D&D 5E Array v 4d6: Punishment? Or overlooked data

[EzekielRaiden]

The simplest way to generate random-but-controlled stats is with cards. You set up a fixed deck of cards and each player in turn deals out all 18 cards facedown into six piles. This guarantees that the characters must have the same stat total, and thus very nearly the same modifier total; racial stat bonuses and an ASI will fix any lingering issues.

For an "old school" feel, just do three cards of each rank 1-6. You'll have higher odds of 18 and of 3 than you would with dice, but not enormously (the math is complicated but it ultimately works out to be similar.) And the average, in terms of taking the total and dividing by 6, will be exactly the same as 3d6: 10.5. For a more "new school" feel, swap out low cards for high cards, e.g. something like four 6s, four 5s, four 4s, four 3s, one 2 and one ace. 18 is noticeably more likely, stats below 6 are impossible, and the overall pattern skews higher (the average stat is 12.5 with this hypothetical deck.) Let that one ace count as a 6, and you've guaranteed no stat below 8 (3+3+2) and made it very likely players will get either really solid average stats or a couple great stats and the others mediocre. Etc.

It gives control and fairness while still being random, and relatively easy to implement if you have a deck of cards on hand.

 

[Oofta]

If I had someone who really wanted to randomize I'd just use point buy to come up with 20 random arrays. Have them roll for which array to use and then roll for placement.

Since I've never hit anyone that wanted to truly randomize their PC since I dropped rolling for ability scores back in 2E (before that we used "modified" versions, i.e. roll 6 sets of stats and pick the one you like), I'm not going to hold my breath.

 

[DND_Reborn]

What I did for my current Monday game is roll 4d6K3 in order. Then you roll a 7th time, and can replace any of the six scores with the 7th roll if you want.

Next, each player rolled twice on a table of random races and could pick either race. If they rolled the same race twice, they could choose any race on the table instead.

It worked very well and everyone is happy with their PCs so far. One player wasn't "feeling" his first concept, so I let him make a new PC to join the party.

I'll probably stick to this method in the future.

 

[Krachek]

Rolling stats is a good tool.
Right from the start of the game, it force the DM to make a stand and established a basic social contract for the future games.

Roll or not.
If roll, In order or not.
and with some tweaks or not to avoid extreme results.

in all cases it already align the campaign into a play style and expectations.

 

[Krachek]

What I did for my current Monday game is roll 4d6K3 in order. Then you roll a 7th time, and can replace any of the six scores with the 7th roll if you want.

Next, each player rolled twice on a table of random races and could pick either race. If they rolled the same race twice, they could choose any race on the table instead.

It worked very well and everyone is happy with their PCs so far. One player wasn't "feeling" his first concept, so I let him make a new PC to join the party.

I'll probably stick to this method in the future.

Usually bad rolls are not the real problems, every body agree on the non fun and to reroll or change the character.
The problematic case happens with high rolls, like 3 18 and no real bad stat. The rolling players want to keep it, the others start to think that is like too much, and the DM wonder how he will handle that!

 

[DND_Reborn]

Usually bad rolls are not the real problems, every body agree on the non fun and to reroll or change the character.

Actually, I think they are... which is why it then leads to "every body agree" to reroll/change the PC.

The problematic case happens with high rolls, like 3 18 and no real bad stat. The rolling players want to keep it, the others start to think that is like too much, and the DM wonder how he will handle that!

How do you handle someone being taller and shorter people want to be taller, etc.?

If everyone agrees on the system, then everyone has to agree to accept whatever results they and others might get.

If you want to avoid that, you use a system which keeps rolls within the bounds and equal.

One I like was from this thread and I adjusted it slightly:

Roll 1d6. Generate a score by adding 7, generate a score by subtracting from 18.
Roll 1d6. Generate a score by adding 8, generate a score by subtracting from 17.
Roll 1d6. Generate a score by adding 9, generate a score by subtracting from 16.

This generates 6 scores between 8 and 17, averaging 12.5. It also guarantees that if you have a 17, you also have an 8. It also makes it so your best three scores can be are 17,16,15, along with a 10,9,8.

 

[Krachek]

Actually, I think they are... which is why it then leads to "every body agree" to reroll/change the PC.


How do you handle someone being taller and shorter people want to be taller, etc.?

If everyone agrees on the system, then everyone has to agree to accept whatever results they and others might get.

If you want to avoid that, you use a system which keeps rolls within the bounds and equal.

