Put more neutrally, the attractiveness of any hypothesis is determined in part by physical circumstances that have no logical connection to the diagnostic problem at hand, but a strong pragmatic bearing on it (kind of like origami). The factory service manuals tell you to be systematic in eliminating variables, but they never take into account the risks of working on old machines. So you have to develop your own decision tree for the particular circumstances. The problem is that at each node of this new tree, your own un quantifiable risk aversion introduces ambiguity. There comes a point
...more
