Hacker News new | ask | show | jobs
by dogsgobork 1369 days ago
I'm reminded of the anecdote about head injuries rates going up in war after helmets were issued. More head injuries due to fewer deaths.
3 comments

Isn't this a case of survivorship bias, like when they took a statistical view of all the areas where WWII airplanes got hit by bullets and they wanted to increase the armour, but a statistician told them to armour the parts where no bullets had hit... because they only surveyed planes that came back whole and the ones that didn't come back had likely been hit in parts of the plane they didn't add to the survey.
Yep, the canonical Wikipedia example that aaronbrethorst mentioned above.

(https://en.wikipedia.org/wiki/Survivorship_bias)

Canonical example is the header image on this Wikipedia article https://en.wikipedia.org/wiki/Survivorship_bias
No, survivorship bias would be looking at motorcycle injuries by only talking to motorcycle injury survivors. Your conclusion would be that motorcycle injuries only result in minor small damage like skin abrasion, broken fingers, etc. Your conclusion would be that brain injuries basically never happen because all the survivors you spoke to were fine.

If you're looking at casualties from motorcycle incidents, you by definition are not looking at survivors.

Same thing every belligerent discovered during WW1 once the introduced steel helmets instead of leather.