Furthermore, the precipitation excess amount tended to be larger in debris flow cases than in slope failure cases, and the optimal pairs of APIs for predicting disasters were different among disaster types and locations. Thus, this study suggested that the proposed method can detect anomalies in rainfall credibly, and the current method which uses a single snake line diagram is not necessarily sufficient to capture every disaster. The proposed method does not require complex statistical analysis such as calculating return periods. In addition, it can provide flexible CLs that respond to climate change and land use history of each location. As a result, it is expected to be more efficient than the current warning method.