Post

Just gotta reply back with "That's XCOM baby!" and call it a day

Darkest Dungeon just literally shows numbers 10% lower than the reality, so "90%" is actually 100%. (I don't actually love this myself, but it was certainly An Approach to the issue.)

Fire Emblem is the one that fudges the numbers more with high% more likely to hit and low% more likely to miss. XCOM's RNG is infamous in large part due to the percentages being exactly what the user sees (though it does cheat in favor of the player under specific conditions on lower difficulties)

"Is it a bug or a forces beyond my control conspiring against me?"

That's the fun thing. We are ALL that player. Because we only ever remember the cases where it acts oddly. But some amount of "feels incorrect" strings of events are pretty much guaranteed with enough rolls of the dice.

When I worked on Diablo 3, I was testing an event that had a 50% chance to appear. It did not after 12 attempts, so I went to an engineer about it. They looked at me silently for a few seconds and said. "That's a horrible sample size, come back after a thousand attempts."

How dare? No one said it could happen to *me*.

People still complain about those though. Look at the complaints magic gets about its shuffler even though it's actually rigged for the player in best of 1.

Didn't someone from Firaxis say in a talk once that they massaged the RNG they used in one Civ game so it was no longer properly calibrated, purely to cut down on how often where players would lose a 90% chance?

But what if there was a 1 in 4 chance of something good happening? People take that bet all the time!

No one’s more susceptible to this thinking than me playtesting my own game. “Wait, that was supposed to be about a 1-in-30 chance but it happened twice in a row?! Did I program this right?”

A few cheeky thumbs on the scale but for the most part no

Never show percentages and chances! If you have to communicate them, use dice or a mock slot machine, or something where the player has the agency of "starting" the draw.

"oh so that explains the emails I just got from 4 other guys telling me they got lucky"

It happens sometimes! Common case for me is if I read a probability as an integer instead of a float and it gets truncated to 0, or messing up the luck polarity on certain rolls

Only on lower difficulties, and even then there's a lot of conditions attached. The biggest one being that the bonus only applies if you have four soldiers or less in play, so if you've got squad size upgrades with 5-6 soldiers in play you don't get anything at all unless you start losing people.

Oh, so XCOM was the one that doesn't cheat behind the scenes? Guess I misremembered 😅

@gamemakerstoolkit.com It's like when the #Apple iPod appeared and people were angry about the song randomiser in it. Because it was actually random: it paid no attention to whether you got three songs in a row by the same artist.

Fire Emblem devs got so frustrated about it that they started lying to players. Everything that says it has more than a 50% chance of succeeding actually has a higher chance than it shows and everything that says it has less than 50% has a lower chance than it says.

I had a friend who was angry about the fact that it was raining post a message saying "20% chance of rain my ass". In fairness to them, I'm pretty sure they realized the logical flaw there sure afterwards, but probability hard! It's just not how our brains work