But at the same time, simplifying strategies runs counter to the researchers' goal of solving underfitting and trying to find more suitable strategies. Figure 3-5 gives a rough example of underfitting and overfitting when a policy changes between simple and complex. Since the internal operating logic of trading assets is not yet clear, any trading strategy can only mine and utilize part of the intrinsic characteristics of the data, which is the light gray area where the two circles overlap, while the remaining white area is where the trading strategy does not have Actual utility but objectively existing part. When the complexity of the strategy increases, the strategy is likely to make more use of the characteristics of the data, which is manifested as an increase in the light gray overlapping area, and the under-fitting problem is alleviated. But at the same time, the white area where the strategy is ineffective may also increase accordingly. This part is the cause of overfitting after optimization. #whycrypto