No no, we always do the stats for one particular not well defined type of crime over two weeks, that's just basic statistics. You want as small and unrepresentative sample as possible, for science reasons.