Ability Score generation method that helps MAD characters

[heavyfuel]

Hey!

Does anyone have a suggestion on one such Ability Score generation method?

I was thinking something like Point-buy, but where higher scores get considerably more expensive.

Something like

8 - 0
9 - 1
10 - 2
11 - 3
12 - 4
13 - 5
14 - 6
15 - 10
16 - 14
17 - 22
18 - 30

And give everyone 50 points.

Any other ideas?

[JNAProductions]

Hey!

Does anyone have a suggestion on one such Ability Score generation method?

I was thinking something like Point-buy, but where higher scores get considerably more expensive.

Something like

8 - 0
9 - 1
10 - 2
11 - 3
12 - 4
13 - 5
14 - 6
15 - 10
16 - 14
17 - 22
18 - 30

And give everyone 50 points.

Any other ideas?

I run 5th edition, but here's what I do:


Ability Scores: Pick six numbers. These are your pre-racial stats. No number may be higher than 18 or less than 8. You may modify them freely up until the game starts, and may pick duplicate numbers.

It might not work as well in 3.5, but it's what I do.

[heavyfuel]

Knowing my players, we'd end up with a bunch of Mary Sues. Not really the kinda game I'd like to run

[radthemad4]

You could make a bunch of arrays you're fine with and tell players to pick one. e.g. some games at my table let players choose between

[18, 16, 14, 12, 10, 08]
[16, 16, 16, 14, 12, 10]
[16, 16, 16, 12, 12, 12]
[16, 16, 14, 14, 14, 14]

Folks typically chose the 1st or 2nd one

[Kraynic]

I got talked into running a Pathfinder 1E game for a couple of guys that are usually running games and have never been able to play in a game together. In order for people to be able to be able to create pretty much any character they wanted, I had them roll their stats using 1d6+1d4+8 (10-18). I allowed a minimum total of 80 and max total of 90. I allowed them to roll until they got 2 sets within that range and choose the one they wanted. Max starting stat was 18 after racial adjustments. It definitely gave them stats to make anything they wanted with the incentive to spread their stats to bringing lower scores up with racial modifiers.

To be honest, what I really did was think of what range I wanted to see stats in and then used AnyDice to fiddle with formulas until I settled on something that would result in what I wanted.

https://anydice.com/

[PairO'Dice Lost]

The downside to making high scores really expensive is that SAD characters can afford to sacrifice in other areas because they only need the one score while MAD characters want multiple high stats--a wizard will be fine with several 8s so long as he gets his starting 16 or 18, but a monk doesn't want three or four 12s, he wants three or four 16s.

So I'd suggest considering not just ability score generation but ability score advancement. The quadratic price of ability-boosting magic items makes three +4 items more expensive than one +6 item, and the every-four-levels ability boost is great for someone who wants +5 to one stat by 20th level but pretty bad for someone who wants +2 to three stats, and since differences between SAD and MAD characters don't tend to get really bad until the mid levels, it can be more effective to focus on the advancement side than the generation side.

Something I do in most of my games is stick with pretty generic ability generation schemes but then hand out level-based ability boosts much more readily--anything from "+1 to three stats every four levels" in relatively standard games to "+1 to all stats at every even level" in games where I removed ability-boosting items to encourage more magic item variety. SAD characters do benefit from a boost to secondary and tertiary stats but aren't hugely affected, while MAD characters really appreciate the all-around better stats.

[Biggus]

The only trouble I can see with the method you suggest is that it will tend to lead to samey characters. It's so cheap, relatively speaking, to get a 14 and so expensive for every point beyond that, that you'll end up with 14's in any stat you care about even slightly, total dump stats will still be 8's, and full casters will still have 18's in their casting stat, with very few in between (edit: also, full casters can still get an 18 in their casting stat and 14's in any others they care about with a few points left over, so it doesn't really limit them much at all: if you're using that method, I'd give about 40-42PB).

