We investigated the effect of practice variability through execution redundancy in skilled and novice basketball players on free throw skills. Twelve skilled basketball players and 12 novices (mean age = 25.4 years, SD = 4.3) were divided into four groups (skilled constant, skilled variable, novice constant, and novice variable). After a pretest, participants practiced free throw action. The variable groups threw the ball over an obstacle of varying heights on each trial in random order, whereas the obstacle’s height was fixed for the constant groups. After 7 and 14 consecutive days of practice, participants performed two posttests with constant and variable distances from the basket. The results showed that practicing different solutions of a task did not affect the performance of skilled players but had an immediate negative effect on the performance of novice players. Learning a complex task is the result of learning task-related parameters, and practice variability can create a mismatch between task difficulty and new learner skill levels.