D&D 5E Alternate ability generation rule

[77IM]

Thank you Skyscraper for starting this thread. Experimenting with different approaches is fun.

Here's one I thunk up the other day.

1. For each ability score, roll three times on Table A and record each result, and cross it off of Table A.
2. If you roll on Table A and get a result that is already crossed off, go down and find the next available one, use that result, and cross it off. If you are going down Table A and reach the end, start over back at the beginning.

Table A:
1d20: Result
1: 2
2: 2
3: 2
4: 3
5: 3
6: 3
7: 4
8: 4
9: 4
10: 5
11: 5
12: 5
13: 6
14: 6
15: 6
16: 6
17: 6
18: 6
19-20: Treat as a roll of 1. (Don't cross off the 19-20 entry, cross off the result you use.)

3. For each ability, sum the results you recorded.

 

[Sword of Spirit]

I think, in the end, the question I wish to tackle with ability scores is: what represents a formidable ability score? How strong is 18 STR? Is it your run-of-the-mill fighter PC, or is it a rare and extraordinary feat to have 18?

If 18 (and 20 by the time you have had a couple of ability increases) is what everyone has (and is expected to have) in his main ability score, then how do you make extraordinary the magic item that provides 19 or 20 in an ability score? How is the 500 pound ogre strong, if any PC fighter is just as strong or stronger?

Personally I like that the PC that can challenge an ogre in a strength matchup is, like, pretty darn exceptional. I wish that the giant with 21 STR is a mighty opponent strong enough to unroot small trees. I want that item that provides 19 or 20 STR be extraordinary and a great find. In looking at the stats that appear in monsters and in magic items, I'm under the impression that, apart from monsters that come into play at levels 10+, there are few monsters that have an ability score that will be better than the strong ability score of a given PC in a group, with all PCs getting an ability score in the range of 18-20 at the time of creating the PC or at most by level 8.

So really, it's a relative thing, ability scores. Me, I wish for that relative value to position PCs well below ogres and giants in their strength; and well below mind flayers in their intelligence. However, there might be the rare PC that has a 18 in STR and the other that has 19 in INT. And perhaps they'll find a magic item that raises one PC's CON score to 19. Those scores will be looked upon as rare and valuable. The PC with 18 STR can actually challenge an ogre! That's, like, normally unthinkable, but when he does, the entire village will be in awe!

This is what I hope to achieve. Wonder.

If you're looking for advice on how to do that, I can give you a recommendation. Use point buy character creation but increase the point buy costs so it's really expensive to get high values. Or use a random 3d6 roll, with whatever other variables you like (there are good suggestions on this topic and others) to make sure no one gets totally hosed. Then require all ability score increases gained through leveling to be spent on feats--don't let players increase their ability scores as they level up. You might want to soften that by specifying one or two levels at which ability score increases may be spent directly rather than converted to feats.

That should more or less create the feel you are looking for.

 

[Mishihari Lord]

I like the 3d6 in order, change one to 14, method proposed. It preserves randomness, it lets you choose a class to be reasonably proficient with, and it's dead simple to use.

Another bonus with a method like this is that it lets you filter out prospective players that aren't going to fit with your group. IME a need for high ability scores goes with a lot of other player traits that aren't compatible with my playstyle or group. This would let me know they probably won't be a fit before in-play problems come up.

 

[clearstream]

I like the 3d6 in order, change one to 14, method proposed. It preserves randomness, it lets you choose a class to be reasonably proficient with, and it's dead simple to use.

Another bonus with a method like this is that it lets you filter out prospective players that aren't going to fit with your group. IME a need for high ability scores goes with a lot of other player traits that aren't compatible with my playstyle or group. This would let me know they probably won't be a fit before in-play problems come up.

My reservation with that method is that someone with fantastic stats gets a free 14, making them flawless, while someone with bad stats gets one rather meh one. It's often useful to go back to the problems. We want heroic stats and items to feel heroic (hence preferring 3d6 over other options like 4d6 keep best, stat array or point buy). We want our players to be free to choose their class. We want playable characters with some decent stats and ideally some flaws (so as noted, we don't want to give the guy with great rolls more rolls or some bonus stat that he can use to take away his flaws).

I feel like you end up needing to roll them straight and then apply some kind of test for rerolls. Back in the day we did roll 3d6 straight but we had a test that you couldn't have net negative bonuses. If you did, you could choose to reroll from scratch. But to avoid forcing choice of class upon players, I feel we need to add an option to swap your best roll onto one of your class stats. That's it really.

Using Troll I roll 10, 10, 10, 13, 12, 7. That nets zero so I keep it. I want to play a Sorceror so I swap my 13 with my 7. I then choose race which I can minimax a bit if I want.

 

[aramis erak]

I like the 3d6 in order, change one to 14, method proposed. It preserves randomness, it lets you choose a class to be reasonably proficient with, and it's dead simple to use.

Another bonus with a method like this is that it lets you filter out prospective players that aren't going to fit with your group. IME a need for high ability scores goes with a lot of other player traits that aren't compatible with my playstyle or group. This would let me know they probably won't be a fit before in-play problems come up.

My reservation with that method is that someone with fantastic stats gets a free 14, making them flawless, while someone with bad stats gets one rather meh one. It's often useful to go back to the problems. We want heroic stats and items to feel heroic (hence preferring 3d6 over other options like 4d6 keep best, stat array or point buy). We want our players to be free to choose their class. We want playable characters with some decent stats and ideally some flaws (so as noted, we don't want to give the guy with great rolls more rolls or some bonus stat that he can use to take away his flaws).

I agree with the flaws.

