I pay attention to how the inside of my head works. Extreme introspection. I'm kind of obsessed with knowledge, learning, and optimization, so it's become a sort of nervous tick.
The brain is an intuitive tool: you can pick it up and use it to reasonable effect without learning how. As with most intuitive tools, you can use it to *great* effect if you have better understanding of technique. This is why some people have shitty handwriting, and others are scribbling out professional-grade calligraphy just by using a slant pen: continuous cycles of practicing, of examining the results, and of recognizing and deliberately correcting your mistakes leads to picking up a slant pen and writing a thousand-dollar wedding invitation.
I'm the guy who learned to make wedding invitations while everyone else was learning to write barely-legible cursive.
The systems simulator is just my own constant tool: whenever I approach a problem, I simulate it. If I'm playing a video game, I'm looking at the sprites on the screen in terms of their direction and speed of travel, and accounting for any known behaviors: I see where things *will* be *simultaneously*, instead of just their current position, direction, speed, and maybe path. I instantly know if things will intersect. The same goes for real-world physics, although that domain is more complex *and* has variables I can't always measure, as well as many I don't often interact with and can't readily project. I do the same in economics, loosely correlating changes with the pressures they put on other changes and shifting the whole system at once. I manipulate data structures in my head when coding, essentially emulating an abstract representation of a computer processor.
Some of those are more or less abstract, and more or less accurate. I'm *very* good with video games--no surprise there. Real-world physics has the stated problems: don't know all variables, haven't observed every aspect in great detail; I can catch a ball for the same reason you can, but I can't fire a sniper rifle because I need a *lot* of time to (poorly) simulate wind resistance against a bullet. Economics is actually a pretty simple system, as long as you're dealing with billions of people and not dozens. Computer programs are like economics: analyzing the *whole* program is hard, and you have to do it in pieces with propagating effects--this is generally a good strategy.
Einstein was a scientist, yet he acted like a philosopher: he sat, thought, made a bunch of shit up, and somehow turned out right. He wasn't pulling experiments in a research lab; people are so amazed by Einstein partly because nearly everything he declared as truth--stuff he was *right* about--was impossible to test in his time. We're only now getting the equipment and the opportunity to vaguely identify that Einstein *might* have been correct, and that evidence exists which is fully explained by his theories, but not necessarily which would lead us to synthesize those same theories without a nudge in the right direction. Where do you think he got it from?