Our starting point is the converse observation: the
actual running time of a sorting algorithm could well
be adaptive even ifno worst case adaptive analysis
(showing asymptotical improved time complexity for
input instances with low presortedness) can be given