Can Stephen Curry really know? - Conscious access to outcome prediction of motor actions

Abstract

The NBA player Stephen Curry has a habit of turning away from the basket right after taking three-point shots, presumably because he can predict the success of his shot. For such a consciously accessible prediction to be possible, Stephen Curry needs access to internal processes of outcome prediction and valuation. Computational simulations and empirical data suggest that the quality of internal prediction processes is related to motor expertise. Whether the results of internal predictions can reliably be consciously accessed is less clear. In the current study, 30 participants each practiced a virtual goal-oriented throwing task for 1000 trials. Every second trial, they were required to verbally predict the success of the current throw. Results showed that on average, conscious prediction accuracy was above an individually computed chance level, taking into account individual success rates and response strategies. Furthermore, prediction accuracy was related to throwing performance. Participants with better performance predicted the success of their throws more accurately than participants with poorer performance. Moreover, for the poorer performing individuals, movement execution was negatively affected by the verbalized predictions required, and they did not show variation in speech characteristics (response latency) between correct and incorrect predictions. This indicates reduced quality of conscious access to internal processes of outcome prediction.