Idealizations or approximations might do that as well, but always in relation to a target system they represent. Toy models, in contrast, lead to insights through arguments from analogy.

Previous attempts to test noncontextuality have all assumed idealizations that real experiments cannot achieve, namely noiseless measurements and exact operational equivalences.

Idealizations are like the Bitmoji of science. They don't show everything, but they show enough that we can figure out what's going on. More Great WIRED Stories.

Miren Boehm, Hume's definitions of 'Cause': Without idealizations, within the bounds of science, Synthese, Vol. 191, No. 16 (November 2014), pp. 3803-3819 Free online reading for over 10 million ...