This is a site for discussing roleplaying games. Have fun doing so, but there is one major rule: do not discuss political issues that aren't directly and uniquely related to the subject of the thread and about gaming. While this site is dedicated to free speech, the following will not be tolerated: devolving a thread into unrelated political discussion, sockpuppeting (using multiple and/or bogus accounts), disrupting topics without contributing to them, and posting images that could get someone fired in the workplace (an external link is OK, but clearly mark it as Not Safe For Work, or NSFW). If you receive a warning, please take it seriously and either move on to another topic or steer the discussion back to its original RPG-related theme.
NOTICE: Some online security services are reporting that information for a limited number of users from this site is for sale on the "dark web." As of right now, there is no direct evidence of this, but change your password just to be safe.

Author Topic: Is RPG Optimization Psychosis?  (Read 7001 times)

Mishihari

  • Hero Member
  • *****
  • M
  • Posts: 648
Re: Is RPG Optimization Psychosis?
« Reply #135 on: January 04, 2021, 12:22:53 AM »
The fact is that most modern systems seem to be much higher than that... 50%, 60%, even 75% odds are not uncommon in modern systems (my system, for example uses a baseline for combat of about 60% or about three hits during a five round combat... this was also the baseline for 4E).


I was in a discussion in ENWorld some years ago, when it was mentioned that WoTC had done some actual research and concluded that a 70% probability to-hit made the game the most fun. 

TJS

  • Hero Member
  • *****
  • T
  • Posts: 796
Re: Is RPG Optimization Psychosis?
« Reply #136 on: January 04, 2021, 01:01:21 AM »
I always consider that. But there are three points against that. The first is I've looked at what I said, and it's seems pretty clear. That's not conclusive, because I could have missed something. But for me to reconsider that, you'd have to demonstrate where I miscommunicated, not just make a vague claim. Secondly, they didn't misunderstand my arguments in the same way. You made one argument, and they made other arguments. And most importantly is the third: People really don't understand probability well, and it can be challenging to break down their preconceptions. That's the reason why I posted what I posted. I expected some pushback.

I think the problem is more possibly where you are placing emphasis rather than the points themselves.

Yes, a +1 is a big deal at the outliers. But its not a big deal as you drift away from those edges.

But to an optimizer that +1 is a bonus period and they NEEED IT!

Or when I was a playtester on AQW. We had some nuts flipping out over how "unballanced" one class was over the others because its attack was .01 second faster. Even after we pointed out No. It didnt. Those were vauguarities in personal or serverside processing. But no-no. This is game breaking! And on the flip side others were declaring a different class THE BEST WHY TAKE ANY OTHER??? because they believed is was .01 sec faster. And of course it was not either.

And all this and so much worse behaviors been around a long time in tabletop gaming. As said. Some perceive it as an "I WIN!" button.

But at the end of the day to the more obsessive optimizer a +1 is a bonus across the board and thats all that matters. Better if can pump it up to a +2, 3, 4, etc.
People always put the emphasis on what is most easily quantifiable.

This is why it feels like a big deal in modern D&D if you don't max your attack score at character creation.  It's less that it's a big deal and more that you will always be obviously and in an easily quantifiable way behind someone who did.

See also focus on DPR etc - relatively easy to calculate.

If you're choice is between "overhead strike" and "underhand strike" then they'd better have equivalent effect because they're so obviously comparable.  If you're choice is between "combat badass" or "can turn into a bat" then that's much less comparable.

This is also why reskinning is dangerous - once you divide everything into fluff and crunch and decide the former is interchangeable then everything becomes much more comparable.
« Last Edit: January 04, 2021, 01:04:28 AM by TJS »

Kyle Aaron

  • high-minded hack
  • Hero Member
  • *****
  • Posts: 9320
  • high-minded hack
    • The Viking Hat GM
Re: Is RPG Optimization Psychosis?
« Reply #137 on: January 04, 2021, 01:39:56 AM »
I was in a discussion in ENWorld some years ago, when it was mentioned that WoTC had done some actual research and concluded that a 70% probability to-hit made the game the most fun.
People are going to hate Conflict, then. :D
The Viking Hat GM
Conflict, the adventure game of modern warfare
Wastrel Wednesdays, livestream with Dungeondelver

Chris24601

  • Hero Member
  • *****
  • C
  • Posts: 2338
Re: Is RPG Optimization Psychosis?
« Reply #138 on: January 04, 2021, 09:59:41 AM »
I was in a discussion in ENWorld some years ago, when it was mentioned that WoTC had done some actual research and concluded that a 70% probability to-hit made the game the most fun.
I’d buy those numbers; though I suspect the actual probability for most fun is probably exactly 66.66~% and that 70% was just the closest value using 1d20 above that.

