Please re-read the original article. I believe 'trend' and 'average' are statistical terms! You can't tell me stats prove speed bumps are a good thing when they clearly show that can't be proved in the case of Portsmouth! For example - maybe they work in London because the bumps are only every 400m apart - whist in Portsmouth they're 800m. So a sweeping generalisation like (Speed bumps save lives- is ,excuse my English, a load of bull!).
Statistics are, and always will be a guide to a course of action - but not the be all and end all. Without a set of control conditions they are next to useless (medical trials use strict control conditions btw - good example to prove my point). I could also counter with examples of obviously idiotic statistical logic!! By all means, use statistics for this purpose, then do some on the ground work to actually see if they're accurate.
In the case of the zone where I live - the generalisation isn't true - as is also the case in Portsmouth apparently. No point in shooting the messenger just because the message isn't the one you want to hear!!