You can spend all day shooting at a piece of paper, trying to get the smallest group and still not really learn much, apart from how statistics are a hell of a drug. Instead, can I suggest you start introducing some stressors that shake out where things can be improved and what you need to do to improve them.