There’s just something in human psychology about threes (basically it’s the smallest number that can show a pattern) so “two-out-of-three” seems like the ideal for that... you’ll succeed about twice as often as you fail overall so those feel like reasonable odds when your life isn’t on the line (your PC’s life doesn’t count). By contrast 1-in-3 feels like bad odds.

I found a 60% baseline to work precisely because it’s that... the baseline and there are a couple of easy ways (the most easy is to use a weapon or implement with the accurate property) to improve it from there.

Also, PCs in my system always have a path (subclass) specific minor action they can use that’s either automatic or otherwise greatly increases the odds of succeeding in that path’s role (the most basic being the Fighter’s striker path action, which gives them a second attack using an offhand weapon; including the butt of a polearm, a shield bash or an unarmed strike if wielding a larger weapon; which gives even a baseline PC a 90% chance of hitting at least once per round).

Regardless, 2-in-3 odds need way more than a +5% bonus for that to feel noticeable. I suspect that’s partly why 5e’s advantage is so popular as it’ll take that 2-in-3 and bump it up to the neighborhood of 4-in-5 or more.

Steven Mitchell

  • Hero Member
  • *****
  • S
  • Posts: 2934
Re: Is RPG Optimization Psychosis?
« Reply #139 on: January 04, 2021, 10:29:37 AM »
There is something to the two-thirds chance part.  Anywhere from about 60% to 70% will safely evoke the feeling in most games. 

That said, I think there is also an element of how much real time it takes to have some success.  A simple system with only 45% to 55% chance of success can feel satisfactory to many players if they can get a hit in about the same amount of time it takes in a slower handling system with higher chances.  Plus, some people get satisfaction at being "shot at without results" (nothing so exhilarating).  The psychology of how successful a hit feels (e.g. damage relative to putting the opponent down) enters into it as well. 

Pat

  • Hero Member
  • *****
  • P
  • Posts: 5213
  • Rats do 0 damage
Re: Is RPG Optimization Psychosis?
« Reply #140 on: January 04, 2021, 10:30:50 AM »
I always consider that. But there are three points against that. The first is I've looked at what I said, and it's seems pretty clear. That's not conclusive, because I could have missed something. But for me to reconsider that, you'd have to demonstrate where I miscommunicated, not just make a vague claim. Secondly, they didn't misunderstand my arguments in the same way. You made one argument, and they made other arguments. And most importantly is the third: People really don't understand probability well, and it can be challenging to break down their preconceptions. That's the reason why I posted what I posted. I expected some pushback.

I think the problem is more possibly where you are placing emphasis rather than the points themselves.

Yes, a +1 is a big deal at the outliers. But its not a big deal as you drift away from those edges.

But to an optimizer that +1 is a bonus period and they NEEED IT!

Or when I was a playtester on AQW. We had some nuts flipping out over how "unballanced" one class was over the others because its attack was .01 second faster. Even after we pointed out No. It didnt. Those were vauguarities in personal or serverside processing. But no-no. This is game breaking! And on the flip side others were declaring a different class THE BEST WHY TAKE ANY OTHER??? because they believed is was .01 sec faster. And of course it was not either.

And all this and so much worse behaviors been around a long time in tabletop gaming. As said. Some perceive it as an "I WIN!" button.

But at the end of the day to the more obsessive optimizer a +1 is a bonus across the board and thats all that matters. Better if can pump it up to a +2, 3, 4, etc.
I didn't place the emphasis on that, I just used it as an example. Which I expanded upon in later posts, because the people initially responding to my post were misunderstanding the probabilities, and spelling out how 2/1 = 100% by listing the 3 numbers is easier than using 11/10 = 10%, and listing 21 numbers. Making things more confusing for the people I actually responding to because other people might come along and assume the example I used for illustration is the totality of what I'm saying doesn't seem like a good approach. Nor is repeating all of what I said in my first post, in every post. This really seems like a case where you read it superficially and drew the wrong conclusions, and where there wasn't a way for me to communicate it better.

You pretty much nailed my read of their argument... Pat was putting so much emphasis on the outlier effects while ignoring that almost NO modern game (i.e. the types where you're most likely to see "builds" and planned optimization) build their system to have lots of outliers because, frankly, extremely low/high probabilities of success aren't actually that interesting.
Same as above, plus note the 11/10 example I just used above is the middle of the range, and it still amounts to twice the +1 = +5%. That's good illustration that +1 = +5% is the wrong way of thinking.

