Yea, roll RNG is not rigged...

Deep Blue
Level 3
5 months ago (edited)

I'm an avid PnP D&D player; discovered Solasta by accident, over 50 hours in now and i LOVE it - graphics, music, and especially gameplay all seem really great (i would love to see more content, maybe a DLC or two). But one thing is REALLY bugging me. The dice rolls seem broken, even more so than in other D&D CRPGs.

I had to lower the difficulty from cataclysm and set the game to 'karmic dice' but even after doing so, i have to save scum to win some fights simply because of the luck of 'true random' rolls.

[Spoiler]

In the volcano boss fight i tried to paralyze the boss with an arrow before his turn, needed for him to roll 10 or less for effect to proc. Seems like a piece of cake with 2 shots on a ranger, isn't it?

WRONG.

Spent 5 minuted save scumming and got ONE proc. ONE roll of below 11 out of a dozen or so rolls.

So... i got to thinking if i just got really unlucky and decided to try and record the number of rolls.

True RNG?

17 minutes, 50 shots at the boss.

Expected result: 50% of the rolls below 11, 50% at, or above - about 25/25 spread, give or take.

Actual result: 34% of the rolls below 11, 66% at, or above, 17/33 spread not in my favor.

[/Spoiler]

It wouldn't be an issue if it was just a single occurrence, but the rolls actually seem skewered not in players' favor in all instances - speech checks, skill checks, attack checks - unless you increase them in your favor in the settings. 

I know the issue was talked about extensively, but maybe you could make the rolls a wee bit skewered in the other direction? After all, people play video games to have fun and having to save scum because of RNG is never fun %)



Brie Longbow
Level 4
5 months ago

I agree with you that the system they use is ANYTHING but random. 

I have even gone so far as to roll a d20 myself to compare it to the computer and over 65% favor the computer and not me while it is around 54% favor me on the d20 rolls. 

We had this problem in XCom 2 random rolls my azz. 

I am almost certain the game is rigged and I know that computer software can't in fact be random with dice rolls. 

To many times I have needed 1 hit to kill the minion and I will miss all my attacks (like 8 attacks by my 4 characters) he then rolls a 20 to kill one of my toons. 


sunseekers
Level 5
5 months ago

Perhaps it is just bad unlucky dierolling ?   Every time I really need a good shot with my ranger or anyone with a bow or crossbow, I roll 1's & 2's.  My Dwarf can't hit anything also but at least he's got 10 Dex with no modifiers. 

R4D4
Level 1
5 months ago

Here's a thought:
If its true that its not very random and you get bad rolls, that must surely also go for the enemies as well. so its still fair.

IXI
Level 14
5 months ago

 After all, people play video games to have fun and having to save scum because of RNG is never fun %)

while i like that D&D and other RPGs have become mainstream, the downside is most people, especially newer players did not cut their teeth on 1st D&D.  modern games are overly gentle in comparison.  so when have a crappy experience i just remember, it used to be a lot worse


yellow flower of courage

Silverquick
Level 14
5 months ago (edited)

I agree with you that the system they use is ANYTHING but random. 

I have even gone so far as to roll a d20 myself to compare it to the computer and over 65% favor the computer and not me while it is around 54% favor me on the d20 rolls. 

We had this problem in XCom 2 random rolls my azz. 

I am almost certain the game is rigged and I know that computer software can't in fact be random with dice rolls. 

To many times I have needed 1 hit to kill the minion and I will miss all my attacks (like 8 attacks by my 4 characters) he then rolls a 20 to kill one of my toons. 


Yeah the Devs are going to have to fix this. 

If it wasn't so blatant when it happens it would be different, they are going to have to fix this because its literally ruining people's experience. 

And that's a bad thing. 

Deep Blue
Level 3
5 months ago

Here's a thought:
If its true that its not very random and you get bad rolls, that must surely also go for the enemies as well. so its still fair.

Even if that was true (and from my experience it's not), the problems with that thought are

1) Your party of 4 isn't always equal in strength and quantity to enemies' party. For harder difficulty where 1 normal hit from an enemy often equals half of your tank's health it means that you always try to leverage the odds in your favor with careful planning of the combat. And poor dice rolls, even if they are mutually poor, will screw you over much more than your enemy, since they already got the advantage.

