LYING TO PLAYERS

I prefer transparency because too much lying and you get stuff like being angry about missing a 90% chance to hit because you've been lied to so much that you expect it to be a 100% chance to hit.

Of course I also understand the desire to lie because anything above 66% always feels like it ought to hit.

It also depends on the game. I think the reason it sometimes felt frustrating in XCOM was because you didn't have enough firepower to not. I've seen the post-XCOM XCOM-a-likes (note: not X-COM) often have very transparent determinism instead and that any randomness comes from a hit doing 2-4 damage instead of doing 0 OR 3 damage.

Whereas in the older games (like X-COM maybe. I never actually played original X-COM only Jagged Alliance and later Xenonauts) your squad size was often such that it was tactically sound to bring a backup to every fight.

When it specifically comes to hit chances I've also found it's a lot easier to accept failure when it is shown in some way. Like I play Blood Bowl 2 on PC and the game has two options of showing the chance of success for any given action. It can give you percentages or it can give you the actual die roll. For some reason the die roll is a lot easier to accept because a 3+ roll is somehow easier to parse than that 66% chance of success.
And if the game then also visually rolls these digital dice it somehow becomes easier to accept that I rolled a one than if an action just immediately failed. Even if there is literally no difference in the backend between these two things.

(The die rolling is also why 66% is the magic cutoff number. Because to me personally a 66% chance to hit seems like a sure bet. But a 3+ roll on a die seems a lot less so even though the two are exactly the same)

Maybe it was a meme, but, I recall seeing, somewhere, a screencap of an XCOM game where a soldier's pulse rifle was aiming at an alilen's face. Like, the barrel of the rifle was literally right on the face. The accuracy of that shot? 95%.

While you could parse that into rolling a nat 1 on a D20, what makes me think it was a meme is how ridiculous that kind of setup would be to begin with. Though, I'm not sure how often that situation could actually come up in a game like that, so, maybe it's a moot point.


More to the thread, I kinda want to talk about Heroes of Might and Magic. The game uses one graphic that can represent anywhere between 1 and 32767 of a single unit. Or, to put it another way, the data is stored as two bytes unsigned. As such, accuracy checks are thrown out the window. However, because of how the damage is calculated, it might be possible for a unit that numbers in the hundreds to only deal 1 damage, if the difference between that unit's attack power, and the defending unit's defense power, is large enough.

fudging is good and people should do it more, but maybe not show the % LOL

I looked into it and it's funny that it turns out that XCOM does also in fact fudge the numbers. Unless you're playing on Legendary it secretly skews the numbers in your favor. I looked into Darkest Dungeon too, it has a lot of hidden stuff like adding an accuracy bonus if you miss two times in a row. So I mean if you're working on a game where hit chance drives a lot of descision making then fudging is probably inevitable for some kind of base line player enjoyment.

More foolish than balking at the numbers is assuming that humans don't fall prey to confirmation bias or gambler's fallacy. Idk if "the math checks out you're just stupid" is convincing especially in a game design discussion.

Depends on the player. I usually find RNG-heavy mechanics to be boring design-wise and not rewarding at all. Conventional hit and dodge rates are like gambling: odds always have precedence over skill and knowledge. Winning doesn't make me feel particularly better because I had no merit in it.

Strategic and optimal choices based on knowledge of what will always succeed, now that feels rewarding to me. Rather than gambling, this is akin to skill games and quizzes. The player has to study, observe, pay attention and learn something, improving themselves, and then use that knowledge to make the best decisions in a given situation. Going through that cognitive process is more fun and dynamic to me than picking odds and hoping for the best.
And you can still include some form of risk-taking in resource management for instance or by having randomness govern other things than the hit/dodge rate, such as the enemies' next actions, the composition of that next enemy party you decided to challenge, etc.

What I believe a dev should ask themselves is "How can I implement RNG-based mechanics in a fun, creative or interesting way?"

Fudging the odds, unless for narrative purposes, seems like a complex solution to a simple problem: Why not simply show the true numbers and adjust the rates?
Instead of lying to me after I miss twice in a row, give me a hit rate buff. That way I can at least consider that bonus when I weigh my options on the next turn.

If this is going to be about if RNG in itself is good or bad then I probably have lots of opinions on that also.

But I'm also the kind of person who would waffle and "both sides" it. Nearly all games have some randomness. Even chess has the one random factor of who plays white and who plays black (if you really want to stretch it)

Personally I think risk management is one of the most interesting aspects of tactical games. However this can play out in different ways. Fairly recently I played the game Warhammer 40k: Mechanicum. That game has very little hit chance randomness. There are some special abilities where the description is it has a 50% hit chance but I don't know how those work because risk-averse as I am I never picked any of those :)

But that game's RNG comes in the form of damage amounts. A gun will do 3-5 damage and I see that an enemy has 4 health left. Do I play it safe and move a second character in shooting range so that even if my first guy only does 3 damage I can still finish them off or do I risk it and hope that I will do 4-5 damage and spend the other guy on something that is less of a direct threat but something I will still need to deal with at some point?

