Normal, daily crisis: You analyze some data and dig out some old script that should do the job (maybe with some minor adjustments). But a problem while capturing the data let the script produce bogus output, you never believe. Instead of seeing the obvious thing (= the trace was messed up), you think about the script and discover some mayor shortcomming which renders all results from the script wrong.
Then you start to think about how to fix the problem and here is the crisis: You have no idea! But after looking over the script again you will realize: Hey, the is no problem with the script, it cannot produce wrong results because the person who wrote it (yourself!) was just smart enough not to make the mistakes you were thinking about.
Sometimes you are smarten than you would expect.