2) The rolls in battle are not the only ones, having no disadvantages on a skill/speech checks, i fail about 70% of the rolls that i expect to pass. Most of those checks are probably just flavor but still it kinda blows when u fail a 95% success chance check.

5 months ago

Okay, another thread accusing the game of cheating die rolls against the player.  Their evidence?  Some cherry picked data & extremely small data sets.  One poster above said they tested a whopping 50 rolls, when it takes hundreds, even thousands, to get a big enough data set to verify something like this.  (The default minimum data set size these days is 10K to test anything.)


In D&D, with its d2 system, any one result is a 5% chance.  In statistical analysis, 5% is a lot.  You could string 3 5% chance results in a row (like 3 consecutive 1s) & have a chain with a 1in 8,000 chance - & yet that would not be an unusual result in a 10K sample size.


it has been demonstrated by social scientists in multiple tests & experiments that the human mind deals poorly with randomness.  Threads like these just prove it.


"I live to serve."

Silverquick
Level 14
5 months ago

I am someone who tests "thousands of rolls". 

Deep Blue
Level 3
5 months ago

Okay, another thread accusing the game of cheating die rolls against the player.  Their evidence?  Some cherry picked data & extremely small data sets.  One poster above said they tested a whopping 50 rolls, when it takes hundreds, even thousands, to get a big enough data set to verify something like this.  (The default minimum data set size these days is 10K to test anything.)


In D&D, with its d2 system, any one result is a 5% chance.  In statistical analysis, 5% is a lot.  You could string 3 5% chance results in a row (like 3 consecutive 1s) & have a chain with a 1in 8,000 chance - & yet that would not be an unusual result in a 10K sample size.


it has been demonstrated by social scientists in multiple tests & experiments that the human mind deals poorly with randomness.  Threads like these just prove it.

Prove me wrong. Please. I'd be glad to see that i am, in fact, biased and the game is completely fair. Any reasonable number of consecutive rolls at any point in the game that favor the player. 

"Don't be an armchair critic by arguing your point without corroborationdon't cry for the moon but instead do something to prove that what you have insisted on"

Silverquick
Level 14
5 months ago

That won't work anymore, 

Too many people know now. If it wasn't so blatant when it happens it would be different, but its so blatant and obvious its predictable when it will happen. 

5 months ago (edited)

Part of the problem here is that players believe that “random” means “equally distributed”.  It doesn’t.  The baseline distribution that players are expecting in a sequence of 20 rolls is something along the lines of each of the 20 rolls appearing once.  When they see duplicated rolls in their sequence, they assume that the game is cheating.  Let’s examine that assumption using probability math, first taught in US students in 5th grade or so.

The first roll in a sequence can be anything, as each result has a 1 in 20 chance.  In order to achieve an equal distribution (& this avoid duplication) the next roll can’t be the same as the first, so it must be one of the other 19 - hence a 19 in 20 chance.  The following roll can’t be the same as the first 2, so it must be one of the other 18 - hence a 18 in 20 chance.  And so on down the line to the 20th roll, which can only be the one roll remaining - hence a 1 in 20 chance.

So the overall probability of the chain is the multiple of all the individual chances - 19/20 x 18/20 x 17/20 … on down to x 1/20.  Our 5th grade math tells us that this can be summarized as 19!/20^19 (where ! is the factorial of a number, which is the product of the number times each of the integers down to 1).

Calculating this out, we get 121,645,100,408,832,000/

5,242,880,000,000,000,000,000,000

which is 0.0000023201965953125%

or a 1 in 43,099,804 chance.

So the baseline expectation of an even distribution that players have is hilariously less likely than a result of 3 consecutive 20s or 1s, which would be 1 in 8,000 or 0.0125%.  The simple fact of the matter is that almost any sequence with duplications is far more likely than an even distribution.

Sure, in a large enough sample size, the overal distribution will trend towards a more-or-less even distribution, but that will be composed of a myriad number of smaller sequences that are NOT evenly distributed.  Lucky or unlucky streaks are a fact of existence - & instead of being proof of non-random behavior, they are instead a validation of randomness.  It might not be what our minds expect, but it is true nonetheless.