I think I'd prefer the OE-to-Cyclopedia D&D option.
3d6 straight down
Trade points 2:1 to raise prime requisites and lower others; none can be lowered below 8, nor raised above 16.

Another method I've considered is...
3d6 straight down
swap one pair
If all rolls below 10, you may reroll. Even one 10, stick with them
If highest is 10-13 and no others are 10+, raise second highest to 14; otherwise, raise highest to 14
If all 12+, lower lowest to 9.

In looking at the way the game is balanced, tho', I think 1d4+1d6+6 is a better fit to the bounded accuracy... range is 8-16, average is 12, central plateau is 11-13... nothing really sucks, nothing really pops.

Alternatively, use 6d6 & 6d4, and bases of 6. Assign 2 of the dice as desired to each attribute. Raises the peak to an 18

 

[Sword of Spirit]

A personal philosophy of mine is that any system the creates random stats and requires you to use them (even if you get to play with them a bit) is only suited for a campaign where players do not already come to the table with a character concept, because a character concept generally includes a basic idea of what you are good and bad at, and how good or bad you are. Random stats are essentially a way of saying, "you may not be able to make the character you want to make."

In order to use random stats in a satisfying manner you have to tell the player: Do not come with a developed character concept!

The random stats are unknown raw materials you then make a character concept from.

Otherwise you're going into an iron chef contest with a idea of what you want to make, only to find out that the ingredients they give you are totally unsuited for that dish--and usually well suited to something else.

(As an aside--it's also mildly annoying when the one person in the group who rolls remarkably well with stats that would allow him to make crazy good multiple attribute dependent characters that are otherwise very difficult to do well, chooses to go with something really simple that he could do with one or two decent stats (because that is the character concept he already had prior to rolling), and doesn't even role-play his exceptional stats. This kind of situation won't occur if you come with a blank slate, because you'll see, "wow, these stats would make a great paladin, monk, druid, assassin, or multiclass character" and will make one.)

 

[clearstream]

I think I'd prefer the OE-to-Cyclopedia D&D option.
3d6 straight down
Trade points 2:1 to raise prime requisites and lower others; none can be lowered below 8, nor raised above 16.

An ingenious system! But still pays off for people with good rolls better than for people with bad. I feel like we need options that kick in only if you already have bad rolls.

Another method I've considered is...
3d6 straight down
swap one pair
If all rolls below 10, you may reroll. Even one 10, stick with them
If highest is 10-13 and no others are 10+, raise second highest to 14; otherwise, raise highest to 14
If all 12+, lower lowest to 9.

For me that feels quite fiddly and actually not better than the simple rule may discard if net negative on bonuses. I like the net negative test because if you don't pass it but you have a shiny stat you want to keep, you can. If you do pass it because you rolled great, you can't use it to go even more ultra. And if you fail it and choose to discard you probably really did have an execrable array that would have left you with a marginalised character.

In looking at the way the game is balanced, tho', I think 1d4+1d6+6 is a better fit to the bounded accuracy... range is 8-16, average is 12, central plateau is 11-13... nothing really sucks, nothing really pops.

Alternatively, use 6d6 & 6d4, and bases of 6. Assign 2 of the dice as desired to each attribute. Raises the peak to an 18

That latter is also ingenious. Although I think OP won't want it because it lets players minimax too much: there will be too many 18s. However, if I understand the OP correctly he does want to allow 18s if the dice fall naturally. He doesn't mind outstanding: he just wants it to be rare and valued. Creating wonder.

 

[GSHamster]

Here's my suggestion: Roll 3d6 in order. Take the highest roll and apply it to the main stat for your class. The remaining rolls apply to the remaining stats in order (Str, Dex, Con, Int, Wis, Chr). This is very close to 3d6 in order, but ensures that your highest roll goes to your main stat, keeping you from completely gimping your character.

Example: You want to play a fighter. Rolls are 10, 8, 13, 9, 15, 11. You take the 15 and assign it to Str. Your stats are:

Str - 15
Dex - 10
Con - 8
Int - 13
Wis - 9
Chr - 11

 

[clearstream]

Here's my suggestion: Roll 3d6 in order. Take the highest roll and apply it to the main stat for your class. The remaining rolls apply to the remaining stats in order (Str, Dex, Con, Int, Wis, Chr). This is very close to 3d6 in order, but ensures that your highest roll goes to your main stat, keeping you from completely gimping your character.

Example: You want to play a fighter. Rolls are 10, 8, 13, 9, 15, 11. You take the 15 and assign it to Str. Your stats are:

Str - 15
Dex - 10
Con - 8
Int - 13
Wis - 9
Chr - 11

I like that. It is close to what I was thinking but neatly streamlined. And prevents abuse of a dump stat. I believe I'd still offer players the choice to reroll the array if net negative on bonuses. The DM has to decide whether someone unlucky enough to roll a really bad array should have to keep it? A player could roll 7 8 6 9 4 8... or worse! I accept that such an array isn't too likely and might be playable - or fit someone's character conception. But as a DM I wouldn't make anyone play it unless they chose to. What are your thoughts on that?

 

[Cobalt Meridian]

Many of these systems seem overly complex.
How about:

1. The player can assign a 16 to one stat and a 14 to another. It's purely their choice where they place these.
2. For the remaining stats they roll 3d6 and place them where they fall.
3. Apply race bonuses as normal

In my eyes this has two advantages;
1. It ensures that all players will have at least a 16 and a 14.
2. It gives the players an interesting choice: do they immediately place these two stats in their primary and secondary attributes of their class or do they trust their dice enough to give them two good stats out of four that they roll?

Personally, being risk adverse, I'd put them in the 16 in my primary and the 14 in my secondary but I have a couple of players who would be confident enough of rolling an 18 one time in four who might make a more creative choice.