Retrospectives with system modelling

I would argue that sometimes, at retrospectives maybe the best thing we can do is to try to better understand “the system”, and for that system modelling seems like a very good approach to me.