Hybrid Ability Stats: Rolling and Point Buy

22 posts / 0 new
Last post
I'm just throwing this out there for fun and to see what people think. I read through some of the point buy vs. rolling thread but not all. If someone else has suggested this already I apologize:

Hybrid System:

The thought behind this is to get randomness into abilities while allowing characters to still be relatively on the same level. Personally I like the solid ground that point buy gives me to stand on but I also miss the randomness and rush of the roll.

How to use this system:

Pick a point buy number that you want use. For my example I'll use the standard point buy of 25. I figure I want to make a Dwarven Cleric so I choose these numbers for the abilities:

STR 13
DEX 10
CON 12
INT 10
WIS 14
CHA 14

Of course I am hedging my bets here a little and being conservative with the spread of the scores. I realize there are ways of optimizing with point buy buy I am not so concerned with that.

I add racial modifiers which now gives me these stats:

STR 13
DEX 10
CON 14
INT 10
WIS 14
CHA 12

Here is where the randomness comes in. I'm using a D20 (I'm not exactely sure why but it seemed to fit) to do this:

Score Ability Adj.
1-4 -2
5-8 -1
9-12 0
13-16 +1
17-20 +2

or you could use a more conservative take:

1-8 -1
9-12 0
13-20 +1

Just for fun I'll use the more agressive approach. Here are the six rolls which I will use for the abilities in order:

Roll Adj.
1 -2
10 0
5 -1
2 -2
15 +1
19 +2

Now the final ability scores are:

STR: 11
DEX: 10
CON: 13
INT: 8
WIS: 15
CHA: 14

So maybe I'm taking to much of a middle ground here but also this might be a good compromise. It'll be interesting to hear what people think.
Interesting, to say the least. What I don't like about this method is the ability to spend most of your points getting an 18 only to then roll a 1 and have it bumped down to a 16. In that case, you've clearly decided to specialize, spending an inordinate amount of effort on that stat, and you've gotten punished by simple bad luck; getting punished by luck is the single thing I hate about rolling stats.
You could try applying the d20 randomness to the score before points are spent, which should also imply before the numbers are assigned to stats. So you use your d20 rolls to adjust the point-buy baseline up or down from 8; spend your points as normal from there, and then assign the six resulting stats wherever you'd like.

But in the end I still don't really like the method, though it is certainly intriguing. I think what bothers me about it is the seemingly arbitrary nature of the randomness added to the predictability of point buy. It feels weird using a system of calculated returns and then imposing some randomness on top of it. What feels more natural, at least to me, is imposing a semblance of order on an otherwise random method (like my favorite dice method, Organic Characters).


Anyway, since we're talking about hybrid stat generation, I'll take the opportunity to bring up my favorite stat generation method of all: using playing cards.

Take two aces, the four deuces, three threes, three fours, four fives, and four sixes (a total of twenty cards). Set aside two of the deuces and shuffle the remaining cards.
Now deal these eighteen cards out into six piles of three; their totals indicate your six scores. Take the two discarded deuces and put each deuce on any score you like, increasing that score by two points. You can even put them both on the same score. The catch is that no score can exceed 18.

Now assign each score to an ability, and you're finished. What I like about this method is that while there is some randomness (in how the cards are allotted), every character has roughly equivalent stats. Furthermore, you get to choose where those deuces go, which can give you some control over your final scores. And if you're not satisfied with the power curve, you can change the collection of cards to adjust up or down as you see fit.

Some sample arrays generated via cards (given before floating deuces):
16, 16, 10, 13, 7, 9
10, 11, 12, 14, 14, 10
12, 12, 10, 17, 13, 7
I tried something similar to that, but it ended up just messing up the players. The randomization needs to happen on the front end, so the player can make the adjustments he wants.

The problem with d20 as it is currently constructed is that (with very few exceptions -- some feat access and carrying capacity come to mind), a 13 is no better than a 12. Until this is fixed somehow, point buy will usually be the superior method from a player's point of view.

We play a fairly high-powered game. 4d6, take the best 3. Then you can reroll any one die of any one of the rolls. So if you rolled a 6 - 6 - 1 - 1, you could reroll one of the 1's to try to get a high score. If you don't like the outcome, build on 32 points.

About 40% of the time, the player keeps the die rolls. 32 point-buy is that good.

For 4e, I'm thinking of toning it down, either to the elite array method or going way back to AD&D -- roll 3d6 six times for each score, in order, keep the best score.
Interesting, to say the least. What I don't like about this method is the ability to spend most of your points getting an 18 only to then roll a 1 and have it bumped down to a 16. In that case, you've clearly decided to specialize, spending an inordinate amount of effort on that stat, and you've gotten punished by simple bad luck; getting punished by luck is the single thing I hate about rolling stats.
You could try applying the d20 randomness to the score before points are spent, which should also imply before the numbers are assigned to stats. So you use your d20 rolls to adjust the point-buy baseline up or down from 8; spend your points as normal from there, and then assign the six resulting stats wherever you'd like.

