You will always have intuitions or hints about what to do next, particularly
if you use a compiler that suggests changes to make (like rustc does very
well). Take a step back, ask your système 2 and accept the cognitive
effort of knowing why you do what you do.
Guess what is the most obvious cause of the problem you can think of.
try to prove that it is not that
failing to prove this, try to fix it then.
Very often, when debugging stuff, you’ll be anchored by the story one told
you. Like « It worked yesterday and today it does not » and will generally go
in the wrong path. The one that tells you this story might very well be
yourself. Never trust even your stories and always double check them.
Face the fact that you will be wrong. There is no shame in that.