Putting the 'role' back in role-playing games since 2002.
Donate to Codex
Good Old Games
  • Welcome to rpgcodex.net, a site dedicated to discussing computer based role-playing games in a free and open fashion. We're less strict than other forums, but please refer to the rules.

    "This message is awaiting moderator approval": All new users must pass through our moderation queue before they will be able to post normally. Until your account has "passed" your posts will only be visible to yourself (and moderators) until they are approved. Give us a week to get around to approving / deleting / ignoring your mundane opinion on crap before hassling us about it. Once you have passed the moderation period (think of it as a test), you will be able to post normally, just like all the other retards.

Baldur's Gate Missing isn't fun or how I failed statistics at high school and blamed D&D for it.

Roguey

Codex Staff
Staff Member
Sawyerite
Joined
May 29, 2010
Messages
35,656
i get why randomization is needed, but at the same time:

NWN2 did the smart thing and made it so that any skill check done out of combat would be a natural 20 (cause let's face it you're just going to keep trying anyway). :M
 

Kaivokz

Arcane
Joined
Feb 10, 2015
Messages
1,499
This was exactly my thought when I read VD's argument example. From what I remember from high school, every coin toss taken individually has the same chance of landing on either side. It's a number of simultaneous coin tosses that have a progressively smaller chance of all landing on the same side. So 31 is just as probable as any other number from 1 to 100.

Assuming the coin is perfectly balanced every flip in the succession will be 50/50. If you flip 150 heads in a row, the 151st flip still has a 50% chance of landing heads. Further, any specific sequence of coin flips will have the same odds of happening (eg HHHH, HTHT, HHHT). It’s just our psychology that makes HHHH feel “more unlikely” than HHHT. (Often called the gambler’s fallacy.)
 

Shadenuat

Arcane
Joined
Dec 9, 2011
Messages
11,955
Location
Russia
I can say that modern gamers poorly react to abstraction by default; it's because games went into point and shoot more and more instead of allowing players freedom in playing around with mechanics behind the game.
 

Neanderthal

Arcane
Joined
Jul 7, 2015
Messages
3,626
Location
Granbretan
This was exactly my thought when I read VD's argument example. From what I remember from high school, every coin toss taken individually has the same chance of landing on either side. It's a number of simultaneous coin tosses that have a progressively smaller chance of all landing on the same side. So 31 is just as probable as any other number from 1 to 100.

Assuming the coin is perfectly balanced every flip in the succession will be 50/50. If you flip 150 heads in a row, the 151st flip still has a 50% chance of landing heads. Further, any specific sequence of coin flips will have the same odds of happening (eg HHHH, HTHT, HHHT). It’s just our psychology that makes HHHH feel “more unlikely” than HHHT. (Often called the gambler’s fallacy.)

But suppose one day the coin lands on its edge.
 

Kaivokz

Arcane
Joined
Feb 10, 2015
Messages
1,499
This was exactly my thought when I read VD's argument example. From what I remember from high school, every coin toss taken individually has the same chance of landing on either side. It's a number of simultaneous coin tosses that have a progressively smaller chance of all landing on the same side. So 31 is just as probable as any other number from 1 to 100.

Assuming the coin is perfectly balanced every flip in the succession will be 50/50. If you flip 150 heads in a row, the 151st flip still has a 50% chance of landing heads. Further, any specific sequence of coin flips will have the same odds of happening (eg HHHH, HTHT, HHHT). It’s just our psychology that makes HHHH feel “more unlikely” than HHHT. (Often called the gambler’s fallacy.)

But suppose one day the coin lands on its edge.
When a coin lands on its edge all probabilistic models are immediately nullified and physical events cease to be ordered or predictable.
 

Mr. Hiver

Dumbfuck!
Dumbfuck
Joined
May 8, 2018
Messages
705
Having coins flip 50/50 is all fine and good, but thats not what skills do in RPGs. Its not a 50/50 chance when you are at 70, 80 and higher CtH.