Also, I think you're wrong about how frequently characters succeed in different games. It's not toward the middle. The study by WotC that Mishihari mentioned a couple posts back was in the back of my head when I wrote my first post, and that matches with my experience with a wide variety of games. Competent characters, which generally include the PCs, don't typically have a 20% chance of success, or even a 50%. Most tend to succeed in the 60-90% range. There's psychological reasons for this: People play games to have fun, and little victories like hitting fairly frequently become a system of positive feedback. Old school D&D tends to be a little bit of an outlier, because things like to hit rolls and saves can vary more, based on level. In low level games, the chance to hit or save can be anywhere from 5% to 50%. The reward here is a little different, with low level games feeling very hard or whiffy, but that gives a feeling of accomplishment to the players as characters advance and become more reliably successful at individual rolls. It's more of a delayed reward, which supports the themes of a game that focuses on leveling up rather than starting characters as full-fledged heroes.

Let's look at the full range, on the d20 spectrum, using some more broadly illustrative examples. (Incidentally, this is in line with where I hoped the discussion would go after my initial post, but the conversation got sidetracked a bit because people were misunderstanding how probability works, so I ended up repeating my example in a couple different ways, leading to the confusion you felt.)

Succeed on a / What a +1 bonus means to your chance of success
20: +100% (doubles it, since you now succeed on a 19 or 20, not just a 20)
16+: +20%
11+: +10%
6+: +7%
2+: +5% (chance of success increases from 2-20 to 1-20, or 20/19)

If we think of this in the long term, if you only succeed on a 20, a +1 bonus means you succeed twice as often, over the course of a campaign (well, as long as the conditions hold). Even a +1 bonus to hit when your chance to hit is 50% (11+) means you do an extra 10% damage, over time. The +5% in the last line of the table (2+) is rounded down (from 5.26%), so even when your starting chance of success is high, a +1 bonus means more than 5%, which should clearly show that +1 = +5% is always the wrong way of thinking about it.

And note this is focused on success. If we focus on failure and penalties, which is important when it's about avoiding a negative results (e.g. saves), then this is what happens:

Succeed on a (fail on a) / What a -1 penalty means to your chance of failure
20 (19-): +5% (chance of failure increases from 2-20 to 1-20, or 20/19)
16+ (15-): +7%
11+ (10-): +10%
6+ (5-): +20%
2+ (1): +100% (doubles it, since you now fail on a 1 or 2, not just a 1)

The table flips. That means there's always a way of looking at it where a +1 bonus is equal to +10% or more, and outliers become twice as common.

Though if you play in a game where the PCs have a high chance of positive results, and a low chance of negative results (to hit and saves, for instance), those will tend to fall in the 5% to 10% range. As I mentioned above, old school D&D is a notable exception. Another is getting hit by a monster, in most games (even if the PCs have a high chance to hit, the monsters tend to have a high chance to hit as well, which is a negative results for the PCs).
« Last Edit: January 04, 2021, 10:41:26 AM by Pat »

Zalman

  • RPG Evangelist
  • Hero Member
  • *****
  • Posts: 763
Re: Is RPG Optimization Psychosis?
« Reply #141 on: January 04, 2021, 10:39:34 AM »
There is something to the two-thirds chance part.  Anywhere from about 60% to 70% will safely evoke the feeling in most games. 

Yes, not just as a "fun" indicator, but as a decision-maker at the table. Pat can harp on all he likes about "percentage increase in chance to hit", but players don't care about "percentage increase in chance to hit", only "percentage increase in chance to survive/win". A fighter that hits only on a 19-20 and gets a +1 only increases their chance of surviving that battle by 5%. That's the percentage that people care about (and rightfully so, it's the one that actually makes a difference.)
Old School? Back in my day we just called it "School."

Pat

  • Hero Member
  • *****
  • P
  • Posts: 5213
  • Rats do 0 damage
Re: Is RPG Optimization Psychosis?
« Reply #142 on: January 04, 2021, 10:46:01 AM »
A fighter that hits only on a 19-20 and gets a +1 only increases their chance of surviving that battle by 5%.
That is completely false. A fighter that hits on a 19 or 20 (2 chances in 20) and gets a +1 bonus to hit will now hit on a 18 or 20 (3 chances in 20).

That doesn't perfectly map to their "chance of surviving", because that's a complex situation dependent on many variables. But it does increase their expected damage output by a full 50%, which is a lot.

