That’s equivalent to saying that if you don’t have a motorcycle license, don’t register your bike, don’t have insurance, and don’t wear your helmet, your fatal accident risk increases by over 3x. Put that way, it’s not surprising, nor does it actually tell you anything about the base rate safety of lawful motorcycling. By way of analogy, you could just as easily say “not dousing yourself in gasoline reduces the risk of death by smoking by 98%”, which is both true and useless.
> That’s equivalent to saying that if you don’t have a motorcycle license, don’t register your bike, don’t have insurance, and don’t wear your helmet, your fatal accident risk increases by over 3x.
That's not really how statistics work. Since the reduction was probably calculated against the population average you need to know the relative size of the groups to calculate the risk increase for the inverse group. Additionaly, the group you specified is not the inverse group since you exclude those who have some, but not all, of the safety signals.
Your calculation would be accurate if almost nobody took all safety precautions (that would mean the average risk rate would be affected much by that group) and everbody else took no safety precautions.
What you have calculated is a rough lower bound for the risk increase given unknown population behavior ratios.
> nor does it actually tell you anything about the base rate safety
It doesn't by itself. What it tells you is given a base of rate of 3x more deadly per mile, those who follow all the rules are as likely to die as an average driver (which still isn't an fair comparison.) To be fair, you'd beed to compare agaisnt driver who have a license, registration, insurance and are wear a seatbelt. (Or maybe helmet..)
My interpretation of the original claim was that, on a per capita basis, the rate of fatal accidents among motorcyclists with license, registration, insurance, and helmets is 30% that of motorcyclists who have none of those things, with no particular claim implied about motorcyclists who fulfill between 1-3 of those four criteria. I know that wasn't necessarily the only possible interpretation but I think it's a reasonable one.
> on a per capita basis, the rate of fatal accidents among motorcyclists with license, registration, insurance, and helmets is 30% that of motorcyclists who have none of those things
When presented with unsourced statistics on the web, probably best to assume the weakest interpretation.
Maybe, but I was also assuming good faith, which entails assuming that Workaccount2 is neither dishonest enough to intentionally share misleading statistics nor completely ignorant about the more basic and obvious ways that statistics can be misleading. If you always assume the weakest interpretation you spend a lot of time quibbling over basic points.
I totally get what you are saying, but if you ride motorcycles and have been around motorcycle groups, the stat is clearly saying "as expected, it's the dumb kids doing the dieing".
The comment is written for other riders, I left out a lot of detail for it to be a general comment.
