Larian Banner
Previous Thread
Next Thread
Print Thread
Page 2 of 13 1 2 3 4 12 13
Joined: May 2010
Location: Oxford
Duchess of Gorgombert
Offline
Duchess of Gorgombert
Joined: May 2010
Location: Oxford
Originally Posted by Sven_
Everybody should experience this quite frequently. Even missing that 75% chance four times in a row is not any higher than a 1 in 256 chance, hugely far from winning the lottery.

The trouble is, it keeps happening; i.e. much more often than 1:256. I wouldn't like to commit to a figure, but it happens way too often; certainly more so than landing a critical hit, for instance.

The other thing is it seems rather redundant as even if you do pass the chance-to-hit, it then rolls again to see if the target is realistically damaged or just a glancing blow.

Of course it may simply be communicated to the player rather poorly: if perhaps it's "you have a 75% chance all things being equal, but now I'll factor things like respective skills, strengths and other stuff that's going on into the equation" which would render that number rather meaningless.


J'aime le fromage.
Joined: Jan 2018
W
old hand
Offline
old hand
W
Joined: Jan 2018
Yeah, confirmation bias plays a huge part in this. You are far more likely to remember missing multiple times than hitting consecutively.

Also, you would probably hit as many times on a 20% chance as you miss on an 80% chance, but since most people choose not to attack if their odds are that bad and just take a different option, the data you are working with is incomplete and skewed.

If you are really missing 14 times in a row there could be something up with your game, maybe you somehow triggered a bug, or it could be one of those improbabilities that when considered on the cosmic scale are virtually inevitable to happen from time to time.

Personally I have not noticed anything wrong with the RNG. I’ve missed on a 98% once, and I have also gotten 3 consecutive critical hits. Unlikely things happen from time to time, but generally my results match the expressed success percentages.

Last edited by Warlocke; 19/10/20 05:59 AM.
Joined: Jun 2020
enthusiast
Offline
enthusiast
Joined: Jun 2020
Originally Posted by vometia

Of course it may simply be communicated to the player rather poorly: if perhaps it's "you have a 75% chance all things being equal, but now I'll factor things like respective skills, strengths and other stuff that's going on into the equation" which would render that number rather meaningless.


From my experience, it isn't. They even factor in the percentage when considering advantage/disadvantage on the attack roll, which is why you have displayed chances such as 99%, 36% or 88% given that you get to throw the dice twice and have to take the higher/lower roll.
http://zerohitpoints.com/Articles/Advantage-in-DnD-5#:~:text=Shown%20for%20each%20number%20on,rolling%20a%2010%20or%20greater.

I personally can't relate to what you are experiencing at all. Given that you have expressed to be very unlucky in general, this may be mostly confirmation bias ("Ah, see, I'm unlucky again" -- except for the many times you aren't). smile

Last edited by Sven_; 19/10/20 06:01 AM.
Joined: Oct 2017
enthusiast
Offline
enthusiast
Joined: Oct 2017
I will just quote what I said in another thread
Originally Posted by Sharp

As far as I am aware the currently displayed hit chances are accurate and the dice are not weighted in any direction, although they are considering weighting them in the player's favor in future. If they do that, I would rather have a difficulty option with non weighted dice, since I am not a fan of weighting to begin with. The human mind is notoriously bad at grasping probability, there are a lot of studies on this. So even if the hit chance is perfectly accurate, people will "feel" it is wrong, because our grasp of probability as a species is flawed. If you wish to read up on the subject, here is an example of a very simple puzzle to which most humans will choose incorrectly, even though pigeons will choose the correct answer. Here is another paper on the topic. Game designers are also well aware of this, the probability of hitting is skewed in the player's favor in XCOM for example. My favorite example though is probably this talk given by Tim Cain, at 25 minutes into this video.

I see someone else has mentioned that video already, but the other links are also worthwhile perusing.

Joined: Oct 2020
V
member
Offline
member
V
Joined: Oct 2020
I am not saying there is nothing wrong with the chances displayed, but as always I'd wager the overwhelming part of frustration is confirmation bias and human psychology. A 75% chance is not a guaranteed hit, yet it is treated that way. You wouldn't take odds like that in real life. Say there is a puddle on the road, you are wearing very nice shoes and you would guess that you are likely to be able to jump it, yet would most likely just step around it.

