In high-stakes calculations, there's a thing called a "sanity check" which is a quick & simple analog of a complex calculation that is done perhaps repeatedly u

In high-stakes calculations, there's a thing called a "sanity check" which is a quick & simple analog of a complex calculation that is done perhaps repeatedly using different approaches, often manually with traditionally simple tools such as slide rules and napkins. In addition to the full accepted calculation procedure, as a final double-check if nothing else. Just to make sure things are not coming out in some outlying ball-park which would be bound to indicate an error in the main approach somewhere, since more complex equations having greater number of variables or conversions simply have that many more chances for a mistake to be made. And a completely independent recalculation which is simple, quick, and more understandable by all, which comes up in decent agreement, is a very good sign that the main calculation has been carried out correctly. But most importantly if there is excessive disagreement, that really calls for stopping and checking every single figure backwards & forwards until discrepancies can be pinpointed and corrected. Of course somebody involved needs to not only show an interest in this kind of reliability, but also be able to do something about it in case a sanity check like this turns up a shortage to begin with :\