D&D 5E Point Buy vs Rolling for Stats

[Tony Vargas]

And in 1e, yet people still complain about wizards ruling the roost.

It's just the sheer versatility of prepped casting. What put 'em in Tier 1 in 3e, too.

Maybe I'm just biased. I grew up in a small town (population 450) and I doubt there were 2 people in the county, much less the town with genius level IQs. I mean, there was one, but I don't live there any more.

Yeah, the curve is not equally appropriate to all stats, 'realistically,' it's a simplification/abstraction to try to make stats roughly equivalent in value, the way being able to arrange them (or even using the same method to generate them) implies they should be.

I vaguely remember RQII rolling some stats differently, like humans rolled SIZ on 2d6+6, an instance I remember because I had an RQI SIZ-4 human character.


The 3d6 bell curve doubling as both chargen method and population statistics has a sort of elegance, though.

 

[Lanefan]

Point-buy very clearly /does/ let players play the concept they want, assuming said concept comes to them prior to chargen, that is. They may have to "compromise" it slightly so that /everyone/ at the table gets to do so.

So, not perhaps exactly what they want, but something sort-of close. The same can be said for both roll-and-rearrange and fixed-array. The only method that has a real chance of messing up a concept is roll-in-order.

In essence, it can't be used to 'play exactly what you want,' because what another player wants may completely undermine it. Point-buy, OTOH, allows you /all/ to play what you want.

This one doesn't make sense. How does stat generation method have anything to do with players competing to play the same 'role' (or class) in the party?

D&D isn't real life, it's a game. That line of argument is a profound category error. Don't bother going there again.

I will go there every single time.

If you're intent is - as mine is - to try and reflect a real-ish game world population with people in it who are not all the same, that variability will also extend to played characters. Rolling - with or without rearranging - reflects this randomness nicely, while at the same time reasonably and quite elegantly allowing the PCs to (on average) be a cut above by giving them 4d6k3 or even 5d6k3 instead of the general population's 3d6.

Meanwhile rearranging allows one's rolls to suit one's concept - except in very rare circumstances - which to me largely puts the "I can't play my character concept" argument to bed.

Side note: something that only just occurred to me while typing this (though maybe old news to everyone else): if 3d6 gives what seems like too loose a bell curve for either or both of the PCs and the general population, a tighter - i.e. much lower chance of extremes - 3-18 bell curve can be achieved via rolling 5d4-2. Maybe this would work better for rolling up commoners?

Lanefan

 

[Tony Vargas]

So, not perhaps exactly what they want, but something sort-of close.

Something as close as possible that won't mess with the concepts of the other players nor the DM's campaign, yes.

The same can be said for both roll-and-rearrange and fixed-array.

To a lesser degree, yes, and I thought I'd been fairly clear in that. Arrays can let you build the character you want, as long as it isn't more extreme or more 'all around' than that array. Roll-and-arrange methods give you a similar level of customization, you can choose where to but high and low stats, but not the high-low spread.

The only method that has a real chance of messing up a concept is roll-in-order.

It also has the best chance of inspiring you to build to a concept you'd've never thought of otherwise.

This one doesn't make sense. How does stat generation method have anything to do with players competing to play the same 'role' (or class) in the party?

With a random (or freeform) method, if two players do go for concepts that require similar stat priorities, one can be flat-out better at that 'role' (or class) than the other, undermining his concept. With an array or point-buy, they'll have to settle for ties, or find some other nuances to differentiate themselves, there's no risk that one will just overshadow the other.

I will go there every single time.

Then logic will not go with you.

If you're intent is - as mine is - to try and reflect a real-ish game world population with people in it who are not all the same, that variability will also extend to played characters...
allowing the PCs to (on average) be a cut above by giving them 4d6k3 or even 5d6k3 instead of the general population's 3d6.

I don't see that as quite the same variability, but OK. (And, no, you didn't "go there.")

Rolling - with or without rearranging - reflects this randomness nicely

Rolling is randomness, yes, but there's no difference between the world painted by randomness, and one where the exact same results have been arrived at arbitrarily. For the PCs, a much smaller sample size, OTOH, random generation can give very different results.

something that only just occurred to me while typing this (though maybe old news to everyone else): if 3d6 gives what seems like too loose a bell curve for either or both of the PCs and the general population, a tighter - i.e. much lower chance of extremes - 3-18 bell curve can be achieved via rolling 5d4-2. Maybe this would work better for rolling up commoners?

Yep, that'd compress it a bit. Don't know why I'd never thought of it - I'd seen 4d4, and even 6d3, for that purpose before.

 

[Lanefan]

It also has the best chance of inspiring you to build to a concept you'd've never thought of otherwise.

Agreed.

With a random (or freeform) method, if two players do go for concepts that require similar stat priorities, one can be flat-out better at that 'role' (or class) than the other, undermining his concept. With an array or point-buy, they'll have to settle for ties, or find some other nuances to differentiate themselves, there's no risk that one will just overshadow the other.

Oh, there's still that risk - the overshadowing just won't be stats-based, is all.

And sometimes if one finds oneself playing second fiddle, particularly in a melee class, one just has to wait until the first fiddle goes in first once too often and gets itself killed...

Lanefan

 

[Satyrn]

It also has the best chance of inspiring you to build to a concept you'd've never thought of otherwise.

Sure.

Though I've found that goal is better achieved by randomly choosing the race/class/subclass/background combo.

 

[Tony Vargas]

Oh, there's still that risk - the overshadowing just won't be stats-based, is all.

That's one less risk, then.

Sure.

Though I've found that goal is better achieved by randomly choosing the race/class/subclass/background combo.

"Uh, I got a Halfling Totem Barbarian Sage..."
"Cool, my half-orc Life Cleric charlatan is disguised as a Sage, we can be buds!"

 

[Satyrn]

"Uh, I got a Halfling Totem Barbarian Sage..."
"Cool, my half-orc Life Cleric charlatan is disguised as a Sage, we can be buds!"

See, it's working!

 

[TwoSix]

"Uh, I got a Halfling Totem Barbarian Sage..."
"Cool, my half-orc Life Cleric charlatan is disguised as a Sage, we can be buds!"

Together, they fight crime.

 

[Maxperson]

At least most of your numbers were above 10. I've been in a game where someone rolled a character with 1 14, a 10 and all other numbers all below 10, with no rerolls.

Depending on the edition, that's actually a house rule. In 1e the rules expected multiple rolls to get a suitable character. 2e didn't to my knowledge. 3e had a minimum positive bonus number that the numbers you are stating violate. Not sure about 4e. 5e doesn't seem to have it either.

 

[Oofta]

Depending on the edition, that's actually a house rule. In 1e the rules expected multiple rolls to get a suitable character. 2e didn't to my knowledge. 3e had a minimum positive bonus number that the numbers you are stating violate. Not sure about 4e. 5e doesn't seem to have it either.

I don't know if the DM was just ignoring the rules or not. But in 5E there is no minimum, so you could roll everything well below a 10 and it would be a "valid" character.