One I like was from this thread and I adjusted it slightly:

Roll 1d6. Generate a score by adding 7, generate a score by subtracting from 18.
Roll 1d6. Generate a score by adding 8, generate a score by subtracting from 17.
Roll 1d6. Generate a score by adding 9, generate a score by subtracting from 16.

This generates 6 scores between 8 and 17, averaging 12.5. It also guarantees that if you have a 17, you also have an 8. It also makes it so your best three scores can be are 17,16,15, along with a 10,9,8.

For the Monday game play you just describe, you have a perfectly clear rule,
but still you manage to change a character.
Would you have made such an exception for a character with 3 rolled 18 perfectly fit?
it’s just a question, not a blame.

 

[DND_Reborn]

For the Monday game play you just describe, you have a perfectly clear rule,
but still you manage to change a character.
Would you have made such an exception for a character with 3 rolled 18 perfectly fit?
it’s just a question, not a blame.

No, that is the system they all agreed on. So, if a player rolled 18, 12, 16, 18, 9, 14 and 18 (7th roll), they could have three 18s.

Hopefully, every one else would just cheer on the rolls. When players get bitter about another player's roll, there is a bigger problem IME. D&D isn't a competition game in that sense.

 

[Lanefan]

I'm a fan if D&D traditions, but random stats and hit points are something I'm highly opposed to. The reason is that they are incongruous with other character creation elements. They are one time random rolls that create permanent character traits.

Nothing else does that. If it did, we would roll randomly to determine race, class, background, skills, languages, feats or ASI placement, etc.

We've always had a great deal of randomness in generation of things like previous professions, backgrounds, languages known (other than your own), etc.

Random hit points have no value in modern D&D because they strait up just give some PCs a permanent advantage over others based purely on luck.

That's the point. Just like many things in real life are based on luck.

Random abilities scores can be useful in certain situations, when done in non-standard ways. They are useful when you don't want a player to come to the table with a character concept. This could be because the desired outcome is to be inspired to play different characters (or just differently expressed characters) than you otherwise would, or because there is a desire for a minigame to see how best to use those stats. Unless that is the case, your group should not use this as it will directly oppose your play goals. Standard D&D play goals involve already having a character in mind, or at the minimum wanting to be able to choose every important aspect of them. Random stats interferes with that.

Flip side: if one wants to minimize the character-build aspect of the game, random is the way to go. That way you can't come to the table with everything planned out; you have to take what the dice give you and work with that, and develop the character through play rather than pre-play mechanics.

And in the uncommon situation where the goals of character creation actually do align with one of the random stat friendly motives instead of the more common motives, you should use a variant that makes sure everyone's stats end up equal despite being random. That doesn't mean sharing a minimum value but allowing randomness to give you more--that's still the same thing, and this has all the same issues as random hit points. No, the total value of any generated random array needs to be equal.

Except - and this is the key thing - people aren't equal. So why should the characters be?

Here is an example of how methods and motive working in harmony can actually play out.

I had the goal of potentially having players fully choose their character stats, but inspiring alternate ways to express that character. I created a method of randomly rolling stats that created equal results,

Curious here - what means of rolling did you use to achieve this?

and a point buy method that created results equal to the array, but which couldnt produce all the same values. Rolling can get you higher or lower numbers, while point buy keeps you in a certain range and make balanced stats more efficient.

Each player rolled a set of stats using the random and equal method. Then each player could choose to use any of those rolled arrays (more than one person could choose the same one) or use point buy. It seemed to work because people chose from the various options rather than all doing the same thing.

OK, that's fair - as long as random was still an option, no worries.

 

[Lanefan]

For those who don't want to think through all of that:

Random Stats Quick Fix: Each player randomly generates an array of 6 ability scores. Then each player can choose to use any of those randomly generated arrays, and more than one player can choose the same array. If you want to avoid the possibility of all of the arrays being sub-standard, each player also has the choice to use point buy, which they decide after seeing all the random arrays, not before.

For a one-off I could see this. I did something similar once for a spur-of-the-moment one-off game where time was pressing: there were nine players, so I rolled a d9. That player then rolled one stat. Repeat process until 6 stats have been rolled, forming an array. Every player then had to use this array, unadjusted for anything (so no class or species adjusts etc.) and bang out a character.

But for a "real" campaign, no.

If you want random stats but that method seems unsatisfying to you, then you need to carefully consider what your motivations are and whether they are a good fit for your group and your stat generation methods.

The motivation is, to some extent, reflect the reality that not everyone is created equal (or, better word, equivalent). Further, to gently promote the idea that D&D, at the end, is a game based on luck. If it wasn't, it wouldn't use dice.