Regarding the nature of 'RNG'

[Gamer Supreme]

I might be missing something but I don't think that the RNG is as straightforward as a random number between 1-100 being selected.
My experience is incredibly limited though so I might be entirely wrong and incredibly unlucky.

As far as damage rolls go it appears to me that its really either lowest or highest, and on occasion a number in between, like constant 8 damage with a crusade when its 8-16. The highest roll I've ever gotten is an 11 or 13, most of the time only 8s tho, and that just an example with the Crusader.
Also hit chance, 80% chance to hit is almost always 50/50, is dodge taken into account when calculating hit chance? Because I've missed six 80% chance hits in a row which would indicate to me that its not actually 80% chance to hit. Like with some Fire Emblem games that use different means of generating a number.


Explanation or have I just been unlucky? The tide turns too quickly for me.

[CutestGirlHere.]

For damage rolls, it's just random for what damage number you roll. There's no sort of weighting there, and your experience is most likely just an odd bit of luck, after a while it should start to even out with damage sread.

For accuracy however, there is a good bit of weighting going on. You see, the game always adds an extra hidden 5% Hit Chance to all Hero and Monster attacks that it doesn't tell you about. So someone with 80% Hit Chance actually has 85%, someone with 0% Hit Chance actually has 5% Hit Chance, while someone with 95% actually has 100%.

You missing 6 80%(actually 85% if you count hidden 5% bonus) attacks in a row is mostly just bad luck, it happens every now and again. If you want to avoid missing, you're supposed to try to stack +ACC or -Dodge to reach the 95% Threshold, then the bonus carries it over to 100 so you can't miss.

EDIT:Also yes, Dodge is taken into account. Hit Chance calculation is Accuracy - Dodge + 5 = Hit Chance.

[TSense]

Aother soul that doesn't understand how random works.

I will now list 10 random numbers, anywhere between 1 to 5:
5
5
5
5
5
5
5
5
5
1

This is random. It may not look random to you. But Random doesn't care what you or anyone else thinks. If you want it to "feel" more random, you would need to change the system to change the outcome based on what the outcome was before. But the system doesn't care what it already gave out when it comes up with a new number. It's random.
100% critz for the whole dungeon for everyone? Possible. Unlikly, yes, but remember: Random doesn't care what is likly to happen.

You cannot make any predictions about the next random output based on what already came.

A coin flip is 50%. You get heads. And heads again. And heads again. Does that change anything about your chances for the next throw? The coin doesn't "know" what it had before. Random doesn't care. 50% only means that throwing it *unlimited* amount of times will lead to half of them being tails.

[Gamer Supreme]

Maybe the ancestor has not told you about overconfidence enough :P

[No One]

This will happen to everyone who plays long enough. It's called RNG screw, and if it can happen, it eventually will happen. This is why I try to hit the ACC cap at all times. 80% is in fact a really low hit percentage. As you've found out.

It is far more common than not to mistakenly evaluate probability. Gather some actual numbers to see if you're really missing more often than you should.

[Your Neighbour]

As I always tell people, if you have a casino near you, go observe the roulette wheels, and you will find consecutive 10 rolls of black/red is not that uncommon. I think people just gets too deeply hammered when that 1/100th chance sets in to obliterate their vestel in week 80 while not seeing how many hits the leper actually dodged with -5 dodge.

[CrackerjackTiming]

I'm sure you'd have to dig into the actual game code to determine if all the die rolls really are 100% random, but I can't imagine why the devs would bother trying to deliberately weight it one way or the other. Unless you egregiously stack the rolls, like all your heroes dodging every attack for an entire long Champion dungeon when they supposedly have a Dodge of 0, most players would never notice the difference. Some people would still get runs of bad luck and immediately assume that the RNG is skewed or whatever.

Originally posted by Lowly Noob:

My experience is incredibly limited though so I might be entirely wrong and incredibly unlucky.

Yeah, your experience is limited. I can't imagine that anyone in history, ever, has played enough Darkest Dungeon to provide a statistically significant data set that would be useful for determining whether the RNG is "truly random" or not (protip, there's no such thing as 100% random except maybe in like weird quantum physics BS or whatever, only random enough that humans can't tell the difference). Even hundreds or thousands of hours probably wouldn't be enough. The only way I could see being able to compile enough data to get a good idea on the subject would be if Red Hook tracked every die roll on every player's game through Steam or something, which I'm not sure is even possible.

The point being, an 80% chance will EVENTUALLY work out to 80/20 outcomes, but is guaranteed to do so only if you repeat it a completely impractical number of times. When it comes to playing poker or Darkest Dungeon or whatever, all you can do is boost your chances of success as much as you can and then let the chips fall where they may. You can boost your heroes' ACC enough to land guaranteed hits, but you will never be able to get a 100% crit rate, nor will you ever be able to completely guarantee that enemies can't hit you in most circumstances except by stunning them, which is why stun is so good. You can't control the dice, so focus on the stuff you can control and you'll be fine regardless of how good or bad your luck seems to be. If people can do silly all-Leper no-trinket no-hit no-attack Bloodmoon speedruns and beat the Darkest Dungeon + all bosses on Week -5 before the Ancestor shoots himself in the opening cutscene or whatever, then the RNG can't be that bad.

I definitely feel your pain though. Throughout all the time I've played Darkest Dungeon, the mean healing I've gotten from Wyrd Reconstruction is probably somewhere around 2-3, with at least a third of all rolls being zeros. I've grown to be very reluctant to rely on Occultists even for just healing off Death's Door, when rationally I know that there's only a 7% chance of rolling a zero even when I don't upgrade Wyrd Reconstruction, and the fact that I use Wyrd Reconstruction less and less only makes it less likely that I will see the average result ever increase. It's one of those things where game theory runs counterintuitive to the human instinct of once burned, twice shy.

