RNG questions

[CynicalCanadian]

So as CRPG with dice rolls as a system RNG is going to be a thing. But it seems like this one may need some tweeking. 20 sided die means that you on average have a 50% chance to roll above a 10. But it does not feel that way. I was missing alot and figured it was just a min/max issue but once i started paying attention to the rolls, it just seems buggy. Maybe I am just super unlucky, but I feel rolling a 1, 8 times in a row, across my 6 party members, is not 50% chance to roll 10 or over, but like 5%.

[Razorblade]

Pro tip: don't try to analyze RNG if you're not a data scientist/statistician; there's a reason people get paid good money for that.

Unless you want to conduct the hundreds of test rolls required to prove your theory, it's not worth even positing. Anyone with half a brain is just going to tell you "it's confirmation bias" and move on.

[T.W. Hamill]

There was a bug in Kingmaker (that they patched) that led to the seed it uses to fire up the RNG leading to biased rolls.

It may have crept back in to Wrath since it seems like a LOT of the fixes they did to the engine for Kingmaker arent in the Wrath client, like they started back over with the base client instead of building on Kingmaker.

Try saving, quitting out to desktop, and restarting. It forced it to re-seed the RNG (Random Number Generator) script.

[T.W. Hamill]

Diposting pertama kali oleh Razorblade:

Pro tip: don't try to analyze RNG if you're not a data scientist/statistician; there's a reason people get paid good money for that.

Unless you want to conduct the hundreds of test rolls required to prove your theory, it's not worth even positing. Anyone with half a brain is just going to tell you "it's confirmation bias" and move on.

You can dump the combat log to a text file and just use "find" in a text editor to see how many 1's you rolled pretty easily.

I remember in the pre-fixed Kingmaker i did that and one run i was only getting rolls above 10 about 25% of the time.

It doesnt affect the enemies as much because in a lot of cases their stat blocks are so inflated, they still hit even with low rolls.

[CynicalCanadian]

Diposting pertama kali oleh Razorblade:

Pro tip: don't try to analyze RNG if you're not a data scientist/statistician; there's a reason people get paid good money for that.

Unless you want to conduct the hundreds of test rolls required to prove your theory, it's not worth even positing. Anyone with half a brain is just going to tell you "it's confirmation bias" and move on.

So did a dump, on over 5000 rolls, character with 1 attack action, averaged over 10 on a roll 35% of the time, those with 2x were 25% of the time, those 3x were 15% of the time. On Sheela with over 1500 attack rolls, she scored more critical misses then she did rolls over 20. Owlcat has something wrong with rolls. And considering how many of the fights are stat padded so high to the point that RNG determine the outcome 99% of the time, that RNG system should at least work as intended.

[T.W. Hamill]

Diposting pertama kali oleh \0/CynicalCanadian\0/:

And considering how many of the fights are stat padded so high to the point that RNG determine the outcome 99% of the time, that RNG system should at least work as intended.

Uhh.. the vast majority of fights, your bonuses will be nearly the size of the number you need to roll, unless you built like a potato.

[DarkFenix]

Diposting pertama kali oleh \0/CynicalCanadian\0/:

Diposting pertama kali oleh Razorblade:

Pro tip: don't try to analyze RNG if you're not a data scientist/statistician; there's a reason people get paid good money for that.

Unless you want to conduct the hundreds of test rolls required to prove your theory, it's not worth even positing. Anyone with half a brain is just going to tell you "it's confirmation bias" and move on.

So did a dump, on over 5000 rolls, character with 1 attack action, averaged over 10 on a roll 35% of the time, those with 2x were 25% of the time, those 3x were 15% of the time. On Sheela with over 1500 attack rolls, she scored more critical misses then she did rolls over 20. Owlcat has something wrong with rolls. And considering how many of the fights are stat padded so high to the point that RNG determine the outcome 99% of the time, that RNG system should at least work as intended.

Other people have performed statistical analyses of the rolls and found a completely normal random distribution. Yet you have apparently not only found that they're weighted, but are extra-weighted based on arbitrary conditions.

You should try a better lie, when you tell one as egregious as characters with 3 attacks only rolling a 10+ 15% of the time, it doesn't even need a statistical analysis to know that's false. If it were that low it would be plainly observable.

[CynicalCanadian]

Diposting pertama kali oleh T.W. Hamill:

Diposting pertama kali oleh \0/CynicalCanadian\0/:

And considering how many of the fights are stat padded so high to the point that RNG determine the outcome 99% of the time, that RNG system should at least work as intended.

Uhh.. the vast majority of fights, your bonuses will be nearly the size of the number you need to roll, unless you built like a potato.

