- they can try but all the known rules of thumb are not quick and all the equations are not known. Principles of Spell Design lays out all of the heuristics you can use but you still usually fail, when designing spells with all of those in mind; it's speculated that gods can see all the hidden properties exactly as clearly as the visible properties and that's why it's not hard for gods to design spells.
They launch into all the known rules of thumb, usually with the caveat that casting cantrips doesn't actually require this. You can think of one aspect of magic as lagging the visible aspects like so, requiring more energy to move and moving more slowly when it does but also requiring more energy to stop; you can think of another as reacting badly to close contact with itself, and resisting spell structures like such or such, which is why no spells have structures like that; you can think of this other thing as possible to tug into alignment only by sort of jiggling the spell, and you can tell you've got it when you don't get any reverberations when you do this -
And people have, of course, tried hyperdimensional representations that capture all that, but it's hard, and usually less useful to learn than the heuristics if you aren't specifically doing spell design, and none of them have arrived at equations that if solved for let you invent spells, despite having headbands and plenty of motivation. It's understood that the number of ways magic interacts with itself is just very very large, for high circle spells, and it's not reducible complexity.
It is, someone ventures, sort of like the thing Keltham said, about how knowing how objects move doesn't let you catch them in the air.