✈ Do you like planes? Armoured planes, specifically? Because in that case I’ve got a story for you.
During WW2, the British 💣 bomber command was trying to improve the survivability of their bombers flying over Nazi Germany.
Wisely, they decided on a data-driven approach: they inspected all returning aircraft for battle damage and look for patterns.
The plan then was to armour all sections of the aircraft particularly subjected to damage, so they would hopefully be more resistant in the future.
Smart, right? The brass certainly thought so.
Except… somebody paid attention and pointed out: those planes made it back to base! So bullet holes in those particular places seemed to be not such a big deal.
Instead, the planes should be armoured precisely where they had little damage. Because, the attention-payer argued, planes with damage in those locations were probably at the bottom of the North Sea!
Goes to show how hard it is not only to come up with good metrics, but also how easy it is to interpret them in absolutely the wrong way.
(Photo by bugdog from FreeImages)