D&D 5E Point Buy vs Rolling for Stats

[Oofta]

Apologies to [MENTION=23751]Maxperson[/MENTION] who answered in the negative, but mathematically it would, provided you designed the array and the selection of the individual line of stats properly.

There are two aspects to this. The array could be the entire array produced by 3-18.
The array could be chosen by percentile dice (even 1,000 or 10,000 like Hackmaster) to achieve the same bell curve distribution as rolling 3d6.

If you wanted to reduce the variation, such as set a floor and/or a ceiling, then you'd select only the portion of the array you wanted, such as no ability less than 8, and none more than 15. So all 8s and all 15s would be in play. Provided you design the selection process to maintain the proper bell curve (not a shifted one, but so the probabilities remain intact), then this would satisfy the random stats goal.

The solution I posted at the end of post #215 solves the same problem if you want to set a floor, without the complexity of the arrays. If you wanted to set a ceiling too, then you could just declare any number rolled that's over 15 is a 15 and you can only use the excess to raise a score to 8. Seems simpler than designing an array.

Personally, if I did arrays they would be based on some variation of point buy to generate the array and have the same floor/ceiling as anyone in the group that uses point buy. I'd have to write some up, but it's just a matter of sitting down with a spreadsheet for a few minutes.

I prefer assigning points, but it seems to be a sticking point for some folks who want random character creation.

 

[Ilbranteloth]

Rolling Stat or Point Buy is a game mechanic used for Players.
Dont try to do Population Simulation based on this.

A commoner have stats around 10. Period.

But see, that's where I disagree right off the bat. A commoner isn't equally good/bad in every aspect of their life. Each person is a mix of stats, so to say a commoner is centered around an average of 10 is reasonably, but not all 10s. Certainly a blacksmith is stronger than most clergy, for example.

I also don't see any reason why a system can't be designed that acknowledges the world and not just the PCs. If you want the PCs to be above average, or different than the rest of the population, then you can design around that.

A soldier may have a better strength, so around 14.
An NPC adventurer or important NPC should be more or less equals to players, so he may have stats according to that.

If you need a significant NPC, the DM may use any method, including assigning stats on the fly without rolling.

But a soldier might not. Perhaps the training is sufficient to warrant a higher Strength than the rest of the population. But I'm not so sure that military training as a whole reflects this. English Longbowman, sure, that was a specific type of long-term training. But your average soldier conscripted to a pike and shield wall? Probably not.

Agreed on what the DM can do when creating an NPC. My goal isn't necessarily to roll for every NPC, etc. It's more that I prefer a system that can be used for that if desired. Really, it's a matter of defining what result you want in terms of stats, and the system is designed to support that. If I know that the system is designed to average 72 points, then I can use that as a guideline for creating an NPC, and make a more average individual closer to 63, and an exceptional person 75 to 80.

 

[Oofta]

In my entire (nearly 40 years) of D&D, I NEVER used point-buy; it offended me on many levels. Of course, that offense is subjective and not something I can use to convince others, but I can advocate for some methods over others, and so can anyone else.

We're all allowed our prejudices, I know I certainly have mine.

I do think it's somewhat revealing. Much like sushi, I've tried a few variations of dice rolling, and simply don't find them to my taste. You've never tried sushi point buy because you decided you wouldn't like it even before you tried it.

That's OK, and I don't mean it as a slam. Just pointing out that you come from a perspective of a conviction that has no basis in personal experience.

As far as "many generation methods", I agree. But what they have in common is that a random roll has an effect on my character for their entire lifetime, which for me regularly stretches into years. In addition, with most alternatives the implicit or even stated goal is that some characters will have superior ability scores compared to others in the same group. Much like point buy offends you, a system that has the goal of handicapping some characters and rewarding others based on one time random luck offends me.

 

[Arial Black]

Where in the current edition does it describe how to generate stats for NPCs? Because I don't see it anywhere but it wouldn't be the first time I missed something.

The only thing I could find is options on how to generate ability scores in character creation guidelines and then on page 173 it states that "Adventurers can have scores as high as 20". There are a few NPCs listed here and there which have a 20 ability.

So all we really know is that the minimum is probably 3 (although that's never really stated anywhere) and the high for mortal humanoids is 20.

All discussions of distribution of ability scores for NPCs seem to be based on rules that were written 4 decades ago or so. I don't see how it's any more relevant than trying to say that only elves can be multi-classed fighter mages and hobbits have to be thieves.

One truth about this is that the writers of 5E also played the previous editions, and carry many, many assumptions around in their heads, and (consciously or not) assume that we do too.

Sometimes, JC forgot to actually write down one of these things; they remain true nonetheless.

One of these is that the 4d6k3 method produces suitably heroic arrays compared to the mass of humanity, and that PC arrays (rolled, bought, whatever) are 'better' than 'normal' people, and that 'normal' people are represented as 3d6 in order. That is the default assumption! It remains as true in 5E as it was in 1E, so true that JC didn't even feel the need to write it down!

What about the NPCs in the back of the MM? Well, these exist to make the DM's life easier, and so the averages are assumed, rather than making the DM roll six random stats for every NPC. But the assumption is not that 95% of the population have exactly 10 in all six ability scores, the assumption is that the general population is rolled 3d6 six times in order.

Now, this would be an insane amount of work for a DM to actually do for his campaign city, and for not much benefit over the deliberately averaged commoner in the MM.

