There has been a lively debate in the philosophy of science over predictivism: the thesis that successfully predicting a given body of data provides stronger evidence for a theory than merely accommodating the same body of data. I argue for a very strong version of the thesis using statistical results on the so-called “model selection” problem. This is the problem of finding the optimal model (family of hypotheses) given a body of data. The key idea that I will borrow from the statistical literature is that the level of support a hypothesis, H, receives from a body of data, D, is inversely related to the number of adjustable parameters of the model from which H was constructed. I will argue that when D is not essential to the design of H (i.e., when it is predicted), the model to which H belongs has fewer adjustable parameters than when D is essential to the design of H (when it is accommodated). This, I argue, provides us with an argument for a very strong version of predictivism.

在科学哲学中，关于预测主义的争论一直很激烈：成功预测给定数据集的论点为理论提供了比仅仅容纳相同数据集更有力的证据。我主张在所谓的“模型选择”问题上使用统计结果来证明论文的一个非常强大的版本。这是在给定大量数据的情况下找到最佳模型（假设族）的问题。我将从统计文献中借用的关键思想是，假设H从数据主体D获得的支持水平与构建H的模型的可调参数数量成反比。我会争辩说，当D对H的设计不是必需的（即预测时），H所属的模型比D对H的设计必需时（容纳时）具有更少的可调参数。我认为，这为我们提供了一个非常强大的预测主义版本的论据。