Alternative Stat Rolling Rules

[Crisis21]

So I decided to play around with the rules for rolling stats in D&D. For reference, an old DM of mine told us to roll 4d6 six times, drop the lowest from each and reroll 1s once and then we could arrange the stats however we liked. He was also nice enough to let us try again (from the beginning of course) until we were happy with our stats or got tired of rerolling.

I remember it being a bit tedious, even if I got a good array eventually.

So here's my alternative proposal:

Roll 24d6 at once.
Reroll all 1s.
Select from the results three dice for each stat.
You may have no more than one stat over 16.
At least one stat must be 10 or lower.
You may do this three times and select the best result.


Here's a few test rolls I did to show how this works. Your view of which option has the best stats may differ.

Spoiler: Group 1

Show

18 (6+6+6)
16 (6+5+5)
13 (5+4+4)
13 (5+4+4)
11 (5+3+3)
9 (3+3+3)
Extra (3, 3, 2, 2, 2, 2)

18 (6+6+6)
16 (6+5+5)
13 (5+4+4)
13 (5+4+4)
10 (4+3+3)
10 (4+3+3)
Extra (3, 2, 2, 2, 2, 2)

18 (6+6+6)
14 (6+4+4)
14 (6+4+4)
13 (5+4+4)
11 (5+3+3)
10 (4+3+3)
Extra (3, 2, 2, 2, 2, 2)

Spoiler: Group 2

Show

18 (6+6+6)
16 (6+5+5)
16 (6+5+5)
13 (5+4+4)
11 (5+3+3)
10 (4+3+3)
Extra (3, 3, 3, 2, 2, 2)

18 (6+6+6)
16 (6+5+5)
15 (5+5+5)
12 (4+4+4)
10 (4+3+3)
9 (3+3+3)
Extra (2, 2, 2, 2, 2, 2)

18 (6+6+6)
16 (6+5+5)
16 (6+5+5)
12 (5+4+3)
10 (4+3+3)
10 (4+3+3)
Extra (3, 3, 3, 2, 2, 2)

Spoiler: Group 3

Show

18 (6+6+6)
15 (6+5+4)
14 (6+4+4)
12 (6+3+3)
10 (4+3+3)
10 (4+3+3)
Extra (3, 2, 2, 2, 2, 2)

18 (6+6+6)
16 (6+5+5)
16 (6+5+5)
15 (5+5+5)
13 (5+5+3)
9 (3+3+3)
Extra (3, 3, 2, 2, 2, 2)

18 (6+6+6)
16 (6+5+5)
15 (6+5+4)
13 (5+4+4)
11 (5+3+3)
10 (4+3+3)
Extra (2, 2, 2, 2, 2, 2)

Spoiler: Group 4

Show

18 (6+6+6)
16 (6+5+5)
15 (5+5+5)
15 (5+5+5)
14 (5+5+4)
10 (4+3+3)
Extra (4, 4, 2, 2, 2, 2)

18 (6+6+6)
16 (6+5+5)
16 (6+5+5)
16 (6+5+5)
12 (6+3+3)
10 (4+3+3)
Extra (3, 3, 2, 2, 2, 2)

18 (6+6+6)
16 (6+5+5)
16 (6+5+5)
14 (6+4+4)
13 (6+4+3)
9 (3+3+3)
Extra (2, 2, 2, 2, 2, 2)

Spoiler: Group 5

Show

18 (6+6+6)
16 (6+5+5)
13 (6+4+3)
11 (5+4+2)
11 (5+4+2)
7 (3+2+2)
Extra (2, 2, 2, 2, 2, 2)

17 (6+6+5)
15 (5+5+5)
15 (5+5+5)
13 (5+4+4)
11 (4+4+3)
8 (3+3+2)
Extra (2, 2, 2, 2, 2, 2)

18 (6+6+6)
12 (4+4+4)
12 (4+4+4)
11 (4+4+3)
10 (4+3+3)
8 (3+3+2)
Extra (2, 2, 2, 2, 2, 2)

Spoiler: Group 6

Show

18 (6+6+6)
16 (6+5+5)
16 (6+5+5)
15 (6+5+4)
12 (4+4+4)
10 (4+3+3)
Extra (4, 4, 4, 4, 3, 2)

18 (6+6+6)
16 (6+5+5)
15 (5+5+5)
11 (5+3+3)
10 (4+3+3)
10 (4+3+3)
Extra (2, 2, 2, 2, 2, 2)

