
jonnythm |

So I was about to go to bed when I read a few posts about ability scores and rolling vs. point buy, and it got me thinking.
People who like point buy love the fact that it allows you a guarantee that you will have appropriate ability scores for your character.
People who like rolling love that it generates characters that are more random, or seem off the wall.
People who don't like point buy complain that it creates "cookie cutter" characters that are the same
People who don't like rolling complain that it makes it possible for a great variation in the power level of the characters. (I rolled three 18s! Look at me go!)
I often have heard people forcing characters to have a certain order of ability scores based on the way they rolled (roll STR first then DEX, CON etc.) So I thought of a new way to get that randomness the dice rolling character generators love.
You want odd characters with certain random stats, but you don't like throwing caution to the wind with the dice and possibly having horribly over powered or under powered characters.
Why don't you get a point buy array, and then randomly select which ability score goes where? For a little more randomness you could have multiple arrays and randomly select the array you get.
This would make the cookie cutter feeling go away, and would make the characters determine the class they choose based on their ability scores and not the other way around. It makes the character an individual, without the risk that the player rolls a whole bunch of 8s or a bunch of 16-18s.
Has this idea been used elsewhere? (google searching it didn't reveal anything)
Would you use this method?

jlord |

One way to have randomness in a more reliable curve of stats is to use playing cards, 2 suites, #s 4-9. Shuffle them up and pull two cards for each stat. (Do NOT replace the cards in the deck, you should have no cards remaining at the end.) If I remember correctly it is usually about equivalent to 20 point buy.
doing this you won't get 3 18s because to get and 18 you need to pull both 9s at once and then they are out of the deck for the rest of the stats. if you get an 8, that means you pulled both 4s, so they are out and the rest of your stats will be higher.

![]() |

This would make the cookie cutter feeling go away, and would make the characters determine the class they choose based on their ability scores and not the other way around. It makes the character an individual, without the risk that the player rolls a whole bunch of 8s or a bunch of 16-18s.
The earliest character creation rules were roll dice and assign them to the ability scores in order. The fact that people couldn't play the class they wanted led to a variety of changes including point buy, so people could create the character they wanted. You don't make people roll randomly for race, why make them choose a random class or struggle with poor numbers in the key abilities of the class.

Tark of the Shoanti |

I have always been a fan of settings like Earthdawn and Shadowrun, how the races have their ability score already set. At the same time, this leads to cookie-cutters. Maybe have a set of ability scores per races, as in each races starts off with a locked number, and then you have a pool to buff the starting stats with?

Valandil Ancalime |

...Would you use this method?
I would be willing to use this method. The only thing I might change is let the player place the highest stat where they want, and then randomly assign the rest. That allows a player to have a chance at playing the class they want while keeping randomness a factor.
A method I like to use is similar to jlord, but uses 18 cards divided into 6 stats.
1- randomly divide 18 cards (Set to the power level the DM wants), 3 per stat, in order
2- add up each stat
3- add 1 to a stat and 6 (max 18)to another stat
3.5(optional) switch any 2 stats
4- adjust for race
This gets you the random/organic-ness of rolling, but keeps the fairness of point buy (without the cookie cutter sameness).
The 1 is to make an odd stat even.
The 6 can make a moderate stat good (or a poor stat moderate), if you really want to play a specific class/concept

![]() |

One way to have randomness in a more reliable curve of stats is to use playing cards, 2 suites, #s 4-9. Shuffle them up and pull two cards for each stat. (Do NOT replace the cards in the deck, you should have no cards remaining at the end.) If I remember correctly it is usually about equivalent to 20 point buy.
doing this you won't get 3 18s because to get and 18 you need to pull both 9s at once and then they are out of the deck for the rest of the stats. if you get an 8, that means you pulled both 4s, so they are out and the rest of your stats will be higher.
I've been doing it this way for years. If I remember correctly though, the array you suggested gives you something more like 25 point buy.
Personally, I use a lower selection of cards (same number, same range, but not an even distribution between the cards) so in the end I have one 9, one 8, two fours, and the rest of the cards are made up of numbers 5-7. I use a combination that either gives 15 point buy, or 20 points buy.
Lately, I tend to go closer to 15 point buy.

jonnythm |

Using cards as random stat makers
That's really interesting, you could get a bunch of mediocre stats, but the odds are you'll get some high and low ones. Only problem I see here is that it's harder to explain this to new players, I think I might want to take a shot using this since a few of my players like rolling but are notorious for getting either amazing or awful ability scores.

