There is no baseline to measure against candidates as opposed to competitive programming problems