18 (6+6+6)
16 (6+5+5)
15 (5+5+5)
13 (5+4+4)
10 (4+3+3)
8 (3+3+2)
Extra (2, 2, 2, 2, 2, 2)

Spoiler: Group 7

Show

18 (6+6+6)
16 (6+5+5)
14 (6+4+4)
13 (5+4+4)
12 (4+4+4)
9 (3+3+3)
Extra (3, 3, 2, 2, 2, 2)

18 (6+6+6)
11 (5+3+3)
11 (5+3+3)
11 (5+3+3)
11 (5+3+3)
7 (3+2+2)
Extra (2, 2, 2, 2, 2, 2)

18 (6+6+6)
16 (6+5+5)
16 (6+5+5)
13 (5+4+4)
13 (5+4+4)
9 (3+3+3)
Extra (3, 3, 3, 2, 2, 2)

Spoiler: Group 8

Show

18 (6+6+6)
16 (6+5+5)
16 (6+5+5)
13 (5+4+4)
12 (4+4+4)
10 (4+3+3)
Extra (3, 2, 2, 2, 2, 2)

18 (6+6+6)
13 (5+4+4)
12 (4+4+4)
10 (4+3+3)
9 (3+3+3)
9 (3+3+3)
Extra (3, 2, 2, 2, 2, 2)

16 (6+5+5)
15 (5+5+5)
12 (4+4+4)
9 (3+3+3)
9 (3+3+3)
8 (3+3+2)
Extra (2, 2, 2, 2, 2, 2)

Spoiler: Group 9

Show

17 (6+6+5)
15 (5+5+5)
15 (5+5+5)
14 (5+5+4)
12 (4+4+4)
10 (4+3+3)
Extra (3, 3, 3, 3, 2, 2)

18 (6+6+6)
16 (6+5+5)
14 (6+4+4)
14 (6+4+4)
12 (6+3+3)
8 (3+3+2)
Extra (2, 2, 2, 2, 2, 2)

18 (6+6+6)
15 (5+5+5)
12 (4+4+4)
10 (4+3+3)
10 (4+3+3)
9 (3+3+3)
Extra (3, 3, 3, 2, 2, 2)

Spoiler: Group 10

Show

18 (6+6+6)
15 (6+5+4)
12 (4+4+4)
12 (4+4+4)
10 (4+3+3)
8 (3+3+2)
Extra (2, 2, 2, 2, 2, 2)

18 (6+6+6)
16 (6+5+5)
15 (6+5+4)
12 (6+3+3)
8 (3+3+2)
6 (2+2+2)
Extra (2, 2, 2, 2, 2, 2)

18 (6+6+6)
16 (6+5+5)
16 (6+5+5)
16 (6+5+5)
15 (6+5+4)
10 (6+2+2)
Extra (4, 4, 4, 3, 3, 3)

Spoiler: Group 11

Show

18 (6+6+6)
16 (6+5+5)
16 (6+5+5)
13 (5+5+3)
11 (5+3+3)
10 (4+3+3)
Extra (3, 3, 3, 2, 2, 2)

18 (6+6+6)
16 (6+5+5)
16 (6+5+5)
13 (5+4+4)
12 (4+4+4)
8 (3+3+2)
Extra (2, 2, 2, 2, 2, 2)

18 (6+6+6)
16 (6+5+5)
16 (6+5+5)
14 (5+5+4)
12 (4+4+4)
10 (4+3+3)
Extra (4, 4, 2, 2, 2, 2)

Spoiler: Group 12

Show

18 (6+6+6)
14 (6+4+4)
14 (6+4+4)
12 (6+3+3)
12 (6+3+3)
10 (6+2+2)
Extra (3, 3, 2, 2, 2, 2)

18 (6+6+6)
15 (5+5+5)
13 (5+4+4)
10 (4+3+3)
8 (3+3+2)
6 (2+2+2)
Extra (2, 2, 2, 2, 2, 2)

18 (6+6+6)
16 (6+5+5)
16 (6+5+5)
16 (6+5+5)
12 (4+4+4)
9 (3+3+3)
Extra (3, 3, 3, 2, 2, 2)

Spoiler: Group 13

Show

18 (6+6+6)
16 (6+5+5)
16 (6+5+5)
14 (5+5+4)
12 (4+4+4)
9 (3+3+3)
Extra (2, 2, 2, 2, 2, 2)

18 (6+6+6)
16 (6+5+5)
14 (6+4+4)
14 (6+4+4)
13 (5+4+4)
10 (4+3+3)
Extra (3, 2, 2, 2, 2, 2)

18 (6+6+6)
16 (6+5+5)
15 (5+5+5)
13 (5+4+4)
12 (4+4+4)
10 (4+3+3)
Extra (3, 3, 2, 2, 2, 2)

[nikkoli]

It's cool, and the only thing that makes me a bit confused is if why do you have to have one stat under 10?

