[upofadown]

Many safety standards have surprisingly little science behind them. That is not always a bad thing. Trial and error can work if there is the possibility of changing the design. Once the body of experience is codified into a standard then development stops.

The current bike helmet standards are a good example of what happens in the absence of both good theory and experience. We have a design that prevents an injury that almost never happens (skull fracture) and that fails to protect against a common injury (concussion). The current standards make it impossible to improve things much but there are people out there that are trying:

http://www.helmets.org/liners.htm#cone-head