贝叶斯决策理论对复杂运动决策中运动预期的启发——以网球和足球为例

Inspiration of Bayesian decision theory for action anticipation in complex decision making in sports:Taking tennis and soccer as examples

对运动决策的研究是认知运动心理学的一个重要领域。运动预期被认为是运动决策的核心,受到运动学和非运动学信息的影响。其中,对运动预期研究的一个关键问题是探讨不同信息源对运动结果预期的贡献以及两者之间的相互作用。研究者运用贝叶斯决策理论解释运动预期中不同信息的整合过程,分析运动员在复杂的竞赛情景中如何做出最佳决策,尤其是对该理论在网球和足球领域的潜在应用进行分析。在不确定的情况下,竞技体育中并非所有的选择、结果或概率都是已知的,故有研究者认为概率论和经典的决策理论不能有效解决此类问题。然而新近提出的启发式近似,为运动员在贝叶斯框架下如何快速做出选择提供了理论依据:首先,在复杂和有时间压力的竞赛情景中,启发式近似假设运动员依据竞赛中不同信息源的不确定程度,很可能选择在运动学信息和情境先验之间进行切换启发式,提高运动预期的效率。其次,判断效用通过卷积效应影响两种信息源的整合,降低情境先验的影响程度。

The study of decision making in sports has always been a significant field of cognitive psychology in sports.Action anticipation is the core of decision making in sports,which is thought to be influenced by kinematic and non-kinematic information factors.One of the key issues in the study of action anticipation is to explore the contribution of diverse information sources to the expectation of action outcome and the interaction between them.The researchers explain the integration process of different information in action anticipation in sports and analyze how athletes make the optimal choice in complex competition situations in sports by using Bayesian decision theory,especially its potential applications in tennis and soccer.Although not all choices,outcomes,or probabilities in competitive sports are known in the context of uncertainty,some researchers believe that probability theory and canonical decision theory cannot effectively solve such problems,the newly proposed heuristic approximation provides a theoretical basis for athletes to make a rapid choice in the Bayesian framework.First of all,in complex and time pressure competitive situations,heuristic approximation assumes that athletes are likely to choose a switching heuristic between kinematic information and contextual priors according to the uncertainty of different information sources in the competition,so as to improve the efficiency of action anticipation.Secondly,judgement utility affects the integration of two information sources through the effect of convolution,so as to reduce the impact degree of contextual priors.