[Crisis21]

It's cool, and the only thing that makes me a bit confused is if why do you have to have one stat under 10?

10 or under, and so that no matter how well you roll there is at least one area your character does not excel in. One stat they're, at best, average in. Think of it like an enforced dump stat thing. Because being good at lots of things may be great, but being good at everything starts to take some of the challenge out of the game.

Of course, if you want to play it with more challenge, there's nothing preventing you from deliberately taking lower rolls for some stats.

[Nupo]

Over the years we have gone through a number of different methods for rolling up stats. All the different methods we had would occasionally yield a really crappy set of stats. It was too random. For the last ten years or more we just used the point-by method. With the point-by method we tended to end up with a bunch of cookie cutter characters. All fighters had exceptional strength, constitution, and dexterity, but had crap charisma, intelligence and wisdom. Other classes were just as predictable.

Our current method is to roll 4d6, take the best three, and roll them in order. You then add up how many points it would cost to get these stats by the point-by method. If the total is equal to or greater than the allocated amount of points, those are your stats. Pick a race, make those adjustments and you're stats are set. You then decide what class you want to make it based on your stats.

If when you add them up the total is less than the allowed points, you add to them however you like until you reach the allowed points. You then pick a race and class the same as above.

This method brings back the fun of actually rolling up stats, but guarantees you won't get a crappy set. It also generates some interesting combinations of stats that very few would pick with a straight point-by method.

For example a character I recently rolled up ended up with a high strength and a high intelligence. I decided to make him a 1/2 orc wizard. He welded a great axe, and wore hide armor. Billed himself a barbarian, even though he was really a wizard, and pretended to be dumb. Not something I would have thought of had I just did the point-by method.

[Jane_Smith]

The best solution I have found, that guarantees fairness to some extent, is that have every single player roll 4d6, take the best 3 for all six stats. Then list the results as a "pool". Every player can pick -any set- from that pool, even multiple times. So everyone can use the same stat rolls one player had rolled.

Example;

Player 1 rolls 18, 16, 14, 14, 13, 9
Player 2 rolls 17, 15, 15, 13, 12, 12
Player 3 rolls 14, 14, 12, 10, 9, 7
Player 4 rolls 16, 15, 15, 15, 14, 14

Player 1 likes his/her roll for a wizard, so they keep their set.
Player 2 and 3 are using MAD classes like bloodrager, paladin, monk, etc, and like player 4's set, so use that set.
Player 4 wants to go a rogue, and decides to grab player 1's set.

One of the largest issues with point buy is that unless you go with the 25+ points, rolling stats "almost" always comes out better in overall value then point buy. Especially if you roll 2 stats over 16, which isnt impossible. God forbid you get 3 over 16, it blows point buy out of the water. But then you come accross the issue that your the "Golden Chosen Child", and some other player's highest stat is like, 14. Nobody likes being the low-roller, but point buy is bland and cookie-cutter. So, team game, why not have stat-rolling be a group activity/shared the "wealth" as it were?

[nonsi]

.
I've seen dosens of suggestions for determining character stats over the years. None of them had managed to converge to a high enough probability of reasonable results.
This is the first suggestion I've encountered that almost guarantees results that any player would be content with and any DM could live with.

Bravo :applause:

[jqavins]

My current favorite is a method that one of my DMs came up with. You take a sheet of college ruled paper (this was a long time ago) and roll straight 3d6 for a mid-power group or top-3-of-4 for a high power group, one attribute score on each line. (Now I use a computer to do the rolling; a spreadsheet or quicky BASIC program or some such.) Then you pick six attribute scores in a row from anywhere on the page. It lets you place a good roll in whatever attribute you want, but then you're stuck with those around it for the others. Want to play a fighter? There's probably a 17 or 18 somewhere on the page that you can use for strength, but you might be stuck with an 8 constitution. Or, you can settle for the 16 strength that's somewhere else on the page and goes along with a 14 constitution. Each page of scores is to be used only once then discarded.

