Port State Control (PSC) ensures that foreign ships do not sail until they can proceed at sea without posing a danger to maritime safety and environment. Since 2011, a New Inspection Regime (NIR) has been implemented by the PSC of the Memorandum of Understandings (MoU) member countries to identify potential substandard ships and increase the effectiveness of inspections. The NIR process and whether the target factors in the NIR can effectively reduce risks at sea are worth exploring. We take the Tokyo MoU as an example and retrieve 125,259 cases from 2015 to 2017 using computer programing. Based on the big data, we adopt a binary logistic regression to analyze detention decisions, which we then supplement by applying a decision tree to conduct a multi-factor decision-making analysis. The ships' characteristics such as age, type, deadweight, number of deficiencies, and others, namely, port State, performances of flag State, and recognized organization, may also shed light on the risks associated with detentions. Based on the empirical results, we derive three conclusions. First, the following factors are essential: ship age, ship type, performance of flag States, and the number of deficiencies that NIR considers significant. Second, ships older than 6 years seem more likely to be substandard. If a ship has five or more deficiencies, the probability that it will be detained is high. Third, discrepancies exist when carrying out the PSC among port States in the Tokyo MoU.