RNG based games also often rely on having enough numbers to mitigate the acute randomness. If a thing is decided on one die roll, that's probably not great. But if a thing is decided on carefully deploying five die rolls where one or two can fail without the plan falling apart?

This is why Blood Bowl is a favourite game of mine because most of that game relies on a six sided die. That means that one in six of best case scenario actions will likely fail. There are mitigating factors (rerolls, meaning that as long as you have a reroll left the best case scenario failure is one in 36) and that game is all about choosing to do the least risky things first and the riskier things later (because any failure will also result in your turn ending prematurely).


Of course (to put it back on topic) in all of this if the player is being lied to they can't make an informed decision about what to do. And this is different from not being shown certain information. (A big part of these kinds of games is some kind of scouting, finding out an enemy's HP makes a huge difference in decision making. If you know something dies in one hit or two hits greatly impacts how much resources you have to spend to get a near guaranteed kill or a mostly guaranteed kill)

So I might have waffled about lying earlier, but no it's bad.

author=Marrend
Maybe it was a meme, but, I recall seeing, somewhere, a screencap of an XCOM game where a soldier's pulse rifle was aiming at an alilen's face. Like, the barrel of the rifle was literally right on the face. The accuracy of that shot? 95%.

While you could parse that into rolling a nat 1 on a D20, what makes me think it was a meme is how ridiculous that kind of setup would be to begin with. Though, I'm not sure how often that situation could actually come up in a game like that, so, maybe it's a moot point.

The thing about this is the abstraction that happens when a game is turn-based. The game shows people taking turns kicking each other and it seems impossible to miss a point-blank shot. But it's just a simplification of what is actually happening. The idea is that everyone is always moving, some games even have a different thing when people are "engaged in melee" (see for example DnD and attacks of opportunity) the idea is not that people are standing on their respective squares occasionally striking at each other instead they are fully engaged, dodging and moving around constantly making what seems in a static picture to be a sure thing possibly be a very different thing indeed (I think a variant of the XCOM meme is someone staring down the barrel of a shotgun and it says like 46% hit chance)

author=LockeZ

author=darken
More foolish than balking at the numbers is assuming that humans don't fall prey to confirmation bias or gambler's fallacy.

Of course the absolute dumbest of people do fall prey to that, but that's no reason for a game designer to change their game.

It's not the dumbest of people it's everyone. You probably commit to confirmation bias every day. If you don't think you're biased then idk I have bad news. You keep talking about intelligence, but smart people make bad choices all the time. Games are made for humans, so it's probably good to take player psychology and bias into account. Like not putting a gacha/lizard brain reward system in your game is a descision motivated by "I don't want people to act irrationally around this"

author=Red_Nova
I think RNG-based accuracy is itself a great thing.

I have a huge issue with this in modern turn-based games. It was different back in the 80's and 90's when you were limited in how you could design a battle system due to things like memory and the small amount of RAM old consoles had (even the PS2 in the 00's had some games like Ar Tonelico II that would cause it to seize up due to exceeding system memory). A modern example is Octopath Traveler - weapon accuracy is influenced by your accuracy stat and the enemy's evasion, plus status effects like Blind and evasion/accuracy buffs. Magic, however, is almost always 100% accuracy even when a character is blinded. Once your characters have enough MP to cast more than three spells in a battle and most spells hit all enemies, you quickly realize that magic is vastly superior to weapons and weapon-based skills. It turns into every character either running a magic class as their secondary, or running two magic classes like Scholar and Cleric to wipe out hordes of enemies in just a couple turns.

I don't mind accuracy being affected by status conditions and buffs/debuffs, as long as the base rate is 100%, since players can understand that and work around it when it happens, compared to a random chance to miss your 99% accurate attack when you need it most. Developers can find other ways to make their battles more interesting and dynamic than "haha you randomly missed the killing blow and now you're dead so you have to start the whole fight over". It's why Pokemon players have recommended Flamethrower/Thunderbolt/Ice Beam over Fire Blast/Thunder/Blizzard for years; the 100% accuracy moves that do a little less damage than the 70%/85% accuracy moves are more consistent and reliable. The more powerful moves are too risky or rely on gimmicks that were added in later games to reduce risk. It's not even about risk vs. reward at that point, since the risk so far outweighs the potential reward that there isn't even really any consideration outside of weather-based teams which don't really see any use in-game barring challenge runs or the post-game competitive scene.