I'm questioning your ability to predict future improvements in performance on one task based on experience with a different task.