D&D 5E Attribute Score Method

[DEFCON 1]

I suspect the question of ability scores mattering in AD&D is really that only HIGH scores mattered much in AD&D. Dexterity for example needed you to get a 15 before you saw an increase in the meaningful stat (AC adjustment). Constitution as well I think required a 15 before you got a bonus to hit points.

So rolling your stats more often that not would produce lots of numbers that wouldn't have any effect on your stats regardless of where you put them. You rolled one (maybe 2) stats over 15, they got put in your primary score(s) for what you are playing... the other 4/5 that were mediocre? Didn't matter where they went cause they'd have no appreciable effect.

Whereas in 3/4/5E you get a bump in modifier starting at 12, so most of your rolls will have a numeric impact on where you place it.

 

[clearstream]

I've been in too many groups where i showed up with an average stat line and other people had "rolled" multiple 18s and 16s. Point buy, all the way.

That's something I'm really enjoying about card-based methods. Drawing with no replacement, player abilities all sum to the same value. It is true that net modifiers can vary, but so far I have found minimal to no overshadowing.

The two decks suggested in this thread (@kenada's 445566778899 and my 222223333444455555) show how you can easily slide scores to produce more or less baseline challenge (versus published and DMG guideline CRs and DCs).

@kenada uses assign as desired, I use assign as drawn. The former better enables players to optimise for their desired class, the latter encourages players to try classes they might not otherwise choose (or they can forgo optimising for class). An interesting impact of assign as drawn is that the party is more likely to be asymmetrical i.e. not cover all bases equally. For example, the party might lack a tank or a healer. I don't mind that (in fact, I like it), but it does increase the risk of lacking good answers in combat, or of being stymied.

Whatever a group uses, my experience is that high scores can result in character optimisations that strain the 5e system and make a DM work harder to find appropriate challenge. A DM's taste for such work might inform where they land on that. More problematic (IMO) are methods that make possible heavy overshadowing. Deck-based methods are excellent at obviating overshadowing, as are points-buy or array methods.

 

[DEFCON 1]

Usually I have my players do Point Buy... but my Eberron games I also tried out the playing card method:

A deck of 12 cards, two fours, fives, sixes, eights, three sevens, and one nine. Shuffle them, deal them into six piles of two cards each, flip and add them together and it gets assigned to the stats in order as you flip. You then would select your class and race after the stat generation. Stats would be between 8 and 17, and there was no racial modifiers added (so you wouldn't be going over 17 regardless of the race you chose.)

 

[clearstream]

Usually I have my players do Point Buy... but my Eberron games I also tried out the playing card method:

A deck of 12 cards, two fours, fives, sixes, eights, three sevens, and one nine. Shuffle them, deal them into six piles of two cards each, flip and add them together and it gets assigned to the stats in order as you flip. You then would select your class and race after the stat generation. Stats would be between 8 and 17, and there was no racial modifiers added (so you wouldn't be going over 17 regardless of the race you chose.)

That deck has some nice features. Comparing it to the other methods discussed -

445566777889 = 76/6 =12.67
Range 8 to 17
Example array 8 10 12 14 15 17 net +6 (no racial)
Example array 13 12 13 12 13 13 net +6 (no racial)

445566778899 = 78/6 = 13
Range 8 to 18
Example array 8 10 12 14 16 18 net +8 (at least +9 after racial)
Example array 13 13 13 13 13 13 net +6 (at least +7 after racial)
I believe @kenada's intuition to cap at 17 rightly responds to the powerful stat-lines this method otherwise affords. Were I using this method I'd probably set up the deck so that rather than generating and then forbidding 18+s, it couldn't generate them. If 17 is the intended max after racial, maybe 15 should be the top of the range?

222223333444455555 = 63 = 10.5
Range 6 to 15
Example array 6 7 9 12 14 15 net +0 (at least +1 after racial)
Example array 12 12 11 10 9 9 net +0 (at least +1 after racial

One could propose
3 444 55 66 777 8 = 66 = 11
Range 7 to 15
Example array 7 8 10 12 14 15 net +2 (at least +3 after racial)
Example array 11 11 11 11 10 12 net +2 (at least +3 after racial)

 

[el-remmen]

I have never heard of this before and I am intrigued. I poked around the link you posted (& followed others, which led back here) - how do you determine the scores to draft from? It didn't seem completely random which is what I was expecting.

I have contemplated, but never used 6+2d6. It bumps the low end to a palatable 8, while not messing with the average score too much.

You hit on it, I use the "heroic" method do determining the stat pool (2d6+6), but I am not above tweaking the results to make sure there is a good range to pick from. I do this with the "random" scores esp. to allow for some mid-range flexibility.

Personally, I love this method (and so far players I've introduced it to have ranged from liking to loving it - no one has complained). It makes stat determination into a collective "Session #0" thing and folks discuss at the table how they plan to choose and the kinds of characters they are looking to play.

I wish I had saved a photo of the board. I put up all the stats on a cork board for everyone to see and hand out the actual numbers as they are chosen.

We do two drafts in a row, so each player makes two sets, one of which they choose and the other goes into a pool that is open to players making replacement characters or new players joining the game. You must take a set of stats as it stands (though for brand new players to the game who were not part of the draft I might offer the chance to switch physical and mental stats). But the idea is everyone needs to make two viable versions and choose btwn them and then the remaining ones could matter down the later.

The other rule I instituted during the draft is that when choosing a stat a player can consult one other player at the table to consult for their choice though this person can change round to round (any other consultations happen before any choices are made) to keep the draft from bogging down into discussion of every single choice ad nauseum.

 

[jsaving]

I've been in too many groups where i showed up with an average stat line and other people had "rolled" multiple 18s and 16s. Point buy, all the way.

This is the main reason why people in some of my past gaming groups have supported rolling.

 

[el-remmen]

I've been in too many groups where i showed up with an average stat line and other people had "rolled" multiple 18s and 16s. Point buy, all the way.

I mean, when it comes to all game rolls I trust people to roll (wouldn't play with people who I didn't) but when it comes to making characters, it is just more fun to all get together (well, virtually these days) and roll in view of everyone and talk it through.

 

[(Psi)SeveredHead]

I was just curious what the new trends are. So far it looks like rolling for attributes is the least preferred method.

I think it's been this way since 3e came out. Having said that, I've seen this questions asked elsewhere. GMs prefer point buy. Players have a more mixed attitude. What the GM says goes though.

I voted for point buy. As a GM I never allow stat rolling. As a player I don't participate if stats are rolled.

 

[ccs]

I've been in too many groups where i showed up with an average stat line and other people had "rolled" multiple 18s and 16s. Point buy, all the way.

And thats why we roll in front of the group.

Weirdly its much less common to roll only fantasticly great scores when there's even one witness....

 

[Oofta]

And thats why we roll in front of the group.

Weirdly its much less common to roll only fantasticly great scores when there's even one witness....

But it does still happen. I've seen multiple 18s, low of 14 for one PC while another PC had a single 14 a 12 and everything else lower than 10. It's one of the reasons I'll never use rolling.