Originally Posted by Tourist
We are now suffering such low accident rates that they are too low for statistical analysis. One accident skews the graph for a year so far it gives no useful inferences.
Noooooooo. You don't get statistical skews. You get peaks and troughs but over time there are still trends and facts. If the lack of numbers limit the ability to make a quantitive assessment you can then make qualitative assessments (i.e. based upon experience). The argument is not wholly about numbers of deaths, it is about rate of deaths or, to be more precise, the risk of occurrences which may lead to deaths. Put another way, if for argument's sake we'll accept one death per airshow or one Shoreham a year, how may airshows are we going to have at that rate before we take a look at what is happening?