When the source of action is motion capture one could naively presume that only physically possible poses would result, however, if re-targeting is used to apply motion from captured actions to a model that is proportionally different than the source it is possible to encounter physically implausible poses.