Two phenomena previously observed In multi-finger static maximum voluntary contraction (MVC) tasks -(1) force deficit and (2) enslaving - were compared with the force patterns produced during sub-maximal dynamic tasks. A new tool, the inverse piano, was designed to measure the finger forces during the sub-maximal dynamic tasks. During the dynamic experiments, the keys of the IP elevated according to a computer run program. Subjects (n = 9) were instructed to press down the elevated keys as fast as they can. All finger combinations were tested (totally 15). Force deficit was not observed for the dynamic tasks. Two aspects of the enslaving effects (EE) for the dynamic task were found to be similar with the MVC task: the EEs were relatively large (as much as 47.4% of the maximum force produced) and nearly symmetrical. Proximity effects and occlusion were not prevalent in the experiment where the key combination was known prior to key activation. In the case where the key combination was unknown and randomly chosen, proximity effects were retained, but occlusion was not observed. Inter-finger connection matrices (IFM) calculated for both the MVC and dynamic tasks further stressed the dissimilarity between the force patterns used to complete the tasks.