remember the game developer's most useful tool: when in doubt, lie to the player
Destiny 2's "low health" visual feedback (red vignette, flashing health bar, sound changes etc) kicks in when the health bar is down to about 25%.
Except that's about 50% of your actual hitpoints; Bungie just realised the game feels more exciting if you get the feeling that you narrowly scraped out of a tough fight.
People got VERY mad about this when one of the designers revealed this on Twitter lmao
In Dragon Age Inquisition, whenever you're riding a horse and start sprinting, you don't actually go faster.
The camera just zooms in and there are speed lines so it feels faster
xcom2
humans are notoriously bad at understanding chance and probability, and firaxis decided to implement two things hidden from the player in order to nudge the probability of things closer to Average Human Perception
first is that the Hit% displayed is inaccurate in three ways. first, on Rookie and Veteran you get a 1.2x/1.1x hidden multiplier to your hit% that doesn't show on the tooltip. second, every time you miss a shot, you gain a stacking difficulty dependent hidden bonus modifier to your Hit% (+10 on rookie and veteran, +15 on commander) that is not updated on the tooltip if the shot has at least 50% chance to hit, until you miss, and up to 95%. enemies get a reverse of this: every time you're hit, they gain a difficulty dependent stacking Hit% malus, until they miss again. the only difficulty that gives no one any maluses or bonuses is Legend. third, on difficulties below Legend you get a stacking hidden flat +15/10/10 to aim for every soldier you lost if you have less than 4, and the aliens get a stacking flat -10 penalty in the same vein.
the other sleight of hand is that crit% works completely differently from how it works in literally any other tactics RPG, or, frankly RPGs in general. in just about every RPG ever, you roll to see if your hit connects, and then you roll to see if your hit ends up being upgraded to a critical. xcom2 says "lol, lmao", and instead decides to use the exact same roll to determine crits.
as in, if you have a 30% chance to hit, and a 30% chance to crit, a hit is automatically going to be a crit. the result of this is that low chance shots end up being a nailbiting experience.. both ways. you might feel like a gamble paid off when your "impossible to hit" shot gets "upgraded" to a crit, but in reality it was pretty much predetermined to be a crit. and you'll curse even harder when your dude behind heavy cover, hunkered down, in a smokescreen, covered by a support bot gets one hit killed by a stray bullet that only had a 2% chance to hit because elite troopers have a 10% crit chance.
a third, much more minor, thing is that the hit% predictor rounds up, so you can actually miss a 100% shot because it's actually 99.1%. that one's just funny, and not a sleight of hand.
these changes are universally hated, and the two mods reverting these changes are permanently etched into the first two pages of Steam Workshop's "most subscribed" list for XCOM2
(there are other things the difficulty levels change that's more behind the scenes trickery, but those are so subtle most people will never actually notice them, making them positive examples, not antiexamples)
Notably Three Houses, tho I assume others do something similar.
Anytime one of your units has a hit chance 50% or greater, you actually roll twice and take the better result. So a 49% attack hits 49% of the time, but a 50% attack hits 75% of the time; 90% actually hits 99% of the time. Enemies do not get this benefit.
Similarly, whenever an enemy has a hit chance 50% or less, they secretly reroll if they would hit you. So a 50% attack actually only hits 25% of the time.
All of this is to accord with humanity's busted-ass sense of probability, where a 70% is "basically a sure thing" (rather than "misses about 1 in 3 times"), and same for small enemy hit chances like 30%.
Combined with the way it generates random numbers (in a fixed sequence, so reloading and performing the same actions gives the same results), you can really engineer your way around fate by "using up" either one or two bad rolls at a time on a less important attack so you can get to a good roll on the attack you need to hit.