Groups of high- (n = 14), intermediate- (n = 12), and low-skilled (n = 15) cricket batsmen participated in two experiments to examine expertise-related differences in anticipatory information pick-up that combined temporal and spatial occlusion methodologies. In experiment 1 participants were shown video displays of a bowler delivering one of three different types of delivery with the display manipulated so that only selected local features of the bowler's movement pattern (bowling hand, bowling hand and arm, trunk, lower body, or whole body) were visible and then only for specific time periods prior to ball release. Only the highly-skilled players were able to produce better-than-chance predictions of ball type and then only under a limited set of display conditions. Information from bowling hand and arm cues was particularly critical although continuous visibility of these cues was apparently not essential for information pick-up. In experiment 2 the order in which particular features were made visible throughout the bowler's movement pattern was varied in an attempt to find the sequence of cues that was most favourable for effective information pick-up. The necessity in this experiment to switch vision between different features eliminated the highly-skilled players' capability to anticipate. Expert anticipation is dependent on sensitivity to information arising from a select set of local cues, and forced attentional switches between different cues negate effective information pick-up and, with it, the expert advantage.