However, according to at least one Star Trek NTG episode, making unpredictable changes in your choices can cause annoying tell-tale computer lag while the simulation “recalculates” to take into account the change, such as First Officer Riker choosing to take the turbo-lift to a different deck than was expected. But on a sufficiently complex computer, wouldn’t it calculate “all decks” at the same time, just in case? Perhaps the engineers of the Enterprise saved a few dollars by using isolinear chips from centuries past, and so that explains it? Or their chips run at MHz and not GHz, because the Star Trek stories were actually written in the 1990s?
Holodecks become somewhat complicated, because they have a way of being “too real”. Thus they have to have “safeties”. Thus, rather than dying from a fatal mistake in the “simulated universe”, the simulation abruptly terminates, and you are left staring at the walls of the holodeck. As if to say, “Game Over”, “You are dead,” “Maybe you can do better next time.” When the safeties somehow become frozen (TV-logic), then you have to extra careful, because holodeck bullets can actually be deadly. Sometimes the program won’t exit properly, so that the door to exit the holodeck is nowhere to be found. (more TV-logic) I doubt that a real holodeck would have so many plot-driven failures.