Probabilities and possibilities
Turbine D, thanks for the interesting links.
I would add that, adding to the difficulty is also that known possibilities can be ignored.
Consider Katrina. The risk of such a hurricane wasn't exactly unknown, but for various other reasons steps weren't taken to cope with that. The recent flooding in Australia is another example. Despite what was repeated constantly, the flooding wasn't unprecedented. Flooding had happened often in those areas, and the governments there actually have decades of records of the flooding. For various reasons, they just hadn't done anything to deal with flooding.
This obviously relates to replacement of the pitots. It probably should have been harder to delay their replacement, by making it harder for the various parties involved to decide that isn't such an important issue.