I have seen it done though. In 1E's City State of the Invincible Overlord, every single NPC had all six stats published, and they were all rolled on 3d6 in order, and the hit points were also rolled.

When you consider having to do that for every single member of a city's population, aren't you glad that you can just assume an 'average' commoner? That ease of use doesn't change the 'reality' of the whole population, just as the fact that 5E's point-buy doesn't let you buy a score of below 8 or above 15 doesn't mean that the lowest stat any human can possibly have is 8, or the highest 15 (before racial mods). Such flawed logic is what leads Hussar to incorrectly assume that the strongest possible 1st level human has a score of 16. It's taking the point-buy method as if it were this method that models the entire population, when all it is is a method to make PCs to play in the game.

 

[Arial Black]

We're all allowed our prejudices, I know I certainly have mine.

I do think it's somewhat revealing. Much like sushi, I've tried a few variations of dice rolling, and simply don't find them to my taste. You've never tried sushi point buy because you decided you wouldn't like it even before you tried it.

That's OK, and I don't mean it as a slam. Just pointing out that you come from a perspective of a conviction that has no basis in personal experience.

As far as "many generation methods", I agree. But what they have in common is that a random roll has an effect on my character for their entire lifetime, which for me regularly stretches into years. In addition, with most alternatives the implicit or even stated goal is that some characters will have superior ability scores compared to others in the same group. Much like point buy offends you, a system that has the goal of handicapping some characters and rewarding others based on one time random luck offends me.

How about the method I posted earlier? The one where each player has 12 cards (two of each of 4, 5, 6, 7, 8 and 9). Every single PC adds up to 78 points. Would you walk away from that? If so, why?

BTW, I have now played plenty of point-buy. Still, given a choice I would not choose point-buy, but if the choice was 'point-buy' or 'walk away', I would suck it up and play with point-buy and still make a PC I like. I find it difficult to understand why anyone would rather walk away and not play at all rather than play with a method they do not prefer, whatever that may be.

Playing is better than not playing, all else being equal, and the stat generation method is not an indicator of how enjoyable the campaign will be.

 

[Ilbranteloth]

Personally, if I did arrays they would be based on some variation of point buy to generate the array and have the same floor/ceiling as anyone in the group that uses point buy. I'd have to write some up, but it's just a matter of sitting down with a spreadsheet for a few minutes.

I prefer assigning points, but it seems to be a sticking point for some folks who want random character creation.

Even using only the standard array, that is only one set of stats, there are 720 variations for your array. And it grows exponentially for each line that you add. An array to generate random allocation among all allowable stats between the low of 8 and high of 15 in the standard array would be huge.

It's much easier to design a random solution that accounts for your floor and ceiling.

 

[CAFRedblade]

I run (and sometimes play) at my local store alot. Wednesday Nights AL. I organize things somewhat, and recommend using the stand. array over the point buy for use, as it gives a better even playing field for the AL modules. But when I run homebrew, I like slightly stronger heroes, so have them roll 4d6 drop lowest and re-roll 1s. But can't have an ability score above 18 even after racial modifier. You can take two points from one ability to bump up another by 1 point. And usually this is done in a apply rolls to abilities directly, first roll is str, last is cha. But I sometimes allow for selective placement. I also throw tougher encounters at the party however for this homebrew style. Both myself and the players enjoy the break in play style.

 

[Tony Vargas]

5e has no 3d6 rule that I am aware of. Everyone uses 4d6 drop the lowest.

No, by default players use 4d6. Most NPCs aren't rolled at all. Or do you think the 'Guard' in the SRD rolled 12, 11, 11, 9, 10, 9 before adding his +1 racial bonus for being human?

Less seriously, what kind of demented Lake-Wobegone 'realism' are you talking about?

 

[Oofta]

Even using only the standard array, that is only one set of stats, there are 720 variations for your array. And it grows exponentially for each line that you add. An array to generate random allocation among all allowable stats between the low of 8 and high of 15 in the standard array would be huge.

It's much easier to design a random solution that accounts for your floor and ceiling.

Your math is off, but don't have time right now to come up with the actual number. If 1 array is 15, 15, 8, 8, 8, 8, you don't need to have an array 15, 8, 8, 8, 8, 15 (or any other permutations).

In addition, the numbers still have to add up to whatever point buy goal you have. But I never said you had to come up with every possible permutation - 20 would probably be more than enough.

 

[Oofta]

How about the method I posted earlier? The one where each player has 12 cards (two of each of 4, 5, 6, 7, 8 and 9). Every single PC adds up to 78 points. Would you walk away from that? If so, why?

BTW, I have now played plenty of point-buy. Still, given a choice I would not choose point-buy, but if the choice was 'point-buy' or 'walk away', I would suck it up and play with point-buy and still make a PC I like. I find it difficult to understand why anyone would rather walk away and not play at all rather than play with a method they do not prefer, whatever that may be.

Playing is better than not playing, all else being equal, and the stat generation method is not an indicator of how enjoyable the campaign will be.

I can't account for every possible variation, and the method you list (to me) seems like just a variant of point buy. It's a far way from roll 4d6 drop lowest you get what you get no rerolls that, yes, I would walk away from.

I would think twice about your method though because it could easily add up to very "meh" characters too easily. Not that they would be bad, but if every stat was around 11-13 it doesn't give me much to hang my hat on.