This study aims to improve of secondhand market model used in the Maritime Business Game (MBG) & to judge the relevance of three data mining algorithms. Pruyn developed MBG to simulate the long term consequences of various shipping scenarios. MBG consists of three multilevel models to represent the maritime economy. Starting from the Multi Country model, it goes down to the shipping markets model & ends at the fleet scheduling model. Secondhand market is a sub-part of shipping market module, where the shipowners trade their old ships. In real life, the valuation is done by shipbrokers under the assumption that vessels are in good working condition. Thus’s, physical characteristics have a little influence on the sales price. A structural model based on the physical characteristics can be a good complement to the broker’s judgment.