Alternate ability score generation

[darjr]

My GM for our new S&W campaign used a new stat generation method, new to me anyway.

He rolled up six stats 3d6. Then numbered them 1 to 6. Each player, for each ability, rolled a d6 and that numbered stat was assigned to that ability.

So for example he rolled.

1 14
2 10
3 9
4 11
5 16
6 7

and I rolled

St 6
Dx 3
Cn 2
In 4
Wi 5
Ch 1

so my stats were

St 7
Dx 9
Cn 10
In 11
Wi 16
Ch 14

So I played a cleric.

It wasn't his invention but I've never heard of it before and now I'm interested in more of the same. Anybody know of any others? Especially ones geared toward party balance like this one?

 

[Greenfield]

What if you roll the same number twice?

You could end up with STR 7, Dex 9, Con 10, Int 11, Wis 9, Cha 7.

Then what?

Our groups use the Point Buy method, found in 3.5 DMG. That promotes party balance, in that everyone has the same number of starting points.

Of the random systems I've seen:

Classic: Roll 3 D6 six times, in order.
Neo-Classic: Roll 3 D6 six times, then arrange them as you choose.
High Average: Roll 4 D6 six times, keeping the best three dice from any set.
52-Pickup: Roll 18 D6 and arrange them into sets of three.
Flat Curve: Roll 3 D6 + 1 D12, six times. From each set, you may use the D12 in place of any two of the D6.
Custom Random: Roll 3 D6 six times, then roll anothr 3 D6 worth of "custom points" that you can distribute among the other sets, with a Max of 18 in any one set.

Classic comes from the original D&D, and tended to generate a lot of career shopkeepers.
Neo-Classic also tended to create shopkeepers, but you got to choose what kind of shop they ran.
High Average is a very popular method, raising the average on any given stat to 12-ish, from a hard 10.5.
52 Pickup kept the same average as Classic, but generated far fewer hopeless characters.
Flat Curve also raised the average to around 12, but because it really does flatten the bell curve, the chance of extreme numbers goes up.
Custom Random ends up with an average that's about 0.1 points lower than High Average, but because you can choose where to distribute the points you end up with more ability bonus points in the end. (i.e. fewer 13s and 15s, more 12s and 16s.)

Then there's the infamous "But I rolled it, honest!' method, where someone writes a program on their PC to generate characters by one of the methods above, and just keeps doing it over and over again, keeping a record of the one with the highest scores. Then they set it running before bedtime and get up in the morning to find their "perfect" character waiting for them.

This is one of the reasons my group invented the spell, Dispel BS! Use it early, use it often

 

[Jacob Marley]

What if you roll the same number twice?

You reroll. I think the OP left that part out. At least, that's how I've seen the method done.

@darjr - One alternate that we have occasionally discussed (but never implemented) is: 2d6+6, arrange to taste. It guarantees no exceptionally weak stats while providing some level of randomization.

 

[darjr]

Yea, re roll duplicates. It stops cheating and still provides random rolls of 3d6.

And now that I think about it he had every player roll one of those 3d6 numbers. Could be more of them to provide a bit more variation. Say 8 or 10 or even 12.

I'm not interested in rehasing all the familiar ways, like point by or 4d6 drop lowest or hope nobody cheats. I'm inspired by the attempt of DCC to provide a sort of balance via randomness, but that mostly depends upon tables and the stats still offer lots of room for way out unbalanced stats.

 

[Umbran]

Yea, re roll duplicates. It stops cheating and still provides random rolls of 3d6.

So, why not just do 3d6 in order?

 

[Jacob Marley]

So, why not just do 3d6 in order?

Each character has the same stat array; the d6 just sets the order for that character.

For example: 3d6 are rolled six times resulting in: 17, 15, 11, 10, 10, 8. Each player will use this array.

For each 3d6 roll (17, 15, etc.) Umbran rolls a d6 where 1 equals strength, 2 equals dexterity, and so on, and places his stats, rerolling any stat that has already been assigned; say, str 15, dex 11, con 10, int 8, wis 17, cha 10. Then you pick a class to fit that array.

I do the same resulting in: str 8, dex 11, con 15, int 17, wis 10, cha 10. Then I pick my class.

And so on.

The purpose is to provide each character with them same array, but to provide randomness in how that array is set.

Does that help?

 

[darjr]

Yea, that's it. I'm not sure I'd do this every time but it was fun.

 

[was]

..There are many variant stat generation methods out there. In my last campaign, I rolled 3d6 100 times (long boring weekend with nothing else to do). After recording the results, I had the PC's roll d100 six times and assign the resulting stats as they saw fit.
..It was just a nice change of pace from our usual point buy method.

 

[Grazzt]

Each character has the same stat array; the d6 just sets the order for that character.

For example: 3d6 are rolled six times resulting in: 17, 15, 11, 10, 10, 8. Each player will use this array.

For each 3d6 roll (17, 15, etc.) Umbran rolls a d6 where 1 equals strength, 2 equals dexterity, and so on, and places his stats, rerolling any stat that has already been assigned; say, str 15, dex 11, con 10, int 8, wis 17, cha 10. Then you pick a class to fit that array.

I do the same resulting in: str 8, dex 11, con 15, int 17, wis 10, cha 10. Then I pick my class.

And so on.

The purpose is to provide each character with them same array, but to provide randomness in how that array is set.

Does that help?

We've used this method a few times (mainly for a change of pace) and it produced some interesting and (in some cases) memorable characters actually.

 

[Umbran]

Each character has the same stat array; the d6 just sets the order for that character.

Sorry. Reading comprehension fail on my part. Somehow I read the OP three times over, and didn't get that, but I understand now.