Favorite Stat Generation Method

[VoxRationis]

So what method do people usually use when making characters for AD&D? Do people generally just roll in order and make a character to fit the stats? Or do you use some variant method that allows score assignment? I'm thinking of starting a campaign soon and I'd like to know how to set this up. I'd like some randomness to be involved and don't really want people to set all their scores themselves, but also to enable people to pick a class before rolling.

[LibraryOgre]

For 2e? Generally 4d6, drop the lowest, arrange to taste. Toss out really low sets, but mostly play as you roll.

[VoxRationis]

Really, 4d6 keep 3? Isn't that the method for 3e? I thought there was a change between the editions...

[Mordar]

Didn't AD&D (1st edition) have a list of like six options for stat generation? I sure seem to recall seeing a bullet list with options like:


3d6 in stat order
3d6 in stat order 6 times choosing whichever array you liked best
4d6 drop lowest in stat order
4d6 drop lowest assigned by player


Clearly there had to be at least a couple more...

Makes me want to go find my old books...

- M

[Breashios]

I like to do the following: Player picks their Race first. Then they are born. That is to say they roll their stats in order; strength first. Then, as they mature, they focus on whatever area their character has interest in and improve that about themselves (they spend points to become whatever they are motivated to become). Maybe they worked on strength to become a fighter because they were tired of being picked on in school since they were born scrawny (rolled low strength). Or maybe they built on their strengths – studied really hard to become a bookworm or prayed devoutly to build on their strong faith…

Unfortunately you want them to pick class first. That is actually last in my method. What happens when the party needs a certain class to cover you may ask? Easy, you take the players who overlapped aside and ask for a volunteer to play the class. After you get your volunteer you tell them they can swap two stats (or re-roll the character if you have time) and get 2 bonus improvement points on top of that. (They are playing not their first choice – so they get a small bump to make up for that.)

Roll and Improvement points are as follows: 3d6, but all 1s rolled are counted as 2s and if three natural 2s are rolled the result is a 16. If any stat is now below the racial minimum, raise it to the racial minimum. Characters then get 6 points to spend on improving stats. It costs 1 point per improvement to 15. You may not raise any stat above 15 unless it started at 15 or higher. In the case that a stat was already 15 or higher you may increase that stat once by one by spending 2 points. Finally, reduce any stat to the racial maximum.

You can increase the dice (4d6, remove lowest), increase the points allowed to improve stats or even change the 15 barrier to another number in order to produce characters of the power level you want for your campaign.

[Thrudd]

Original D&D and Basic D&D had 3d6 in order

1e AD&D had four methods listed in DMG
4d6 best 3, arrange

Roll 3d6 twelve times, keep the best 6 and arrange

3d6 in order, roll six times for each stat and keep the best one

3d6 in order, roll enough for 12 characters and pick the array you want

Method 1 was most common. I actually prefer the in-order methods. The method where you roll 6 times for each stat in order tends to give rather high results. I also like doing option 4 and let players trade arrays if they want.

[LibraryOgre]

Really, 4d6 keep 3? Isn't that the method for 3e? I thought there was a change between the editions...

As Thrud mentions, 4d6k3 was the method of AD&D 1st edition. 2e "technically" had 3d6 straight down as the default method, but also said "This isn't very good, here are some different options."

[BigBadHarve]

I'm a fan of 3d6, assign as desired.

You get largely average stats, with the occasional exception for better or worse, but flexibility to be what you want.

I don't subscribe to the notion that players should start out as superheroes, but just slightly above the norm and work their way up to epic level.

[Khedrac]

Did anyone use the method from Unearthed Arcarna?

a) Choose class
b) in stat order as determined by class roll:
9d6 keep best 3 for first stat
8d6 keep best 3 for next stat
7d6 keep best 3 for next stat
6d6 keep best 3 for next stat
5d6 keep best 3 for next stat
4d6 keep best 3 for next stat
3d6 for last stat.

Yes 7 stats - this method included Comeliness as one of the stats.

I think we always used 4d6 drop lowest, arrange in order to suit player.

[viking vince]

No, I never used that method

I give an option - either 4d6, drop the low, 6 times and arrange as desired - or - 3d6 nine times, keep whatever 6 and arrange as desired.

The first method is the better option for a high score. The second is better for avoiding a low score.

[Grod_The_Giant]

Did anyone use the method from Unearthed Arcarna?

a) Choose class
b) in stat order as determined by class roll:
9d6 keep best 3 for first stat
8d6 keep best 3 for next stat
7d6 keep best 3 for next stat
6d6 keep best 3 for next stat
5d6 keep best 3 for next stat
4d6 keep best 3 for next stat
3d6 for last stat.

Yes 7 stats - this method included Comeliness as one of the stats.

I think we always used 4d6 drop lowest, arrange in order to suit player.
That's fun, I like that.

[D+1]

Did anyone use the method from Unearthed Arcarna?

a) Choose class
b) in stat order as determined by class roll:
9d6 keep best 3 for first stat
8d6 keep best 3 for next stat
7d6 keep best 3 for next stat
6d6 keep best 3 for next stat
5d6 keep best 3 for next stat
4d6 keep best 3 for next stat
3d6 for last stat.

