Anna Filina¹, Maryam Vaziri-Pashkam¹; ¹University of Delaware

In daily interactions, we often have to read others’ body movements to predict their goals. In competitive contexts, these bodily cues can be misleading, as opponents may feint to deceive each other. To understand how humans predict actions when deceptive movements are possible, we designed a competitive reaching task. Two participants sat across from each other at a table, and each had a starting point and two targets. One participant (the Attacker) was instructed to hit either the left or right target with their finger, after which their opponent (the Blocker) was instructed to hit the same target as quickly as possible. Motion sensors were attached to participants’ index fingers to measure reaction time, accuracy, and movement trajectories. In Deception blocks, Attackers were allowed to deceive the Blockers. In Control blocks, Attackers were instructed to go directly to the target on all trials. Reaction times were faster in Control compared to Deception blocks, indicating that Attackers successfully misdirected the Blockers. Blockers’ accuracies, however, did not differ between conditions, suggesting that Blockers maintained high accuracy despite the misdirection. Next, to investigate how participants read each other’s bodily cues in this interactive game, we also measured reaction times when Blockers reacted to a video of a dot representing the Attacker’s finger position on a computer screen. Reaction times were faster when participants had access to full-body movements compared to when they responded to the moving dot on the screen, with no difference in the size of this advantage in control compared to deception blocks. These results demonstrate that access to full-body movements is critical for predicting others’ goals, regardless of the presence or absence of deception. They provide insights into how humans interpret and predict actions in settings where deceptive movements can occur.