[nikkoli]

My current favorite is a method that one of my DMs came up with. You take a sheet of college ruled paper (this was a long time ago) and roll straight 3d6 for a mid-power group or top-3-of-4 for a high power group, one attribute score on each line. (Now I use a computer to do the rolling; a spreadsheet or quicky BASIC program or some such.) Then you pick six attribute scores in a row from anywhere on the page. It lets you place a good roll in whatever attribute you want, but then you're stuck with those around it for the others. Want to play a fighter? There's probably a 17 or 18 somewhere on the page that you can use for strength, but you might be stuck with an 8 constitution. Or, you can settle for the 16 strength that's somewhere else on the page and goes along with a 14 constitution. Each page of scores is to be used only once then discarded.

This is crazy interesting, but if im reading this right you could keep going with lines, becasue you will statistically get all 18s eventually if you were out to do that though.

[ericgrau]

10 isn't really a dump stat though. It's just average. That's a bit harder to role play than a bad stat. Though it's not necessarily a good thing RP-wise to be bad at something either.

For reference the elite array represents average rolls by the old method: 15,14,13,12,10,8. Actually the high stat average is 15.5 so elite array is a hair worse than rolling 4d6d1. Unimportant NPCs get a high stat of 13 or 11. Of course if you want a little higher stats then that's fine too.

I just don't understand how "unacceptable" it must be to have a high stat of 14 or 13 and how such a player sees that and hates rolling stats. That's just a little bit under the curve, a -1 really. The PHB does actually have a rule for rerolling really bad stats when your total modifier is +0 or lower. And with a high stat of 14 or 13 you'd have a total of +2 or +3, so you really can't go much lower without being granted a reroll. If, however, you're using a method that generates higher stats then a high stat of 14 or 13 really is unacceptable and you need a higher reroll cutoff. The total elite array modifier is +5, the total modifier in the OP's array is around 8 or 9. So for the OP's method or a similar strong rolling method I'd allow a reroll if your total modifier was +4 or lower.

[jqavins]

This is crazy interesting, but if im reading this right you could keep going with lines, becasue you will statistically get all 18s eventually if you were out to do that though.

No, you get one sheet of paper, period. That's about 35 lines if memory serves. If you do it on a computer, you get whatever number of rolls the DM has set down. It could be reasonable to allow one and only one new whole set if there's really nothing usable, but you can't just keep rolling sheets forever.

[AOKost]

My preferred way to go about figuring out stats is to give my players (or be given) 90 points to distribute as they please, point for point, as long as no state is above 18 before racial modifiers and not below 7 after.

I've had a lot of discussion with people that feel 90 points is too much, and you're virtually giving the player the ability to be super-human/super-powered.

My typical response is that on average, that gives the player the ability to assign 15 to each of the six ability scores, and that's not breaking anything. I myself, and many I've witnessed have rolled much higher stats than that, so I see 90 as a nice round number. So what if players want to optimize for 3 classes. Having fun should be more important than living by Force reinforce Adamantine rules.

[jqavins]

I agree that 90 is a lot, but it doesn't break anything and if you want a high power group then fine, use 90. The average sum of all attribute scores for a normal is 6*10.5 = 63, and adventurers are supposed to be extraordinary. Personally I'd use 75 or 80, but if you want 90 then so be it.

My problem with this personally, and it's a matter of preference, is that any system that let's people set the attribute scores arbitrarily - whether this, point buy, roll dice and assign the results, or whatever - all have a tendency to lead to characters that are very much alike. For a group that does not mind that, go right ahead.

The other extreme is to have players make six roles (using best three of four, best three of five, or some other method to skew the roles high) and use them in order. That's even worse, as it doesn't allow players to enough flexibility to make the characters they want. So I like to use some sort of compromise. There are several possible compromises, including the one I described above; another is to make six roles in a row and then swap any two, but only two. Or swap two pairs, which allows for one's top two rolls to go into two important attributes.

[Terrichance]

The current rolling method that my group has been using, has been to roll 5 D6, drop the lowest two results, and if the final result is a 7 or lower, then allow a reroll of the stat. The numbers admittedly tend to be a bit higher on average, but with this method we usually avoid having to make arrays, making character creation just that little bit faster.

[bekeleven]

When I do score rolling, I also like the "Each player rolls, use any array" method that Jane_Smith outlined above.

Just remember when doing point buy: the rolling method in the PHB averages a point buy score of 28.