![]() |

In my campaigns we roll the stats in order and then the players are each allowed to swap one pair. That way, you get your highest stat in your classes' prime requisite, so you can play what you want, but you can also end up with "interesting" characters - rogues with good Wisdoms, charismatic fighters and so on..

Cyberwolf2xs |

We did something very similar for the Dragon Age RPG.
It's easier there, because you have lower overall scores, less points to spend, and one attribute point always costs one build point.
Basically you roll 1d8 (8 = number of attributes) for each point you are allowed to spend.
(Actually, it is a bit more complicated, as you have to have mechanisms to allow negative scores and to enforce maxima.)
It would take some work to make something like that function with PF, obviously.
But these card draw methods actually sound quite nice.
Has anyone taken a look at the statistics involved? ^^

Evening Glory |

jlord wrote:One way to have randomness in a more reliable curve of stats is to use playing cards, 2 suites, #s 4-9. Shuffle them up and pull two cards for each stat. (Do NOT replace the cards in the deck, you should have no cards remaining at the end.) If I remember correctly it is usually about equivalent to 20 point buy.
doing this you won't get 3 18s because to get and 18 you need to pull both 9s at once and then they are out of the deck for the rest of the stats. if you get an 8, that means you pulled both 4s, so they are out and the rest of your stats will be higher.
I've been doing it this way for years. If I remember correctly though, the array you suggested gives you something more like 25 point buy.
Personally, I use a lower selection of cards (same number, same range, but not an even distribution between the cards) so in the end I have one 9, one 8, two fours, and the rest of the cards are made up of numbers 5-7. I use a combination that either gives 15 point buy, or 20 points buy.
Lately, I tend to go closer to 15 point buy.
What if I wanted 30 point buy? Is there a good card array for that? Gestalt characters love to chew up ability scores, so I need big ones.

Valandil Ancalime |

For 30 PB I would do the following.
First, using cards applies the same value to each +1, Point Buy weights the value of a +1 depending on where it is (going from 17 to 18 is valued at 4). This means that using cards you might arrive at varying Point Buy amounts. For example;
14,14,14,14,14,14 is 30pb and 14X6=84.
18,17,10,10,10,10 is also 30pb, but only adds up to 75.
80 seems like a good value to use.
For my method (18 cards) you need your cards to add up to 73+1+6=80.
6x4
5x4
4x3
3x4
2x2
1x1 = 73
if you use 12 cards (Like jlord; 2 per stat; without my +1+7 after the cards),
9x3
8x2
7x2
6x1
5x2
4x1
3x1 = 80

Diskordant |
What I use is roll 4d6. 1s and 2s get bumped to 3. Drop lowest.
Then I add up all the points in each persons set and allow anyone who has less points to spend bonus points to *improve* their rolled stats. It's a fun system because everyone wants the other players to roll high, and them to roll low so they can get the most customization out of it.
The good is that it generates characters with almost 50% of stats in the 12-14 range, and almost even split between 9-11 and 15-18.
The bad is that this can have huge swings in game power I'd one character rolls 2 18s you're looking at a 26 point buy min. Another issue is that the average party power starts at about 15 with three players, and inches upward as players are added.

Valandil Ancalime |

...
Then I add up all the points in each persons set and allow anyone who has less points to spend bonus points to *improve* their rolled stats. It's a fun system because everyone wants the other players to roll high, and them to roll low so they can get the most customization out of it...
I would be interested to know how many bonus points people get?
Even if you try to level out the characters, rolling still leaves the chance of really obnoxiously powerful stats. In 1 campaign my group was starting, 1 guy rolled a 54pb(3.5e) stat array. If you want to pull everybody up to near his level, then you are going to have a whole group with really high stats. IMO, it is better to set the power level before rolling, and then allow randomness to decide how that power is distributed.