A milder/more staggered increase might work better. Perhaps

15 - 8
16 - 11
17 - 15
18 - 20

or

15 - 8
16 - 12
17 - 18
18 - 26

I think maybe about 32-34PB using the first set of numbers or 36-38PB using the second would allow MAD characters to be decent while forcing SAD casters to choose between an 18 in their casting stat and really crappy other stats, or a 16-17 with a good array of other stats (personally I've found that standard 32PB isn't too bad for MAD characters, I don't think it needs increasing very much beyond that).

Edit 2: another possibility instead of/as well as this, is to make the usual dump stats do something more generally useful, along the lines suggested here:

https://forums.giantitp.com/showsinglepost.php?p=24832457&postcount=24

[upho]

I was thinking something like Point-buy, but where higher scores get considerably more expensive.Wha...? You some kind of {scrubbed} MAD-man? Refusing to play unless your vanilla monk's OP enough to satisfy your kung-fu fanboy fantasies? Yeah, I thought so... Can't you just play a SAD caster like everybody else and stop this nonsense?

Seems OK, although personally I think I would prefer a more consistent exponential increase of the point cost per +1 bought, preferably with the upper "sweet spot" in terms of bang for the buck being around a score of 16. For example, if 14 is set as the lowest score which costs more than one point per +1 added to the minimum score of 8, a cost increase growing exponentially by .8 from there would result in:

8 - 0
9 - 1
10 - 2
11 - 3
12 - 4
13 - 5
14 - 7
15 - 9
16 - 12
17 - 18
18 - 28

In detail:

9 to 13 - 1 point cost per +1 bonus Straight 1 for 1 when adding up to +5 for a score of 13.

14 - 7 IOW, it costs 7 points to add +6, about 1.17 points per +1, which is 17% more than the cost of each +1 when buying up to +5.

So with a .8 exponential cost increase, a score of 15 (adding +7) instead costs about (0.17 increase x 1.8 exponential) 30.5% points more per +1, in total about (7 x 1.305) 9.14 points:

15 - 9 About 29% greater cost per +1 than when buying up to +5.

Adding +8 should instead cost about (0.17 x 1.82) 55.1% points more per +1, in total about (8 x 1.551) 12.4 points:

16 - 12 50% greater cost per +1 than when buying up to +5 to a score.

Adding +9 should instead cost about (0.17 x 1.83) 99.1% points more per +1, in total about (9 x 1.991) 17.9 points:

17 - 18 100% greater cost per +1 than when buying up to +5 to a score.

And finally, adding the maximum +10 should instead cost about (0.17 x 1.84) 178.5% points more per +1, in total about (10 x 2.785) 27.8 points:

18 - 28 180% greater cost per +1 than when buying up to +5 to a score.

Looking at this point-buy and assuming 50 points, especially the relatively low cost of a 16 should be great for MAD builds, while the 28-point cost of an 18 remains very expensive in comparison. And at least when using less than very generous point buys, 16 is IME typically also the score of least interest to SAD builds.


Any other ideas?Short of giving different points dependent on class, unfortunately not. That is, not when it comes to ability score generation per se, as the relative benefits MAD builds may gain from point-buy tweaks tend to be rapidly diminishing after the very earliest levels IME. I think MAD vs. SAD balance issues are more - and more easily - mitigated for example by tweaks to the available races and the related magic items. For example simply ditching the cost increase for stacking enhancement bonuses to different stats along with the slot standard headband and belt restrictions has a far more noticeable difference IME (and making sure such item enchantments are easy to find and use of course). Not to mention the many things not directly related to ability scores which may be tweaked for far greater effects.

But I guess you're already well aware of this and have implemented other related changes. And it probably doesn't hurt to also tweak stat generation.

[Kaleph]

Hey!

Does anyone have a suggestion on one such Ability Score generation method?

I was thinking something like Point-buy, but where higher scores get considerably more expensive.

Something like

8 - 0
9 - 1
10 - 2
11 - 3
12 - 4
13 - 5
14 - 6
15 - 10
16 - 14
17 - 22
18 - 30

And give everyone 50 points.

Any other ideas?

I like 32-point-buy more, but your method does exactly what you're asking for, so I gues it's ok. I did a couple of checks and, although the results where slighty better than what I typically work with, they didn't seem totally bonkers.

