And to extend your correct remarks and combine them with the article "Russian roulette" section, at that time NASA had no idea if the insulation failure was the

And to extend your correct remarks and combine them with the article "Russian roulette" section, at that time NASA had no idea if the insulation failure was the usual statistical thing, or if there was a procedure/material failure in that ramp and all ramps going forward, or there is some weird interaction such that the fumes from a slightly different tile glue ruins all current and future insulation. What I'm getting at is part of what took months after the accident was ruling out a new, previously unknown problem, that might have doomed every single launch going forward rather than merely the russian roulette scenario. The article has a pretty bad case of 20/20 hindsight WRT this topic. Immediately after the accident NASA knew the risk of a ramp failure was about 6/100 and a resulting vehicle loss was about 1/100 but those are LOWER bounds, the upper failure bound due to a manufacturing change or supplier error might very well be 100% on both counts and it took months to rule that out. If the problem had been a minor catalyst change in the foam chemistry such that every ramp fails under launch conditions, every time going forward, then launching Atlantis would result in two dead crew not one.