Low-poly topology and subdivision topology are two different techniques.
Lowpoly technigue bewares about polycount
Sub-d doesn’t care about polycount, as it will be subdivided many times. We care only about the cage.
Lowpoly uses triangles a lot, especially if the mesh won’t deform.
Sub-d also uses triangles, but sparsely, and not in major areas, where consistency of flow is of utter importance.
In both techniques you care about visual result, which should look smooth, edges go parallel\perpendicular to forms or deformation (not diagonally or chaotically).
Lowpoly doesn’t have a concept of loops\supporting loops. As the model is what you see is what you get in the engine.
With lowpoly, you also care about smoothing groups, as they mimic abrupt or smooth changes.
With sub-d, you use many loops to make the mesh denser to control its edges.
I would look for lowpoly topology examples. There’s nothing complex about it. It just has to conform to the poly budget of a particular project, and have optimization (deleting unnecessary edges, which don’t shape any form).