Yes 7 stats - this method included Comeliness as one of the stats.
Note that this method ONLY applied to humans, not any other race...

[Breashios]

... and the Unearthed Arcana method led to quite a few overpowered characters and sometimes with one bad roll, a total dud. I do not recommend it unless you want a high powered group where you are certain no one will suffer from jealousy issues. At least when you get a high powered character using one of the other methods, everyone knows that was your destiny - an awesome moment in time. They are not expecting to do so well (for that game). Using the UA method, another player might be more or less guaranteed to get what they need to roll to play the class they want, but one player might be noticeably less incredible than everyone else.

[VoxRationis]

Yeah, I was kind of looking quietly askance at the UA method... I think I'll go with something else. Maybe 4d6k3 in order or something like that...

[Psikerlord]

I prefer a homebrew variant:

1. Everyone rolls 4d6 drop the lowest, including the DM. Allocate how you like.
2. Anyone can choose anyone else's array, but possibly at a 1-3 point penalty (table discretion - purpose of this is to differentiate the arrays slightly, and give the person who actually rolled the array a benefit).

Allows for randomness but avoids imbalance between PCs if some roll really high and others roll really low.

[Thrudd]

FYI, the actual 2e methods from the PHB (that I had to look up, never played 2e much)

Method 1 - 3d6 in order
Method 2 - 3d6 in order, roll twice for each ability and pick the best
Method 3 - 3d6, arrange as desired
Method 4 - 3d6 rolled twelve times, choose the best six and arrange
Method 5 - 4d6k3, arrange
Method 6 - Point buy/dice roll hybrid - All scores start at 8, roll 7d6. Add each die to whatever ability you like, multiple dice can go to the same ability as long as the score doesn't exceed 18.


I don't like the UA method, either.
I think a more reasonable variant of that method, with the goal of giving players a better shot at qualifying for a specific class they want, would be to divide 24d6 among the abilities however you want (always taking the bests 3). The minimum is 3d6 for any ability. This would allow someone to roll 4d6 in each ability if they didn't really know what class they wanted. At the most extreme, they could roll 3d6 for five abilities and 9d6 for the sixth one, giving you a pretty good chance to get an 18 in that one score. In this method, you're still rolling the abilities in-order, so you can get surprising results, but the player has more control over what they might end up with.

[Xuc Xac]

Roll 18d6 then assign 3 dice to each stat.

[nyjastul69]

We always roll 4d6 drop the low roll for each ability. Generate 7 stats and drop the low stat.

[MeeposFire]

One thing to remember outside of prerequisites to play certain classes high stats are not that big of a deal in AD&D. The difference between average and fairly high in most stats is fairly small and becomes smaller as you gain levels.

One exception if you have an 18 in str as a fighter and you get to roll for percentile str. That can actually be big enough to really help. Otherwise even at a 17 it is just not that big of a deal (+1 to hit and damage) and I rather boost int (bonus languages converted to prof is sweet).

[VoxRationis]

One thing to remember outside of prerequisites to play certain classes high stats are not that big of a deal in AD&D. The difference between average and fairly high in most stats is fairly small and becomes smaller as you gain levels.

One exception if you have an 18 in str as a fighter and you get to roll for percentile str. That can actually be big enough to really help. Otherwise even at a 17 it is just not that big of a deal (+1 to hit and damage) and I rather boost int (bonus languages converted to prof is sweet).

Really? I was thinking the opposite. A lot of these tables seem to work on a cubic curve, where there's a plateau phase near the average with little difference from one value to another, but near the ends there's a big difference. A thief with 18 Dexterity is much more superior to one with 16 than the one with 16 is to a thief with 14.

[MeeposFire]

Really? I was thinking the opposite. A lot of these tables seem to work on a cubic curve, where there's a plateau phase near the average with little difference from one value to another, but near the ends there's a big difference. A thief with 18 Dexterity is much more superior to one with 16 than the one with 16 is to a thief with 14.

Dex is one of the better ability scores for sure but even then you really need the best ability scores to make it worth it. Notice that unlike the others dex actually improves relatively early (15 is all you need to get a bonus) and it increases at an acceptable rate after that point. Most of the other abilities for most classes do not operate that way.

High ability scores are great to have but you will have a fine character just having average ability scores in AD&D since the difference between average and above average is not very much.

[Potatomade]

My group uses 3d6, arrange as you wish. It lets you make the character you want for the most part, but means rare classes (like Paladin) are legitimately rare, and you will probably have one or more bad stats. We like squishy characters.

Another one we tried was a modified Method 6. All ability scores start at 8, roll 5d6, can add one or more die to each ability score as wished, unless that die would put it over 18. THEN, you can choose to roll a d4, and subtract that number from one of the ability scores and add it to another (still not going over 18). This allowed players to build whatever they wanted, and gave them the option of having a genuinely bad stat if they wanted a weakness (while shoring up some other part of their build). The only reason we stopped using this was because it only allowed for one bad stat, and we love overcoming weak characters.