And with the amount of rolling in this game with varied chances the odds of having a bad streak every once in a while is high enough, and that is what sticks in your mind. Not the times four 75% hits connected in a row.

Joined: Jun 2020
enthusiast
Offline
enthusiast
Joined: Jun 2020
Originally Posted by VincentNZ

And with the amount of rolling in this game with varied chances the odds of having a bad streak every once in a while is high enough, and that is what sticks in your mind. Not the times four 75% hits connected in a row.


Additioinally, you are going to find many players who experience -- in some way or another -- just the same as you (this includes my own experience too). This is a game meanwhile played by hundreds of thousands, despite the Early Access. So if you're looking for a confirmation, you'll find it. I'm almost tempted to do some actual testing now, armed with pen and paper (heh) but as said, I can't see anything obvious "off". I think this is a really fascinating topic though, in particular because of human psychology at play. Without it, casinos and gambling outlets would be lost too (gambler's fallacy, etc.)

I've seen players claiming that the dice were actually rigged right on these boards, insofar as that the opposition would barely ever miss, which obviously also isn't happening, certainly not over the mid to longer term. Over a sequence or two, much goes (and will if this is truly random). Additionally, actually rigged dice against the player would be majorly stupid from a developing standpoint, as it would undermine all player trust -- there's much easier and "Fair" ways to make the game hard. Such as always putting the player at a numeral and positioning disadvantage, giving creatures abilities they don't have in the pen&paper, etc. (which BG3 sometimes does too).

If you'd roll the dice long enough, you're guaranteed curious sequences. smile






Last edited by Sven_; 19/10/20 08:47 AM.
Joined: Oct 2020
Location: Italy
member
Offline
member
Joined: Oct 2020
Location: Italy
I missed 6 time in a row with 81% hit chance. Nothing is impossible but here the probability is less than 10^-3.
The RNG is screwd, everyone noticed it. Truth to be told, RNG is screwd in most software, non only videogames, but there are tricks to level the distribution of outcome. I'm sure Larian's guys and gals are not stupid, if they are going to put their heads on it they will find a solution (DOS2 didn't have such a problem).

The only thing to hope for is that they are reading these posts laugh

Joined: Mar 2020
Location: Belfast
old hand
Offline
old hand
Joined: Mar 2020
Location: Belfast
Originally Posted by Eddiar
30% I understand.
50% I also understand.
But 75% and over? Really?

And how do you miss with the guy standing right in front of you? Why not just say dodge? That would make me feel a little better.

The question is: what do those percentage number represent? Those are adaptation of d20 rolls and include things like advantage (rolling twice And picking higher number). Unlike XCOMs or PoEs it doesn’t represent actual calculations behind the hit. Perhaps 75% was a double d20vs10 roll?

I would recommend jumping into combat log next time (can be expanded in bottom right corner) and see for yourself what actually happened.

And no matter what your chances are rolling 1 will always fail, and rolling 20 will always be a critical hit.

Last edited by Wormerine; 19/10/20 09:27 AM.
Joined: Jun 2020
enthusiast
Offline
enthusiast
Joined: Jun 2020
Originally Posted by Sharet
I missed 6 time in a row with 81% hit chance. Nothing is impossible but here the probability is less than 10^-3.


This is a 1 in ~20.000 chance, not that hugely far off Tim Cain's Fallout example. It's long odds, but given the number of rolls and players, it should occur with some frequency.

Last edited by Sven_; 19/10/20 09:36 AM.
Joined: Oct 2020
journeyman
Offline
journeyman
Joined: Oct 2020
Every game with randomness gets these same posts. And in none of those games, this one included, I haven’t encountered anything weird with how chance to hit works. Sure, I miss attacks, but never felt it’s ”more than supposed”.

Played through this Early Access (with several characters) and failed a bunch of rolls along the way. That’s expected. Failed some quite important rolls in and out of combat. That’s how it’s supposed to work. But nothing prevented me continuing the game. I help downed companione back up, resurrect dead ones, lick my wounds at camp, and then continue playing.

I don’t doubt that fail streaks happen to people, that’s expected as well. With tens of thousands of players rolling hundreds of rolls all the time, some of them are likely to get those nasty streaks. But I do think that many of us remember the negative outcomes and forget the positive ones.