Chris24601

  • Hero Member
  • *****
  • C
  • Posts: 2338
Re: Is RPG Optimization Psychosis?
« Reply #143 on: January 04, 2021, 11:41:40 AM »
A fighter that hits only on a 19-20 and gets a +1 only increases their chance of surviving that battle by 5%.
That is completely false. A fighter that hits on a 19 or 20 (2 chances in 20) and gets a +1 bonus to hit will now hit on a 18 or 20 (3 chances in 20).

That doesn't perfectly map to their "chance of surviving", because that's a complex situation dependent on many variables. But it does increase their expected damage output by a full 50%, which is a lot.
Pat, I think you have to look at the replies here and realize that you are the one not approaching this like most people do. It may make perfectly logical sense to you, but other people aren't processing the odds in the same way you are.

Going from 2 average damage per round to 3 average damage per round may seem significant to you, but when the average damage per round with a 60% hit rate is 12, both 2 and 3 feel horribly sub-optimal.

Here's something I learned from all my playtesting for my system... You can be 100% right on the math, but if the people actually using the math don't process it the same way you do, then you're never going to convince them.

For example, when I was doing testing on flat-bonuses vs. advantage-style mechanics, one of my tests was offering a flat-bonus statistically better than even the re-roll mechanic could provide (a +6 bonus) and they STILL preferred the re-roll because the re-roll could save ANYTHING. Roll a 2 when you need a 10? +6 doesn't help you. Re-roll might.

Hat tip to the playtester who coined the term, but "save vs. failure" beat out even far statistically superior flat bonuses precisely because it was just that... a literal second chance and humans are just hard-wired to remember the times those pay off way more than the times they fail (you'd already failed, so having it continue to fail isn't notable... whereas a failure turning into a success is memorable).

From my testing, regardless of the statistical odds, the odds were ALWAYS perceived as better before the roll if you had a re-roll than a flat bonus unless the flat bonus actually took your odds to "sure thing" territory.

Another element from my testing... from my experience people almost NEVER judge on odds of success... they judge on odds of failure (i.e. risk). They don't see "I needed a 19-20, but now need an 18-20" as "I have 50% better odds of success" they see "I went from a 90% chance of failure to an 85% chance of failure."

Which, to bring this back to optimization, means that optimizers aren't chasing +1's to take improve their odds of success from 19-20 to 18-20... they're taking them to reduce the rate of failure... ideally to as close to "sure thing" as the system will allow.

As a practical example of this, back in Palladium days the default standard was to get at least a +3 to strike because a natural 1 always missed and a 1-4 result missed. With a +3 to strike anything that didn't auto-miss at least required the defender to make a parry or dodge roll in order to avoid a potential hit.

Zalman

  • RPG Evangelist
  • Hero Member
  • *****
  • Posts: 763
Re: Is RPG Optimization Psychosis?
« Reply #144 on: January 04, 2021, 12:51:11 PM »
But it does increase their expected damage output by a full 50%, which is a lot.

Here I disagree. A "50% increase" is "a lot" only if it's 50% of a significant number in the first place. No player I've ever encountered thought +1 was "a lot" on a d20 roll. At this point, the argument seems entirely disingenuous to me.
Old School? Back in my day we just called it "School."

Zalman

  • RPG Evangelist
  • Hero Member
  • *****
  • Posts: 763
Re: Is RPG Optimization Psychosis?
« Reply #145 on: January 04, 2021, 12:56:18 PM »
Another element from my testing... from my experience people almost NEVER judge on odds of success... they judge on odds of failure (i.e. risk). They don't see "I needed a 19-20, but now need an 18-20" as "I have 50% better odds of success" they see "I went from a 90% chance of failure to an 85% chance of failure."

Interesting. My own experience is that the difference is generally viewed in terms of the larger chance, whether that happens to be success or failure, probably since it's the number that is relevant to outcome, and thus player decision.
Old School? Back in my day we just called it "School."

Steven Mitchell

  • Hero Member
  • *****
  • S
  • Posts: 2934
Re: Is RPG Optimization Psychosis?
« Reply #146 on: January 04, 2021, 01:48:21 PM »
Another element from my testing... from my experience people almost NEVER judge on odds of success... they judge on odds of failure (i.e. risk). They don't see "I needed a 19-20, but now need an 18-20" as "I have 50% better odds of success" they see "I went from a 90% chance of failure to an 85% chance of failure."

Interesting. My own experience is that the difference is generally viewed in terms of the larger chance, whether that happens to be success or failure, probably since it's the number that is relevant to outcome, and thus player decision.

My experience is that players are all over the place, such that as a group their behavior defies description.  I can categorize some of the common bits, but even then there is a lot of variety and crossover.  To wit:

A. Player is all about the math.  (A.1 To the point of doing stupid stuff because the formula says so. A.2 Classic optimizer with modest tactics after the build.  A.3 Focused on winning on all levels. Etc.)

