Maurice Brosnan examines the limits of statistics, the importance of defensive work and possible ways to improve the process