Joined: Oct 2020
Location: Italy
member
Offline
member
Joined: Oct 2020
Location: Italy
Originally Posted by Sven_
Originally Posted by Sharet
I missed 6 time in a row with 81% hit chance. Nothing is impossible but here the probability is less than 10^-3.


This is a 1 in ~20.000 chance, not that hugely far off Tim Cain's Fallout example. It's long odds, but given the number of rolls and players, it should occur with some frequency.


The problem is that it happens consistently to the same player and across the player base. As for the fact that if you have a 3:1 critical miss/critical success ratio is indicative the RNG doesn't work properly.

Last edited by Sharet; 19/10/20 09:38 AM.
Joined: Oct 2020
member
Offline
member
Joined: Oct 2020
Originally Posted by Sharet
Originally Posted by Sven_
Originally Posted by Sharet
I missed 6 time in a row with 81% hit chance. Nothing is impossible but here the probability is less than 10^-3.


This is a 1 in ~20.000 chance, not that hugely far off Tim Cain's Fallout example. It's long odds, but given the number of rolls and players, it should occur with some frequency.


The problem is that it happens consistently to the same player and across the player base. As for the fact that if you have a 3:1 critical miss/critical success ratio is indicative the RNG doesn't work properly.


Or rather, players are anecdotally relating their experience of this happening to them which is not necessarily the same thing as it actually happening. Unless we look at the numbers (the game engine numbers, not the statistical math based on the anecdotes) this is just a bunch of people going "I feel like...". Which is fine and can totally be relevant feedback but in terms of specific criticism of a random number generator "I feel like I miss a lot" isn't really too meaningful due to, amongst other things, Gambler's Fallacy and Confirmation Bias.

Sometimes I get really annoyed at my companions too because it feels like they mess up a lot but as a gut feeling from looking at the stats of NPCs and pretty solid knowledge of the D&D system, I'm fairly confident (but unable to provide solid evidence for) that the system is working as intended and that if anything needs tweaking it can probably be fixed by modifying enemy statistics rather than looking at the RNG.

Last edited by Khorvale; 19/10/20 10:02 AM.
Joined: Jun 2020
enthusiast
Offline
enthusiast
Joined: Jun 2020
Originally Posted by Sharet
[quote=Sven_]

The problem is that it happens consistently to the same player and across the player base.


Depends on the playing hours and number encounteres, and specifically the number of rolls, but this could very well happen to the same player, yes, say every week or such. It's nowhere close to winning the lottery or any other hugely unlikely event.

An 80% chance would be akin to needing to roll at least a 5 on the D20.[Linked Image] I came comparably close to not doing that 6 times in a row just on my first fresh attempt on rolladie.net.



Last edited by Sven_; 19/10/20 10:05 AM.
Joined: Sep 2020
addict
Offline
addict
Joined: Sep 2020
Dice rolling tends to do this. Good thing about it is that enemies have the same string of bad luck that we do. Had so many fights where they all missed for a whole round.

Joined: Aug 2014
T
member
Offline
member
T
Joined: Aug 2014
In PnP (3.5) there were plenty of ways to boost your spell DC. So far in this game I feel like its a total waste to cast crowd control because your chance of landing the spell is like 60% at most. Its wierd that it feels like this because the cantrips (or any damage spell) have similar chance to hit. I guess it matters if its a cantrip that you can cast ad infinitum but even with a hardhitting finite spell I would rather cast that at 60% than a CC spell.

Joined: Oct 2020
member
Offline
member
Joined: Oct 2020
Originally Posted by Torque
In PnP (3.5) there were plenty of ways to boost your spell DC. So far in this game I feel like its a total waste to cast crowd control because your chance of landing the spell is like 60% at most. Its wierd that it feels like this because the cantrips (or any damage spell) have similar chance to hit. I guess it matters if its a cantrip that you can cast ad infinitum but even with a hardhitting finite spell I would rather cast that at 60% than a CC spell.


This is a good observation I feel. Since most (all?) crowd control spells are finite resources with typically a mediocre chance of success I personally gravitate towards cantrips and direct damage instead for most encounters, and save my CCs for "later" but I'm inevitably disappointed by my CCs when I finally use them.

