I think the way we try and explain physics with analogies and simplifications is often more confusing than just showing the equations.
Pre-Einstein we assumed "Galilean relativity" almost by default: the idea that if you're moving at a constant velocity, the results of any experiment you can perform are the same as if you were standing still.
Maxwell's electrodynamics is a wonderful, beautiful thing and well worth learning, but the relevant part here is that it implies that light always propagates at a fixed speed. And we can perform experiments that verify this, and we can - more or less - perform those experiments while moving at speed (or we can find clever ways to test the theory using the motion of the Earth, as with Michaelson-Morely), and we observe that they still hold.
So then just think about objects moving at constant velocities, and how to reconcile their observations with each other - for simplicity, you probably want to start with 1D space. Plug the observations in, write some numbers, and special relativity more-or-less drops out as the only way things could possibly work. It really is a lot less confusing than it sounds.