My eyes glazed over and my head started to spin when I went onto this page... http://en.wikipedia.org/wiki/Odds_ratio I think I will stick with my KISS method, the numbers are good enough for me... Divide how many ACTUAL deaths helmet-less, compared to ACTUAL deaths helmeted, close enough to give the general outcome... JMO If you want to also add the % of each type of rider and total number of riders into the equation to get a "better" number that also works for me but that's as complicated as it needs to get to show possible outcome WHEN there is an ACTUAL head injury in an accident. (not talking odds of getting INTO an accident, or the odds of getting a head injury in an accident)...