I have a theory of statistics: if you can double them or halve them and they still work$$$ they are really good statistics.