But in the end I still don't really like the method, though it is certainly intriguing. I think what bothers me about it is the seemingly arbitrary nature of the randomness added to the predictability of point buy. It feels weird using a system of calculated returns and then imposing some randomness on top of it. What feels more natural, at least to me, is imposing a semblance of order on an otherwise random method (like my favorite dice method, Organic Characters).


Anyway, since we're talking about hybrid stat generation, I'll take the opportunity to bring up my favorite stat generation method of all: using playing cards.

Take two aces, the four deuces, three threes, three fours, four fives, and four sixes (a total of twenty cards). Set aside two of the deuces and shuffle the remaining cards.
Now deal these eighteen cards out into six piles of three; their totals indicate your six scores. Take the two discarded deuces and put each deuce on any score you like, increasing that score by two points. You can even put them both on the same score. The catch is that no score can exceed 18.

Now assign each score to an ability, and you're finished. What I like about this method is that while there is some randomness (in how the cards are allotted), every character has roughly equivalent stats. Furthermore, you get to choose where those deuces go, which can give you some control over your final scores. And if you're not satisfied with the power curve, you can change the collection of cards to adjust up or down as you see fit.

Some sample arrays generated via cards (given before floating deuces):
16, 16, 10, 13, 7, 9
10, 11, 12, 14, 14, 10
12, 12, 10, 17, 13, 7

Interesting thoughts Fiegmund. Putting the randomness up front may be a good idea. I think I like that.

As far as spending the points for an 18 and then in the unlucky event getting the roll that gets you -2 for that stat I understand the concern. But is that any different than rolling poorly for stats.

Maybe the -1, 0, +1 adjustments are the better way to go. It may provide more variety without punishing players for poor rolls too much.
Yeah I like this idea. But I maintain there should be multiple possible methods of ability score generation, including just straight die rolling. My personal favorite is the organic method where you can reroll one and switch two around.
I had always thought a "community roll" was the best idea/
One system that I'd like to try out some day is random rolls with a safety net.

First, assign ability scores using the standard array or whatever point buy you're comfortable with - that's your safety net.

Then, roll your dice in order - 3d6 or 4d6 drop lowest or something else, depending on taste, and take the higher of the assigned or the rolled number for each ability score.

This means you will never end up with a character worse than the point buy you started out with, but with a good chance of a character that is better, possibly even with a high score in an unusual ability score for his class.
A method that I sprung on my players was to have them each roll an ability score array using the standard method, roll 4D6 and drop the lowest.

So the highest array that a player rolled happened to be 16, 16, 14, 13, 10, 9

So I took this array and gave it to all the players to distribute as they see fit.

Seems flawless right? Wrong!

The last time I used this method, a player rolled, no joke, 18, 18, 17, 16, 16, 14. This player was new to the group and was super excited about these ability scores, and though I would have rather gone with a more balanced array, one with strengths and weaknesses, I didn't want him to feel cheated out of his uncanny streak of luck so I allowed it. Afterall, ability scores don't count for beans at higher levels anyways. So when I told everyone that they would be using this array, the kid who rolled it up through a fit. See. it wasn't enough that he had ridiculously high ability scores, he wouldn't be satisfied unless he had better scores than the rest of the group. Powergaming little tard.

So my point is. point buy for the win!
One system that I'd like to try out some day is random rolls with a safety net.

First, assign ability scores using the standard array or whatever point buy you're comfortable with - that's your safety net.

Then, roll your dice in order - 3d6 or 4d6 drop lowest or something else, depending on taste, and take the higher of the assigned or the rolled number for each ability score.

This means you will never end up with a character worse than the point buy you started out with, but with a good chance of a character that is better, possibly even with a high score in an unusual ability score for his class.

Now this is interesting as well. Personally I'd go with a 25 point buy and 3d6 but again that is my style of play. I think I'll save this a give it shot at some time. Nice thinking Firelance.
A method that I sprung on my players was to have them each roll an ability score array using the standard method, roll 4D6 and drop the lowest.

So the highest array that a player rolled happened to be 16, 16, 14, 13, 10, 9

So I took this array and gave it to all the players to distribute as they see fit.

Seems flawless right? Wrong!

The last time I used this method, a player rolled, no joke, 18, 18, 17, 16, 16, 14. This player was new to the group and was super excited about these ability scores, and though I would have rather gone with a more balanced array, one with strengths and weaknesses, I didn't want him to feel cheated out of his uncanny streak of luck so I allowed it. Afterall, ability scores don't count for beans at higher levels anyways. So when I told everyone that they would be using this array, the kid who rolled it up through a fit. See. it wasn't enough that he had ridiculously high ability scores, he wouldn't be satisfied unless he had better scores than the rest of the group. Powergaming little tard.

