I have come across situations where researchers have tried to analyse complex dynamic systems and they will often make drastic simplifying assumptions in order that the problems can be solved with a set of equations. Indeed if the problem included a few more of the real world complexities then the problem would not be soluble by these kinds of methods and instead you may have to resort to running computer simulations of the problem and approaching it that way instead.
As you can probably see already, I am struggling to find the right language to describe this scenario. I suspect there is a technical expression to describe problems that can be solved in the simple way vs the complex way... perhaps the simple scenario might be described as being "soluble with a closed form equation"... is that the phrase I am looking for? Please advise.
EDIT: Another distinguishing factor is that the "closed form (or whatever it's called)" solution is much preferred by academic journals over the other type because it can much more reliably be reproduced than a computer program.
EDIT: I first came across this distinction in the field of neural networks and then later in economics. In the case of economics the non-(closed form?) option is an "agent based model" which economists and economic journals appear loathe to employ.