Im of a mind outright misses should be drastically reduced as skills get higher and all such unsuccessful attempts should be changed into defense of the enemy being successful.
So instead of you missing - the enemy defends with visual feedback clarifying that. Same for you in opposite situation.

Distance should be a great modifier to this all the time, increasing chance to hit at low levels when closer, slightly decreasing it when larger even on high levels, but the problem is that most often RPGs use very constrained small space for combat.
Yet even so, most "misses" at high levels should become enemy or yours successful defense.

So, you have 20% in small guns or knives at the beginning, in a fight you run toward enemy to get close - and that makes you and the enemy chance to hit become very high if you get very close.
You can hit more easily, but so can the enemy - and then it becomes a matter of who can defend, dodge or block or whatever better.
 

Neanderthal

Arcane
Joined
Jul 7, 2015
Messages
3,626
Location
Granbretan
This was exactly my thought when I read VD's argument example. From what I remember from high school, every coin toss taken individually has the same chance of landing on either side. It's a number of simultaneous coin tosses that have a progressively smaller chance of all landing on the same side. So 31 is just as probable as any other number from 1 to 100.

Assuming the coin is perfectly balanced every flip in the succession will be 50/50. If you flip 150 heads in a row, the 151st flip still has a 50% chance of landing heads. Further, any specific sequence of coin flips will have the same odds of happening (eg HHHH, HTHT, HHHT). It’s just our psychology that makes HHHH feel “more unlikely” than HHHT. (Often called the gambler’s fallacy.)

But suppose one day the coin lands on its edge.
When a coin lands on its edge all probabilistic models are immediately nullified and physical events cease to be ordered or predictable.

One can only match move by move the whims of probability and thus defy the tyrannous stars.
 

HeatEXTEND

Prophet
Patron
Joined
Feb 12, 2017
Messages
3,928
Location
Nedderlent
Assuming the coin is perfectly balanced every flip in the succession will be 50/50. If you flip 150 heads in a row, the 151st flip still has a 50% chance of landing heads. Further, any specific sequence of coin flips will have the same odds of happening (eg HHHH, HTHT, HHHT). It’s just our psychology that makes HHHH feel “more unlikely” than HHHT. (Often called the gambler’s fallacy.)
Utterly indulgent statistical wankery.
 

Kaivokz

Arcane
Joined
Feb 10, 2015
Messages
1,499
Assuming the coin is perfectly balanced every flip in the succession will be 50/50. If you flip 150 heads in a row, the 151st flip still has a 50% chance of landing heads. Further, any specific sequence of coin flips will have the same odds of happening (eg HHHH, HTHT, HHHT). It’s just our psychology that makes HHHH feel “more unlikely” than HHHT. (Often called the gambler’s fallacy.)
Utterly indulgent statistical wankery.
Not really, especially the second part makes sense if you think about it. If you take any specific sequence of coin flips—or dice rolls, or anything with a discrete uniform distribution—all of those are equally likely to happen. It’s obvious that you’re just as likely to roll all 6’s as 1’s, but less obvious that you’re equally likely to roll 1 3 5 6 2 4 6 as 6 6 6 6 6 6 6. It’s interesting, isn’t it?
 

Mr. Hiver

Dumbfuck!
Dumbfuck
Joined
May 8, 2018
Messages
705
Only if you count probabilities of each toss separately. If you count probabilities for 6666666 group toss, its much lower then the other more random variations for such groups. And on repeated throws its very improbable you will get that same sequence versus other more random variations.

Thats different then any specific sequence having the same odds of happening.

Anyway, the 50/50 odds are irrelevant for RPG gameplay since we never get such a clear distinction. And its specially nonsensical to use that as the only relevant measure of complex tasks.
 

Kaivokz

Arcane
Joined
Feb 10, 2015
Messages
1,499
Only if you count probabilities of each toss separately. If you count probabilities for 6666666 group toss, its much lower then the other more random variations for such groups. And on repeated throws its very improbable you will get that same sequence versus other more random variations.

