If the same set of methods is put into different market conditions, the results may be completely opposite: follow the trend and repeatedly stop losses during shocks, sell high and buy low when encountering a unilateral market, and it is easy to go against the trend. This article explains the state identification methods of trends and shocks from observable perspectives such as high and low point structure, moving average arrangement and fluctuation speed, discusses method matching, exit processing when state switching, and the use of position margin and failure conditions to set boundaries for judgment errors.