In basketball games, it is important to spontaneously adapt to the individual fake frequency of different opponents. This study investigated if participants (n = 34) are able to adapt to player-specific fake-frequency schedules on a trial-by-trial level. Therefore, stimulus material of three different basketball players was used, and each player was presented with a different frequency of head-fake trials (i.e., 20%, 50%, 80%). The analysis of the mean RTs revealed that the head-fake effect was present for 20% head fakes and disappeared for 50% head fakes. Also, the head-fake effect reversed (though not significantly) for 80% head fakes. Results indicate that participants adapted to the individual frequency of head-fake usage of the basketball player displayed in a trial. We further analyzed whether this result can be explained by stimulus-driven episodic retrieval of previous responses that were executed at the last occurrence of the stimulus. Indeed, the findings from a multilevel modelling technique point out that for the RT data, the interaction between type of pass and fake frequency (which is indicative for player-specific adaptation of the head-fake effect) was no longer significant as we statistically controlled for stimulus-driven episodic retrieval effects. This implies that episodic response retrieval accounts for a substantial amount of variance in the head-fake effect. By transferring these findings to head-fake effects in basketball, players can increase the efficiency of a head fake by repeating the same head movement but changing the direction of the pass.