D&D General here's how to stop jealousy in between lucky players and unlucky ones

[ichabod]

if that's the solution why not just give them a point buy and do away with randomness?

Because it's not a binary choice between crappy point buy and 4d6 drop low, and some people like randomness. There's different ways to roll stats with less variance between players, and other ways to introduce randomness without rolling for stats directly.

 

[nevin]

There are still ways to introduce randomness, but one of the two desires must be either reduced or tweaked:

Every character gets an equivalent starting point.
No one can predict what their stats will be in advance.

Which makes sense. The first one is, quite literally, saying every character can predict, with reasonable accuracy, exactly what their stats will be. The second is, quite literally, saying that there very likely (but not necessarily always) will be significant differences.

The problem is, starting statistics serve two masters, and those masters are often at one another's throats. On the one hand, those starting values are permanent power which can heavily define what a character is capable of for the rest of the game. As @Zardnaar said, starting with much higher stats is equivalent to getting several feats for free at character creation. On the other hand, for a lot of players, those starting values are also used as sources of inspiration for unexpected characters and as personality/behavior markers. To have the same numbers, even if they are put in different spots, thus necessarily means "these characters have exactly the same personality," which is a huge no-no for most people.

I'd say about a third of the time, you cannot please both masters, and no one likes the choice that that induces. Unfortunately, the only actual fixes are unacceptable to at least a very large minority of the audience. Firstly, it is unacceptable to divorce power from these stats, so that they can simply serve as roleplay guides, because many (I would argue most!) folks want statistics that actually signify growth, doubly so in 5e where "growth" has been so thoroughly strangled. Secondly, it is unacceptable to divorce the personality/behavior markers from these stats, because folks who expect that have few to no alternatives and don't get any value out of "just make it up yourself"-type things (because they want to be surprised, and/or challenged, by the character they get, as opposed to what they see as a plastic, manufactured character someone "wants" to have.)

Unfortunately, unless and until someone can come up with a method that is (a) truly simple, (b) genuinely random, (c) not particularly predictable, and (d) guaranteed to not produce "unfair" results for one player vs another, this problem will continue to exist. Because the community as a whole wants two characteristics that cannot work together--even in principle. Like how 60% of people can want to reduce taxes, and 60% of people can want to increase services, and yet none of them need to be irrational to do this, because the (minimum) 20% overlap between the two can be made up of people who have nuanced opinions that get glossed into something irrational only when they are forced to join up with rigid political factions (e.g., "I want to increase taxes on the rich, which will pay for decreasing taxes on the poor AND pay for more services." Or, "I want to decrease spending on everything except the army and infrastructure projects, and thus decrease taxes for everyone." Neither of these positions is irrational if understood in isolation, but when part of a political environment, you can have major issues. This is one of the poorly-discussed issues with democracy, that the individual voters can be rational while the electorate can be collectively irrational.)

There is no way to have random and fair in the same method. If you want fair point buy for all characters let them put them where they want them and everyone is the same. it is however unfortunate that random generates excitement and some people would rather risk lower stats for the chance to get higher stats so for them this method is the most boring/unfun option.

 

[nevin]

Because it's not a binary choice between crappy point buy and 4d6 drop low, and some people like randomness. There's different ways to roll stats with less variance between players, and other ways to introduce randomness without rolling for stats directly.o

I get that but randomness is unfair by definition. I don't see the the people that like to gamble to have the chance at being better than everyone else being happy they got a handout from the lucky roller.

 

[ichabod]

There is no way to have random and fair in the same method.

I would have to disagree with this. We had a campaign where each player rolled one stat and the DM rolled two. Then each player had the same stats, but they were random, and not stats you would ever get with point buy. In my current house rules I use point buy (just beefed up from the PHB, because point buy in the PHB is worse than a typical rolled character). There are however, two random options, and players can use one or both. First, they can roll d% and get a random set of stats that could be bought with my point buy (there's only 97 possible ways to spend the points). Second, they can roll to randomly determine which values go to which stats. All of these methods are random and fair.

And again, you can use different dice mechanics for rolling stats that reduce between player variance (or increase it), so you can find the level of variance that is acceptable to your table. You can make the randomness more fair.