B. Player is all about the character.  (A.1 Unconsciously adapts the math via experience of what works mixed with good judgment. A.2 Repeatedly makes mistakes in evaluating odds due to gambler fallacy and other issues.  Etc.)

C. Player is orthogonal to both above, perhaps stirring up trouble or acting almost solely on an intuitive level or too many other ways to list.

Suppose a pre-gen character in a game with 3 attacks.  One very accurate, low damage.  Another is the opposite.  The middle one is a mix of accuracy and damage that is also the most mathematically efficient.  I've known different players that will always pick the accuracy or the high damage or the efficiency as long as all 3 options are available.  That's not even getting into the situational differences, limited resources providing the attacks, area effects vs number of targets, etc.  Not to mention players that will cycle between the 3 options with no appreciable pattern or even a logic that they can articulate.

Pat

  • Hero Member
  • *****
  • P
  • Posts: 5213
  • Rats do 0 damage
Re: Is RPG Optimization Psychosis?
« Reply #147 on: January 04, 2021, 02:54:06 PM »
But it does increase their expected damage output by a full 50%, which is a lot.

Here I disagree. A "50% increase" is "a lot" only if it's 50% of a significant number in the first place. No player I've ever encountered thought +1 was "a lot" on a d20 roll. At this point, the argument seems entirely disingenuous to me.
Pat, I think you have to look at the replies here and realize that you are the one not approaching this like most people do. It may make perfectly logical sense to you, but other people aren't processing the odds in the same way you are.
Zalman, that's because, and I'm not trying to be mean about it, you seem to have no understanding of probabilities. You stated that someone that hits on 19 to 20 and then gets a +1 bonus has a 5% greater chance to survive a battle. The math error is relatively minor compared to the more fundamental one: That the survival rate simply can't be calculated by looking at one +1 bonus. Combat, and survival, involves a wide range of variables, as well as a number of subjective decisions. That +5% is completely, utterly meaningless.

I'm not sure what point you're trying to make with all these replies and snide remarks. Because while the tone of your posts seems to be strongly opposed to what I'm saying, and you adopt elements from my posts like the +5%, you use those elements out of any context that makes sense, and never seem to address any of the points I've made.

Chris, you seem to have a better grasp of probabilities, but otherwise you're doing exactly the same thing. It's really bizarre.

If you're not interested in what I'm talking about, why reply? And if you'd like to talk about how the probabilities work in a wider context, or about other topics like how players react to the different ways of presenting the same mechanics, then why not just say so? I've literally never said anything about any of that, in this thread, so why are you pretending you're rebutting anything I've said? Those are actually topics I would have been interested in talking about, if the waters hadn't been poisoned like this.


Shasarak

  • Hero Member
  • *****
  • Posts: 3979
Re: Is RPG Optimization Psychosis?
« Reply #148 on: January 04, 2021, 03:10:17 PM »
Optimisers should beware of monsters that know this maxim: the nail that sticks out is the first to get the hammer.

Minimizers should beware of monsters that know this maxim: the limp guy is the first one left behind when shit gets real.  :P

You assume that Minimizers want their character to live.  And why would you when it is so easy to roll up a new one 3d6 down the line.
Who da Drow?  U da drow! - hedgehobbit

There will be poor always,
pathetically struggling,
look at the good things you've got! -  Jesus

Zalman

  • RPG Evangelist
  • Hero Member
  • *****
  • Posts: 763
Re: Is RPG Optimization Psychosis?
« Reply #149 on: January 04, 2021, 03:17:22 PM »
I'm not sure what point you're trying to make with all these replies and snide remarks. Because while the tone of your posts seems to be strongly opposed to what I'm saying, and you adopt elements from my posts like the +5%, you use those elements out of any context that makes sense, and never seem to address any of the points I've made.

Apologies, I don't mean to sound snide. I do disagree with your original statement, which I believe was that it's more "useful" to think of a +1 bonus as a "50% increase in the chance of success", and to thus consider it a "large" bonus.

I disagreed, and attempted to explain why to you, but you seem stuck in repeating mathematical truisms about comparisons that are, in my opinion, irrelevant to the context. As to the other points you brought up, like "survival is more complex than that," I find them entirely orthogonal to the conversation, so yes, I have nothing to say about them.

I understand probabilities just fine; I also understand relevance, which is something you don't seem to want to discuss.

In any case, I don't feel like you are engaging in conversation in good faith here, so I won't be continuing this line of argument with you.
Old School? Back in my day we just called it "School."