Thats different then any specific sequence having the same odds of happening.

Anyway, the 50/50 odds are irrelevant for RPG gameplay since we never get such a clear distinction. And its specially nonsensical to use that as the only relevant measure of complex tasks.
Only if you count the other sequences as a group and exclude 66666 because it seems distinct to you. What about 12345? 112211? 54321? 111222333444555? Etc. Are these “more random” variations than 66666?

I was only responding to @AwesomeButton’s statement that a coin landing on heads repeatedly is less likely than not, which is true... but it is true of any one possible sequence, including the one where exactly half heads and half tails are flipped. I wasn’t suggesting we use discrete uniform distributions to model any and all situations (though dice are (ideally) discrete uniform distributions which we can fudge in various ways (assigning certain outcomes to ranges of values)).

Anyway, what you said highlights my point: the reason we compare “ordered” results with “random” (eg seemingly disordered) results is because of our psychology and perspective, not because 66666 is less random than 15434 or 34552.
 

Mr. Hiver

Dumbfuck!
Dumbfuck
Joined
May 8, 2018
Messages
705
Yes, those others are more random variations then 6666666. Because this is a specific sequence of repeating same numbers. Since the pool contains 10 different numbers, getting specifically that one is such a repeating pattern is less probable then other distributions. It must be because it excludes all other numbers and combinations which are greater in number. In other words, there is that single specific sequence and a bazzillion of all possible other combinations.
Therefore it is less probable to get that specific sequence then any of the others.

And then its even more improbable - rare to get that same sequence again on repeated tries, rather then any of that great variety of other combinations.


I wasn’t suggesting we use discrete uniform distributions to model any and all situations
I know, i didnt say you were, but it always comes up in discussions about hit chances and seemingly weird results we sometimes get in games with dice rolls.
 

Kaivokz

Arcane
Joined
Feb 10, 2015
Messages
1,499
Getting a repeating number is not less probable than any other distribution.

Rolling a six sided die ten times gives us ~60 million possible permutations. Each of those is unique and has the same chance of happening. The probability of rolling the sequence "6253462341" is roughly 1/60,000,000 which is equal to the probability of rolling the sequence "6666666666."

However, if you treat "1234..." as equivalent to "2134..." then you are changing the game. It's like writing 3 over the 2, 4, and 5, and then saying that 3 is more likely to be rolled than 1 or 6. It is true than the range of 2-5 is more likely to be rolled than 6, but that's because you're comparing a range of outcomes to an individual outcome in a discrete uniform distribution.

If you want to compare "distributions only containing the same number" with "distributions that do not only contain the same number" that is a human bias; you could also compare "the following six distributions: 12345, 11345, 11145, 11115, 11116, 11111" to "distributions that are not the preceding six distributions" and then make the same argument that 11116 is less random than other distributions because you've put it into this group which contains 6 out of ~60 million possibilities.
 

Mr. Hiver

Dumbfuck!
Dumbfuck
Joined
May 8, 2018
Messages
705
If you want to compare "distributions only containing the same number" with "distributions that do not only contain the same number" that is a human bias;
How the F is that human bias? Those numbers exist as such with or without me or my bias. A sequence of all same numbers IS objectively different and singularly specific then mutlitude of combinations that include the other numbers in that group. Its not my imagination or bias.

Getting a repeating number is not less probable than any other distribution.
it is for reason i explained above.

However, if you treat "1234..." as equivalent to "2134..." then you...
False equivalence fallacy. I was very specific in my example, which IS NOT the same as that example.
 

Kaivokz

Arcane
Joined
Feb 10, 2015
Messages
1,499
What you're saying is getting the same result repeated is less likely than all other possibilities, not any other possibility.