 

[EzekielRaiden]

Hey @EzekielRaiden did you use ChatGPT to publish your post ?
I say that here because ChatGPT is renowned for his science to hide truth using poker,
and
the order to assign stats scores should be left to players' choice
( for example the highest scorer provides a 10,12,13,15,17,18
and an unfortunate player will choose a 12,10,18,17,13,15 )

No? I have never used ChatGPT to write a post for me. Ever. Not even a part of a post (other than quoting from it to specifically and openly discuss what it has said/done, not to take credit for its work.)

 

[le Redoutable]

No? I have never used ChatGPT to write a post for me. Ever. Not even a part of a post (other than quoting from it to specifically and openly discuss what it has said/done, not to take credit for its work.)

Ok excuse me for using your post to point out the poker aspect of chatGPT

 

[EzekielRaiden]

I would have to disagree with this. We had a campaign where each player rolled one stat and the DM rolled two. Then each player had the same stats, but they were random, and not stats you would ever get with point buy.

But they aren't actually random then. Everyone is using (to copy a phrase I've seen often in this context) "carbon copy" stats, and thus not truly random. Once one person knows their stats, everyone knows their stats.

In my current house rules I use point buy (just beefed up from the PHB, because point buy in the PHB is worse than a typical rolled character). There are however, two random options, and players can use one or both. First, they can roll d% and get a random set of stats that could be bought with my point buy (there's only 97 possible ways to spend the points). Second, they can roll to randomly determine which values go to which stats. All of these methods are random and fair.

But are they? Getting 13/13/13/12/12/12 isn't even remotely as good as getting 15/14/13/12/10/8 or 15/14/14/10/10/8, or even 15/13/13/13/11/8 even though all four are 27 PB slates. Even with a +1-to-three race or standard human, many potential expenditures of 27 point buy points are crappy allocations.

And again, you can use different dice mechanics for rolling stats that reduce between player variance (or increase it), so you can find the level of variance that is acceptable to your table. You can make the randomness more fair.

And you do so precisely by removing randomness and increasing control. The more you make randomness "fair," the more you control the outcome. That is precisely the problem. People want an outcome that feels genuinely outside of their control, free to be any possible character that could theoretically be made, and yet also remains fair and does not grant undue benefit to some participants over others.

And those two desires conflict at a fundamental level. To ensure fairness IS to reduce randomness in this context.

 

[le Redoutable]

here a system ( based on character pts )

pts based on age

 

[ichabod]

But they aren't actually random then. Everyone is using (to copy a phrase I've seen often in this context) "carbon copy" stats, and thus not truly random. Once one person knows their stats, everyone knows their stats.

Oh, no, those are random. Those are truly random. You got them by rolling dice. You are just not being clear on the kind of randomness you are looking for. The within player variance (the spread in the possible values a single player could get) is exactly the same as each player rolling separately. It's just that the between player variance (the spread in values across the different players) is now zero.

But are they? Getting 13/13/13/12/12/12 isn't even remotely as good as getting 15/14/13/12/10/8 or 15/14/14/10/10/8, or even 15/13/13/13/11/8 even though all four are 27 PB slates. Even with a +1-to-three race or standard human, many potential expenditures of 27 point buy points are crappy allocations.

That's a fair argument, but that can be worked around. You can limit the possibilities to the non-crappy point buy allocations. You can do other things. Note that the sum of the median for six stats with 4d6 drop low is 73. So roll five stats, and the sixth is 73 minus the sum of the first five. Then you don't even have the carbon copy problem.

And you do so precisely by removing randomness and increasing control. The more you make randomness "fair," the more you control the outcome. That is precisely the problem. People want an outcome that feels genuinely outside of their control, free to be any possible character that could theoretically be made, and yet also remains fair and does not grant undue benefit to some participants over others.

And those two desires conflict at a fundamental level. To ensure fairness IS to reduce randomness in this context.

I would say that to ensure fairness is to reduce between player variance. I'm just arguing (a) that it's a spectrum and not a binary choice, (b) each table can choose where they want to be on that spectrum, and (c) that there can still be randomness involved at the fair end of the spectrum. The freedom to make any possible character is another issue, and I agree that it can only be done in a limited fashion while having randomness. But there is also a long tradition of RPGs not having that sort of freedom. I remember rolling badly of the life path tables in Cyberpunk 2020...

 

[AmandaBarbarian]

Is it common to try to avoid player differences in stats?