[No One]

Every time you miss you get an invisible +4% to hit until the next time you hit. They do fudge the rolls, in the player's favour. Not to mention the constant invisible +5% to hit, which used to be 10%.

I regularly hit 100% hit rate. Used to display as 90%, now it's 95%.

They key to healing with occultists is to never hit death's door. It's not that hard. Heal early, heal often, and do something to make bleed = 0%. I make sure my heroes outlevel wyrd.

[CrackerjackTiming]

Originally posted by No One:

Every time you miss you get an invisible +4% to hit until the next time you hit. They do fudge the rolls, in the player's favour. Not to mention the constant invisible +5% to hit, which used to be 10%.

Really? Interesting. I wouldn't count the invisible +5% to hit chance as being in the player's favor, though, because my understanding is that it applies to monsters too.

Originally posted by No One:

They key to healing with occultists is to never hit death's door. It's not that hard. Heal early, heal often, and do something to make bleed = 0%. I make sure my heroes outlevel wyrd.

Is it really worth it to ever upgrade Wyrd Reconstruction? I'm no mathologist, but it doesn't seem like raising the maximum healing by a few points is worth the bleed side-effect getting worse.

[CutestGirlHere.]

Originally posted by author:

Is it really worth it to ever upgrade Wyrd Reconstruction? I'm no mathologist, but it doesn't seem like raising the maximum healing by a few points is worth the bleed side-effect getting worse.

Every hero's resist stats go up another 10% every time they level up. So a level 6 hero will usually wind up with around 90-100 Bleed Resist on average, while a max level Wyrd Reconstruction only has an 85% Bleed Chance. So basically almost everyone would be immune to the Occultist's Bleed by that point, with the exception of the Plague Doctor(only 80% Bleed Resist) and the Leper(70% Bleed Resist).

Additonally, raising the maximum healing makes it a lot less likely for the Occultist to roll a 0. At level 1 there's a 7% chance of him rolling a 0(since it's a scale of 0-13), while at level 5 there's only a 4% chance(since it's a scale of 0-22).

[CrackerjackTiming]

Originally posted by Power.:

Additonally, raising the maximum healing makes it a lot less likely for the Occultist to roll a 0. At level 1 there's a 7% chance of him rolling a 0(since it's a scale of 0-13), while at level 5 there's only a 4% chance(since it's a scale of 0-22).

Again, not a mathologist, but a 3% difference doesn't sound like "a lot." Remember, you could instead spend that money putting a hero in the Sanitarium and then immediately canceling the treatment, which would increase the Occultist's healing ability by roughly the same amount in practice.

[CutestGirlHere.]

Gold is ridiculously plentiful, and by the time you actually get level 5 upgrades you should be drowning in money. You're not exactly wasting much money in exchange for a 3% difference and increasing the size of your heals another 9 health. Also what does the Sanitarium have to do with that?Unless you're locking in Hippocratic it's not going to be increasing his healing(especially not if you cancel locking it in, it's just gonna risk being replaced later), and even any +Healing Trinkets and Quirks don't change your chances of rolling a 0 or not, only upgrading the skill does.

You get the potential for much larger heals, a reduced chance of a 0, and unless your heroes have -Bleed Resist trinkets/diseases/quirks there's no chance of Bleeding(aside from the few exceptions), it's just improving the skill in every way pretty much.

[CrackerjackTiming]

Originally posted by Power.:

Gold is ridiculously plentiful, and by the time you actually get level 5 upgrades you should be drowning in money. You're not exactly wasting much money in exchange for a 3% difference and increasing the size of your heals another 9 health. Also what does the Sanitarium have to do with that?

It was just a joke.

You're correct, I just don't like Wyrd Reconstruction because I mostly roll zeros regardless of what level the ability is. Like seriously, 50%+ zeros. It's really bad.

[CutestGirlHere.]

OH, my bad. Really tired so I didn't realize it was a just a joke at first.

[No One]

Originally posted by CrackerjackTiming:

Is it really worth it to ever upgrade Wyrd Reconstruction? I'm no mathologist, but it doesn't seem like raising the maximum healing by a few points is worth the bleed side-effect getting worse.

All I have to say is this: 0% chance of 3 point bleed vs. 0% chance of 9 point bleed.

[Satoru]

Please note that while the Eldritch Gods do love to toy with humanity, the RNG Gods are in fact entirely neutral and are not out to get you. They care not what the dice roll. They are true neutral gods.

That being said, the game design itself always has a constant downward pressure on you. This means that even moderately bad rolls push you down, which has a potential negative feedback loop. But again this is a conscious design philosophy in the game. You're not 'supposed' to be winning.

Most 'RNG' in games is in fact 'fake RNG' to make you feel better most of the time.

https://youtu.be/MtzCLd93SyU?t=1173

Skip to 19:00. But basically Sid, being a pure engineer and game designer of course made a totally neutral RNG system for Sid Rev. But play testers didn't actually understand that you know a 50% chance to iwn is still 50%. And that you can still lose 2-3 battles in a row, even at 50%. So they added weighting into the RNG to make it 'feel better' to their players. Lose a battle? You get additional bonuses to later battles to 'make up' for it. Despite this not being actually 'random'.

DD actually gives you a minor bonus to future rolls if you keep missing.