Uhh, that is not the point, the point is that on d20 you have a 50% chance to roll over 10, and the game is not like its around 40% base and it gets lower the more attacks you get. But yea fan boy ignoring the entire point, and trying to blame the person for bad game design, BTW its the same bug that took them a year to fix on Kingmaker. lol

[CynicalCanadian]

Diposting pertama kali oleh DarkFenix:

Diposting pertama kali oleh \0/CynicalCanadian\0/:

So did a dump, on over 5000 rolls, character with 1 attack action, averaged over 10 on a roll 35% of the time, those with 2x were 25% of the time, those 3x were 15% of the time. On Sheela with over 1500 attack rolls, she scored more critical misses then she did rolls over 20. Owlcat has something wrong with rolls. And considering how many of the fights are stat padded so high to the point that RNG determine the outcome 99% of the time, that RNG system should at least work as intended.

Other people have performed statistical analyses of the rolls and found a completely normal random distribution. Yet you have apparently not only found that they're weighted, but are extra-weighted based on arbitrary conditions.

You should try a better lie, when you tell one as egregious as characters with 3 attacks only rolling a 10+ 15% of the time, it doesn't even need a statistical analysis to know that's false. If it were that low it would be plainly observable.

Not a lie, I have no reason to lie? I want it to be balanced, and its not. And btw 8 critical missed in a row for a 2% chance to roll is a statistical anomaly if its balanced like a d20, and yet it happened more then two times. And clearly someone knows nothing about the RNG bug Kingmaker had, and took Owlcat over a year to fix. But yea its a totally perfect rng system that constantly gets brought up by hundreds of people, on a level no other game with a dice rolling system has seen. Ok fan boy.

[Gracey Face]

Do you have any evidence kingmaker had an RNG glitch? This is the first I have heard of it.

[Red Phantom]

I swear these threads are contagious

[Planeforger]

Diposting pertama kali oleh \0/CynicalCanadian\0/:

So did a dump, on over 5000 rolls, character with 1 attack action, averaged over 10 on a roll 35% of the time, those with 2x were 25% of the time, those 3x were 15% of the time. should at least work as intended.

Any tips on making one of those dumps?

I've tried manually tracking the numbers a few times, and the numbers (after hundreds of rolls) seem to even out to 50% 1-10 and 50% 11-20...but I'd be keen to try it on a larger scale.

[Skor]

Diposting pertama kali oleh \0/CynicalCanadian\0/:

Diposting pertama kali oleh Razorblade:

Pro tip: don't try to analyze RNG if you're not a data scientist/statistician; there's a reason people get paid good money for that.

Unless you want to conduct the hundreds of test rolls required to prove your theory, it's not worth even positing. Anyone with half a brain is just going to tell you "it's confirmation bias" and move on.

So did a dump, on over 5000 rolls, character with 1 attack action, averaged over 10 on a roll 35% of the time, those with 2x were 25% of the time, those 3x were 15% of the time. On Sheela with over 1500 attack rolls, she scored more critical misses then she did rolls over 20. Owlcat has something wrong with rolls. And considering how many of the fights are stat padded so high to the point that RNG determine the outcome 99% of the time, that RNG system should at least work as intended.

Sounds about how my first game went. Interestingly enough, my second game wasn't nearly as bad. But yeah, I feel your pain. Unfortunately, the 'statistics keyboard warriors' will talk down to you if you bring up anything about it.

[jhughes]

Diposting pertama kali oleh Planeforger:

Diposting pertama kali oleh \0/CynicalCanadian\0/:

So did a dump, on over 5000 rolls, character with 1 attack action, averaged over 10 on a roll 35% of the time, those with 2x were 25% of the time, those 3x were 15% of the time. should at least work as intended.

Any tips on making one of those dumps?

I've tried manually tracking the numbers a few times, and the numbers (after hundreds of rolls) seem to even out to 50% 1-10 and 50% 11-20...but I'd be keen to try it on a larger scale.

Combat log text file is in:
C:\Users\<user>\AppData\LocalLow\Owlcat Games\Pathfinder Wrath Of The Righteous\”

[jhughes]

Diposting pertama kali oleh \0/CynicalCanadian\0/:

Diposting pertama kali oleh T.W. Hamill:

Uhh.. the vast majority of fights, your bonuses will be nearly the size of the number you need to roll, unless you built like a potato.

Uhh, that is not the point, the point is that on d20 you have a 50% chance to roll over 10, and the game is not like its around 40% base and it gets lower the more attacks you get. But yea fan boy ignoring the entire point, and trying to blame the person for bad game design, BTW its the same bug that took them a year to fix on Kingmaker. lol

That is Untrue.. What you said does not happen all the time. The RNG is an RNG. It is 'POSSIBLE' that in 5000 rolls in that you roll less than ten 40% of the time. It is not very probable and it will not happen all the time. Stating otherwise is just ♥♥♥♥♥♥♥♥.