So my point is. point buy for the win!

Or it could just be a problem with the player.
Hybrid System:

The thought behind this is to get randomness into abilities while allowing characters to still be relatively on the same level. Personally I like the solid ground that point buy gives me to stand on but I also miss the randomness and rush of the roll.

If you introduce randomness into a point buy system, you totally destroy the most important aspect of point buy and that is the lack of randomness. The reason I would choose to use point buy over rolling is that point buy has no randomness. If you have a system that introduces randomness into point buy, you would give me no reason to really choose point buy. I might as well just roll.
<\ \>tuntman
If you introduce randomness into a point buy system, you totally destroy the most important aspect of point buy and that is the lack of randomness. The reason I would choose to use point buy over rolling is that point buy has no randomness. If you have a system that introduces randomness into point buy, you would give me no reason to really choose point buy. I might as well just roll.

I don't think you understand what my suggestion is here. And what I'm suggesting here is that point buy can allow for a baseline. Rolling on top of point buy allows for some randomness without the huge potential disparity between characters. It allows for variety rather than people figuring out optimization.

With all that said, of course, if you do not like the suggestion there is no reason for you to use it.
I don't think you understand what my suggestion is here. And what I'm suggesting here is that point buy can allow for a baseline. Rolling on top of point buy allows for some randomness without the huge potential disparity between characters. It allows for variety rather than people figuring out optimization.

With all that said, of course, if you do not like the suggestion there is no reason for you to use it.

The 4 point variance (from -2 to +2) is quite a swing. If you were a spellcaster, you would want your primary casting stat to be no less than 15. With your hybrid system, every spell caster would be compelled to put a 17 to their primary stat in order to ensure the stat is no less than 15. That would be more than half of your points on the one stat. If your intention is to allow for a baseline, you could baseline one good stat and that's probably about it and the rest of your stats you risk being average to way below average. For these reasons, I would choose not to use it.
<\ \>tuntman
The most important thing is not what stats you end up with but that players have equal stats .. not equal as in identical but equal as in no player is 'gimped' vs another.

I still like rolling the dice tbh, though I now have players use the playing card method after seeing it in this post (above).

The most extreme stats you can get are 112=4 233=8 344=11 455=14 556=16 666=18 (you can add +2+2 to give you either 18, 18, 16, 11, 8, 4 (uber but with a harsh weakness) or 18, 16, 14, 11, 8, 8 (to eliminate the weakness).

The most boring stats you can get are: 633=12 453=12 462=12 462=12 165=12 155=11 .. with the extra cards you get a +16, +12, +12, +12, +12, +11 .. which is still reasonable

We use the cards that come with the wizard dnd mini's .. grab 2 lvl 6 creature cards, 2 lvl 5 creature cards etc.

Best thing is that it keeps everyone happy .. the player that likes to build a character given the stats they get and doesn't even reorder them is happy, the player that has an idea for a character and is aiming for the best stats for that particular toon is happy, at most they are a few points off, nobody gets an uber set of stats and dominates and nobody gets a gimped set of stats and .. suicides ;)

Take two aces, the four deuces, three threes, three fours, four fives, and four sixes (a total of twenty cards). Set aside two of the deuces and shuffle the remaining cards.
Now deal these eighteen cards out into six piles of three; their totals indicate your six scores. Take the two discarded deuces and put each deuce on any score you like, increasing that score by two points. You can even put them both on the same score. The catch is that no score can exceed 18.

Now assign each score to an ability, and you're finished. What I like about this method is that while there is some randomness (in how the cards are allotted), every character has roughly equivalent stats. Furthermore, you get to choose where those deuces go, which can give you some control over your final scores. And if you're not satisfied with the power curve, you can change the collection of cards to adjust up or down as you see fit.

Some sample arrays generated via cards (given before floating deuces):
16, 16, 10, 13, 7, 9
10, 11, 12, 14, 14, 10
12, 12, 10, 17, 13, 7

The playing card method is wonderful, definitely the best I've encountered so far.

The problem with rolling is not randomness, it's imbalance. When your total modifiers equal +3 and the guy next to you totals +12 it just kind of sucks.

A system in which the total of your ability scores is always the same but the individual scores are randomized retains the fun of rolling without the inevitable imbalance.
I think this is the sort of compromise that satisfies few to none of the parties.

Point buy offers control and it offers equivalence. To someone who values either, your proposal offers a degree of control and equivalence but with a promise to muss it all up at the end.

"Munchkin" rollers (where rolling is just a mechanism to generate as high scores as possible - and yes, I have done this for computer games on occasion) won't find it satisfying, because it attempts to severely restrict broad-ranging high scores.

