I came closest to that when I wrote about testable hypotheses and the relation between theory and empiricism. Indeed, in contrast with (the common interpretation of) experiments, only all conceivable covariates can be held constant.

But that’s a different story, I would say. That would relate to a discussion on how to observe reliably, not to my statement that I thought that simulation models were purely theoretical without the possibility to deal with observations.

All what you describe here is not very popperian… no hard science here at all, but correlations in between mathematical constructs and observations… don’t you think that it is a bridge too far to talk here about experiments? I mean, I would… but I thought that you were more strict in this issue Cheers, I.