[Edited to correct a math error, that does not change the conclusion.]


"I live to serve."

Deep Blue
Level 3
5 months ago (edited)

Part of the problem here is that players believe that “random” means “equally distributed”.  It doesn’t.  The baseline distribution that players are expecting in a sequence of 20 rolls is something along the lines of each of the 20 rolls appearing once.  When they see duplicated rolls in their sequence, they assume that the game is cheating.  Let’s examine that assumption using probability math, first taught in US students in 5th grade or so.

The first roll in a sequence can be anything, as each result has a 1 in 20 chance.  In order to achieve an equal distribution (& this avoid duplication) the next roll can’t be the same as the first, so it must be one of the other 19 - hence a 19 in 20 chance.  The following roll can’t be the same as the first 2, so it must be one of the other 18 - hence a 18 in 20 chance.  And so on down the line to the 20th roll, which can only be the one roll remaining - hence a 1 in 20 chance.

So the overall probability of the chain is the multiple of all the individual chances - 19/20 x 18/20 x 17/20 … on down to x 1/20.  Our 5th grade math tells us that this can be summarized as 19!/20^20 (where ! is the factorial of a number, which is the product of the number times each of the integers down to 1).

Calculating this out, we get 121,645,100,408,832,000/

104,857,600,000,000,000,000,000,000

which is 0.000000116009807976573%

So the baseline expectation of an even distribution that players have is hilariously less likely than a result of 3 consecutive 20s or 1s, which would be 1 in 8,000 or 0.0125%.  The simple fact of the matter is that almost any sequence with duplications is far more likely than an even distribution.

Sure, in a large enough sample size, the overal distribution will trend towards a more-or-less even distribution, but that will be composed of a myriad number of smaller sequences that are NOT evenly distributed.  Lucky or unlucky streaks are a fact of existence - & instead of being proof of non-random behavior, they are instead a validation of randomness.  It might not be what our minds expect, but it is true nonetheless.

Interesting speech. Would be more interesting if it had any relation to the topic at hand. So far i only see random numbers strung together with some empty rhetoric. But I'm a reasonable man, i will give you another chance to redeem yourself. 

Show me ANY proof that THIS PARTICULAR GAME does an even remotely fair random numbers distribution.

Just in case the above sentence is too hard to comprehend, I'll reiterate: "Proof where?" (c) WehSing

5 months ago (edited)

You apparently missed the point of my post.  What you apparently think is a “random” distribution is not in fact random. It is an extremely unlikely result - much less likely than the streaks that players have been complaining about in this thread.  I don’t have to prove that this game’s results are random, because I have already shown that they don’t fall outside the bounds of normal random distribution. You need to prove your conspiracy theory that the game results are not random.  I have proven mathematically that the results complained about are well in the realm of normal random distribution.  When are you going to prove that they are not?

Just in case this post was too hard to comprehend, let me reiterate: I have submitted a proof of my point.  When are you going to make even a token effort to try to prove your point?  


"I live to serve."

Deep Blue
Level 3
5 months ago

Well, i gave you a shot, you blew it; 

I provided data, even if it was 'cherry picked'. All you did was spout number theories. Not gonna argue anymore, hope the devs are smarter that this and will actually address the problem of roll RNG.

Though am still amazed at the empty rhetoric at it's finest - "There is a chance that anything can happen so the game's rolls are fair, I've proven my point beyond any doubt" 




This thread should now be closed; it will serve no other purpose from now on other than provoke previous poster to stroke his own... ego... by making more speeches about how everything and anything can happen or spouting insults. 

5 months ago (edited)

You are correct in one thing - this thread can serve no further purpose.  You have no interest in reading any point of view other than one that reinforces your previously determined point of view & flatters your (apparently not inconsiderable) ego.

I am amused though that someone who repeatedly ended his posts with quotes about providing proofs never once did so in this the thread.  And this is the real reason this thread should be closed - you can’t have a debate where one side makes rational arguments & the other figuratively puts its hands over its ears & says “Nyah Nyah Nyah - your logic & facts are irrelevant.” 🤷‍♂️


"I live to serve."