"Organic" rollers (who just like creating from the "raw materials" of the scores) won't find it satisfying either, since you get 80% of the way via point buy and then throw in a bit of statistical noise at the end.

Ultimately, you end up importing the things rollers dislike about point-buy and the things point-buyers dislike about rolling. Aggregating the stuff people don't like and calling it a "hybrid" generally just makes everyone unhappy.
here is a variation that has the rolling first.

first you take 6 rolls, thinking ether 3d4, 2d6, or just plain 3d6 for high random affect, and placing each stat to what you want it to be in. then you get a pool of points to build off that, i was thinking somtinhg around 30 for 3d4, 25 for 2d6 and 18 for 3d6. its like how life is: your genetics give you your base, and your life makes them grow (or not.)
but i still wouldnt allow above 18 until race mods are added in.
that seems like a good way of doing it.
was that kinda what you were shooting for?
oh and about the begining rolls, unless you got all realy low numbers, you dont reroll them. that would unbalnace the system.
I think this is the sort of compromise that satisfies few to none of the parties.

Point buy offers control and it offers equivalence. To someone who values either, your proposal offers a degree of control and equivalence but with a promise to muss it all up at the end.

"Munchkin" rollers (where rolling is just a mechanism to generate as high scores as possible - and yes, I have done this for computer games on occasion) won't find it satisfying, because it attempts to severely restrict broad-ranging high scores.

"Organic" rollers (who just like creating from the "raw materials" of the scores) won't find it satisfying either, since you get 80% of the way via point buy and then throw in a bit of statistical noise at the end.

Ultimately, you end up importing the things rollers dislike about point-buy and the things point-buyers dislike about rolling. Aggregating the stuff people don't like and calling it a "hybrid" generally just makes everyone unhappy.

I see what you are saying. Personally, I find that there is some sort of excitement with random rolls but I just wanted to see if there was a way to mitigate the problem by combining some randomness but with a saftey net.

I have no idea why there can't be a middle ground. The attempt to make ability scores HAVE to be either roll or point buy doesn't make sense. The fact that you assume that most people don't like the middle ground is too bad. Would you rather not have that choice?

And I don't see how this would make anyone unhappy. This is simply a choice that can be laid out there along with pure roll and pure point buy. If you don't like it, don't use it.
Seriously, read the playing card post. It accomplishes exactly what you're looking for - middle ground, the best of both worlds.
If you want to hybridize the point buy method with rolling, why not use the "base" portion of the point buy and add rolling to it. The PB method uses bases of 8 + varying points the player assigns. Perhaps taking the base 8 and adding a d10 roll for each score would work. The ability scores would range from 9-18 with an average of 13.5.

Or, if that seems too excessive, how about 2d6+6? The scores would rnage from 8-18 with an average roll of 13, but with a bell curve to assure more mid-ranged scores.

This would make sure that it is highly unlikely you would have feeble characters with all single digit scores, yet also highly unlikely you would get superman characters with all 15-18 scores.

Of course, if anyone could use these methods and wind up with singles digit characters, it would be me. In my last campaign, we used a method we found in the back of the 1e Unearthed Arcana:

[INDENT]You get to roll 9d6 for one stat, 8d6 for another, 7d6 for the third, 6d6 for the fourth, 5d6 for the fifth, and 4d6 for the last (each time taking the best 3 dice rolled). You had to decide beforehand how many dice you would roll for each stat. As a Human Paladin, I rolled 9d6 STR, 8d6 CHA, 7d6 CON, 6d6 WIS, 5d6 INT, and 4d6 for DEX. My scores wound up being 15 STR, 16 DEX, 15 CON, 13 INT, 12 WIS, 16 CHA. So even rolling 4d6 for my DEX, I still wound up with a 16 and rolling 6d6 for WIS, I only got a 12.[/INDENT]
I see what you are saying. Personally, I find that there is some sort of excitement with random rolls but I just wanted to see if there was a way to mitigate the problem by combining some randomness but with a saftey net.

Here is how I make my players generate their ability scores:

Use the standard 4d6 drop lowest and arrage as desired. If the player is unsatisfied with the results of these stats, they use the standard point-buy.

The standard point-buy give scores that are slightly lower than the average you get by rolling 4d6. However, with the point-buy, you are guaranteed to get ability scores that are at least satisfactory. I like the randomness of the standard rolling method. As a safety net, I use the standard point-buy in case the player rolls too low for his liking.

In my experience, about 75-80% of the time, the player likes his rolls. The other 20-25% of the time, they are OK with the point-buy scores. The last time I DM'ed, one player rolled an 18, but settled for the point buy because his other scores were to low for his liking.
<\ \>tuntman