Joined: Aug 2014
T
member
Offline
member
T
Joined: Aug 2014
We usually had that Fighter with 2-hander with all the feats that boosted hitchance, our group relied on him to have a near 100% chance to hit. As the streamlining of classes right now (dont know how it is in 5.0, but in BG3 atleast), it doesnt feel like there is any way to become a Excellent Blademaster or whatever. Just level up, chose MAYBE 1 thing, "Accept". Congratulations, you're now level 2 but you dont feel like you're any stronger.

Joined: Sep 2020
addict
Offline
addict
Joined: Sep 2020
I also agree on the point of low success chance with spells, but that's because you only get limited spell slots.

Only having a 60% chance to land a spell seems ridiculously low and makes it not worth casting anything but magic missile and cantrips.

Joined: Oct 2020
T
apprentice
Offline
apprentice
T
Joined: Oct 2020
I dont think the RNG is flawed here. I normally sense that if it would be skewed big way.

But in computer terms RNG doesnt exist. Its often generated by an algorithm using the computers clock.

It wouldnt be damaging if the Larian showed the algorithm doing the RNG and give us access to it so we can RNG thousands of times and generate spread sheets.

Sometimes there are little errors like for example rounding wrong.

to preserve symmetries - 2,5 rounded to 1 significance is for example 2 - not 3.

"Rounding to the nearest integer
Rounding a number x to the nearest integer requires some tie-breaking rule for those cases when x is exactly half-way between two integers — that is, when the fraction part of x is exactly 0.5.

If it were not for the 0.5 fractional parts, the round-off errors introduced by the round to nearest method would be symmetric: for every fraction that gets rounded down (such as 0.268), there is a complementary fraction (namely, 0.732) that gets rounded up by the same amount.

When rounding a large set of fixed-point numbers with uniformly distributed fractional parts, the rounding errors by all values, with the omission of those having 0.5 fractional part, would statistically compensate each other. This means that the expected (average) value of the rounded numbers is equal to the expected value of the original numbers when we remove numbers with fractional part 0.5 from the set.

In practice, floating-point numbers are typically used, which have even more computational nuances because they are not equally spaced."

https://en.wikipedia.org/wiki/Rounding

children often learn wrong rounding in school


Last edited by Tav3245234325325; 19/10/20 12:59 PM.
Joined: Jul 2014
enthusiast
Offline
enthusiast
Joined: Jul 2014
Originally Posted by Tav3245234325325
I dont think the RNG is flawed here. I normally sense that if it would be skewed big way.

But in computer terms RNG doesnt exist. Its often generated by an algorithm using the computers clock.

It wouldnt be damaging if the Larian showed the algorithm doing the RNG and give us access to it so we can RNG thousands of times and generate spread sheets.

Sometimes there are little errors like for example rounding wrong.

to preserve symmetries - 2,5 rounded to 1 significance is for example 2 - not 3.

"Rounding to the nearest integer
Rounding a number x to the nearest integer requires some tie-breaking rule for those cases when x is exactly half-way between two integers — that is, when the fraction part of x is exactly 0.5.

If it were not for the 0.5 fractional parts, the round-off errors introduced by the round to nearest method would be symmetric: for every fraction that gets rounded down (such as 0.268), there is a complementary fraction (namely, 0.732) that gets rounded up by the same amount.

When rounding a large set of fixed-point numbers with uniformly distributed fractional parts, the rounding errors by all values, with the omission of those having 0.5 fractional part, would statistically compensate each other. This means that the expected (average) value of the rounded numbers is equal to the expected value of the original numbers when we remove numbers with fractional part 0.5 from the set.

In practice, floating-point numbers are typically used, which have even more computational nuances because they are not equally spaced."

https://en.wikipedia.org/wiki/Rounding

children often learn wrong rounding in school



Frankly, this puts a smile on my face.

I can assure you Larian does not recreate a bicycle, they are also not incompetent and don't know how to properly generate random value in their games. It's not some indie studio getting its feet wet with first game, they don't need random forum poster teaching them how naive random number generators, like those based on internal clock work.

It's ok, you can let them handle this.

Page 2 of 13 1 2 3 4 12 13

Link Copied to Clipboard
Powered by UBB.threads™ PHP Forum Software 7.7.5