However, people often assume that because getting a repeated result is less likely than all other possibilities, it is therefore less likely than any other possibility.
 

anvi

Prophet
Village Idiot
Joined
Oct 12, 2016
Messages
7,530
Location
Kelethin
I never cared about missing. And RNG is fine for swings and misses, RNG is not so fun in loot and other ways. But if you are losing a fight because you got unlucky with RNG, you are weak. Gitgud. Then you win and don't even think about RNG.
 

Kaivokz

Arcane
Joined
Feb 10, 2015
Messages
1,499
Mr. Hiver look at this:

(1,1) (1,2) (1,3) (1,4) (1,5) (1,6)
(2,1) (2,2) (2,3) (2,4) (2,5) (2,6)
(3,1) (3,2) (3,3) (3,4) (3,5) (3,6)
(4,1) (4,2) (4,3) (4,4) (4,5) (4,6)
(5,1) (5,2) (5,3) (5,4) (5,5) (5,6)
(6,1) (6,2) (6,3) (6,4) (6,5) (6,6)

I've put in bold the case that you are interested in. But why is it special? What about the cases defined by (x, 1), (x, 2), etc.?
Any row or column forms a similar set with an ordered pattern. We can find many patterns in a diagram like this. What makes that diagonal row special?
 

HeatEXTEND

Prophet
Patron
Joined
Feb 12, 2017
Messages
3,928
Location
Nedderlent
Yes, those others are more random variations then 6666666.
The point is every single sequence of numbers into literal infinity has the exact same chance at comming up than any single other number in the (infinite) group. It's a never-ending single elimination bracket between every number until the end of time, in other words
utterly indulgent statistical wankery
:negative:
 

Mr. Hiver

Dumbfuck!
Dumbfuck
Joined
May 8, 2018
Messages
705
Oh yeah, infinities. Just love those. My favorites.

Kaivokz
Ive explained the exact example i am considering. Getting seven sixes (or any other specific number) repeating from a pool of ten numbers by complete random process. Your example may be "similar" but they are not the same.
ill say it once again. If you choose seven numbers from a pool of ten different numbers - by completely random selection - chances to get a combination of just one number repeating seven times must be lower then all other possible combinations.
IF NOT it must mean that the process is somehow skewed in favor of that particular pattern.

I dont have anything else to add, so best to just agree to disagree. Predefined columns of two numbers or any other different example simply does not apply here.
 

Garbage

Learned
Joined
Jun 10, 2019
Messages
121
Location
nearby dumpster
i get why randomization is needed, but at the same time:

NWN2 did the smart thing and made it so that any skill check done out of combat would be a natural 20 (cause let's face it you're just going to keep trying anyway). :M

I don't think that's so good, in tabletop gaming failure can be as fun as success and put you in amusing/tense situations you would have never gotten yourself into otherwise.
That needs to translate perfectly into games, the player shouldn't be immediately aware if they succeeded or failed a certain skill check but instead want to see how the situation plays out regardless.

To me, something like that would feel the same as savescumming so your favourite characters don't ever die in Valkyria for example. Half the tension comes from an encounter where you screwed up and you need to rush to save your favourite characters before they're killed off for good... And the more memorable moments of your campaign when you lose them for good.
 

Kaivokz

Arcane
Joined
Feb 10, 2015
Messages
1,499
Numbers in columns is just a way to represent what you're talking about. You can do the same thing with 10 numbers but the chart would be huge.

If you choose seven numbers from a pool of ten different numbers - by completely random selection - chances to get a combination of just one number repeating seven times must be lower then all other possible combinations.
Of course, because you're comparing the chance of getting all 7's to the chance of getting literally anything else... except that's the exact same chance as getting six 7's in a row and then one 6 compared to getting literally anything else.

You're saying "combination" which, in formal language, refers to probabilities when order does not matter (55566 = 56565). But if that's what you mean, then as I said, it's like writing 3 on all sides of the die except 1 and 6 and then saying 3 comes up more often on six sided dice than 1's or 6's.
 

As an Amazon Associate, rpgcodex.net earns from qualifying purchases.
Back
Top Bottom