Does 1d12+5 for ability scores work OK?

[Azzy]

If you are interested in determining the probabilities of various dice combinations, AnyDice is a great tool.

 

[CTurbo]

I think I'm gonna go with 9d2 or 1d4+14 from now own

 

[robus]

Try (3d12, pick the middle one) + 5

As d12 is obviously the best die, this is great!

 

[Enevhar Aldarion]

Since I generally hate how people "roleplay" low stats, I tend to try systems that eliminate any scores that give worse than a -1 modifier, so I use 2d10+8, drop the lowest, for a range of 9-18. If I want it so it takes a racial modifier to have an 18 to start, then I will use 2d8+8, drop the lowest.

 

[Coroc]

Whats the purpose? do you want to create more swingy stats? Higher than average stats? Or just a simpler (quicker) method to create stats?

There are several methods to achieve each of your goals.

e.g.

3d6 12x select the 6 best

4d6 drop lowest

What also might work is 7d6 record the dice, add dice to 8 as you see fit, but you cannot exceed 18 and you must stay below or equal 18 so you cannot add 2 sixes but need e.g. a four and a six or 2x five to reach 18.

For 3 high and 3 low scores in a quick way : select the three stats you want to roll with expecting a high score, roll 14 + 1d4, select the three low scores, roll 8+1d4
This one probably helps you most.

 

[SouthamptonGuild]

It depends what you are trying to achieve. Assuming you're doing 5e D&D then if you want an OP party do roll 7 sets of 4D6 keep 3 (r74D6k3), that should give you a big swing up in power.

If you are more interested in party balance then the best way is point buy. I find standard array to be a bit rubbish for species with +2, and if you must have random then you can use some of the following:

Roll 5 sets of 3D6, total the results and then stat 6 is 72 -n.

Everyone rolls 3D6, 6 times to make an array of stats and then your party chooses which individual array all characters will be made with.

Basically, I roll terribly, and having six dice rolls at the start of my character's career mean that they will just never be as good at their job as someone who rolled well in a different job is really irritating. Imagine playing a game where your best stat was 14. It would be fun once, but the second or third time round it would start to pall.

 

[TwoSix]

My own preferred method is random but bounded.

1) 2d6+5 5 times.
2) Sum those 5, then subtract that number from 77.
3) If the last number is lower than 7 or higher than 17, make it either a 7 or 17 and add/subtract the remainder from the lowest/highest remaining score.

 

[billd91]

Not a Maths buff here (at all). Just wondered. Thanks.

Is there anything in auto-starting with a 16 in your main stat (after racial addons) and totally randomising the rest?

I've played in games with a variation on this. We played 2e AD&D with something like this method. We rolled 4d6-drop lowest until we got 16 or higher, then we also took the next 5 rolls. The PCs were guaranteed to have at least one stat over 15 as a result, but not necessarily limited to 16 in that stat.

Middle Earth Role Playing (a variation on Rolemaster) had PCs roll all stats (on a d100 scale) but then had them put their lowest in their class's prime stat and replaced it with a 90 (if memory serves).

 

[necrotizer]

In addition to points buy, standard array, and a wide array (18, 13, 10, 9, 8, 5), I let players individually choose a method from a variety of options all balanced around a 12 average. I think you'll recognize one of them.

NORMAL METHOD (12 average)​

• Generates ability scores between 6 and 18, with an average of 12.
• Rolls above and below 12 are equally likely.
• Repeat 6 times: Roll three dice (3d6); reroll ones. [Functionally equivalent to 3d5 + 3]
• Assign the numbers to the ability scores you choose.

DROP-LOW METHOD (12 average)​

• Generates ability scores between 3 and 18, with an average of 12.
• Rolls above 12 are more likely than rolls below 12.
• Repeat 6 times: Roll four dice (4d6); drop the lowest roll.
• Assign the numbers to the ability scores you choose.
• Subtract 1 point from any ability. No ability may fall below 3. [Adjusts average to 12.]

JACK METHOD (12 average)​

• Generates ability scores between 9 and 15, with an average of 12.
• Rolls above and below 12 are equally likely, but most scores will be 12.
• Repeat 6 times: Roll three four-sided dice (3d4); reroll ones and add 3 to the total.
• Assign those numbers to the ability scores you choose.

UNIFORM METHOD (11.5 average)​

• Generates highly variable ability scores ranging from 6 to 17.
• Risky, as ability scores of 6, 12, or 17 are all equally likely.
• Repeat 6 times: Roll one twelve sided die (1d12); add 5.
• Assign the numbers to the ability scores you choose.

MAD METHOD (11.5 average)​

• Useful for characters that are multiple ability-score dependent (MAD).
• Generates at least two above-average scores at the expense of a lower overall average.
• Generate 6 numbers:
  • 5d6 (drop 3 lowest); add 6. (range: 8-18, average 15.93) [95% of the time greater than 13]
  • 4d6 (drop 2 lowest); add 6. (range: 8-18, average 15.34) [95% of the time greater than 12]
  • 4d6 (drop 1 lowest). (range: 3-18, average 12.24) [95% of the time greater than 8]
  • 2d6; add 3. (range: 5-15, average: 10)
  • 2d6; add 2. (range: 4-14, average: 9)
  • 1d6; add 3. (range: 4-9, average: 6.5)
• Assign the numbers to the ability scores you choose.

 

[Shiroiken]

The last time I was a player, that DM made us all roll 4d6 drop lowest twice and choose, but there were no rerolls or redos period. That guy REALLY believes in the random.

I personally don't like rerolls or redos, unless it's forced rolling. I've offered either rolling or point buy as an option, but if you roll you're stuck with it. This represents the risk of getting higher stats (statistically likely), but with a descent chance of being worse than point buy. Eventually I relented (my players hate point buy), and now allow a rolled character to default to standard array, which is just worse than point buy.

I've used point buy since 3E. Why? I hate the swing factor. Not for PCs, in general -- I can adapt to any group power curve. It's because it almost always results in one player having super high stats, with another having super low stats. Sure, you can set lower bounds and eyeball rerolls, but what's the right level of "cheat"? Also, if the PCs have a possibility of coming up 18, 18, 17, 16, 15, 14, shouldn't there also be a consummate risk? Granted, being simply "average" when another PC has the above stats is punishment enough. It's even worse when you have to play the "standard array" powered character while Superman came on a reroll after that player got a first roll bad enough to be considered unplayable, especially if the chump was about the same number of bonuses below your standard character as your standard character is below the new one.

Variable ability scores within the party was a significant issue in 3E (and probably 4E too), but it's not quite as bad in 5E. With bounded accuracy a character can still contribute, if at a lower capacity (I've done it myself). Of course, if you have hyper competitive players, this is a serious problem, regardless of edition. I was a player in a campaign that died in session 0 because one player rolled way better than another player, causing a massive argument.