As an alternative, I can suggest the following method, that I've used in the past:

32 points to distribute in total,
you start with 8's in all stats,
1 point = 1 stat increase (always),
you cannot exceed 18 in any stat,
the sum of no 2 arbitrarily chosen stats may exceed 34 (in total),
the sum of no 3 arbitrarily chosen stats may exceed 48 (in total).


Everything is, of course, to be considered before racial adjustments. You may get 3x16, or 18-16-14, or 17-17-14, and still have enough points spare to avoid a penalty in the other stats. This worked, but as I said, I like 32-point-buy more, even for MAD characters.

[radthemad4]

The downside to making high scores really expensive is that SAD characters can afford to sacrifice in other areas because they only need the one score while MAD characters want multiple high stats--a wizard will be fine with several 8s so long as he gets his starting 16 or 18, but a monk doesn't want three or four 12s, he wants three or four 16s.

So I'd suggest considering not just ability score generation but ability score advancement. The quadratic price of ability-boosting magic items makes three +4 items more expensive than one +6 item, and the every-four-levels ability boost is great for someone who wants +5 to one stat by 20th level but pretty bad for someone who wants +2 to three stats, and since differences between SAD and MAD characters don't tend to get really bad until the mid levels, it can be more effective to focus on the advancement side than the generation side.

Something I do in most of my games is stick with pretty generic ability generation schemes but then hand out level-based ability boosts much more readily--anything from "+1 to three stats every four levels" in relatively standard games to "+1 to all stats at every even level" in games where I removed ability-boosting items to encourage more magic item variety. SAD characters do benefit from a boost to secondary and tertiary stats but aren't hugely affected, while MAD characters really appreciate the all-around better stats.I'll second this

[Gavinfoxx]

The game balance actually improves if everyone just gets straight 18's down the board. So just do that.

[heavyfuel]

I've read all suggestions, and the array one is my favorite so far.

I'm thinking of allowing the folowing arrays (PF1 game)

Sum Point-buy Equivalent
18 12 10 8 8 8 64 13
18 10 10 10 8 8 64 13
16 14 14 12 10 8 74 20
16 14 12 12 12 10 76 21
16 13 13 12 12 12 78 22
15 14 14 14 14 8 79 25
14 14 14 14 14 14 84 30


Do these look good?

I'm thinking #6 could build a decent enough skill-focused Monk. 14 15 14 14 14 8 before racials.

I think most charcters would go for #3, even casters (though I can see some people going for #1 and #2). Getting +2 to your casting stat is nice, but having decent Con and Dex is probably better.

Despite #7 being the clear winner in PB, I don't think anyone would pick that, so it seems to me they're balanced lol.


So I'd suggest considering not just ability score generation but ability score advancement. The quadratic price of ability-boosting magic items makes three +4 items more expensive than one +6 item, and the every-four-levels ability boost is great for someone who wants +5 to one stat by 20th level but pretty bad for someone who wants +2 to three stats, and since differences between SAD and MAD characters don't tend to get really bad until the mid levels, it can be more effective to focus on the advancement side than the generation side.

Something I do in most of my games is stick with pretty generic ability generation schemes but then hand out level-based ability boosts much more readily--anything from "+1 to three stats every four levels" in relatively standard games to "+1 to all stats at every even level" in games where I removed ability-boosting items to encourage more magic item variety. SAD characters do benefit from a boost to secondary and tertiary stats but aren't hugely affected, while MAD characters really appreciate the all-around better stats.

Definitely something to keep in mind. Thanks for the suggestion!


The game balance actually improves if everyone just gets straight 18's down the board. So just do that.

I agree that it does, but as I said before, I don't like characters like that.

[Biggus]

The downside to making high scores really expensive is that SAD characters can afford to sacrifice in other areas because they only need the one score while MAD characters want multiple high stats--a wizard will be fine with several 8s so long as he gets his starting 16 or 18, but a monk doesn't want three or four 12s, he wants three or four 16s.

So I'd suggest considering not just ability score generation but ability score advancement. The quadratic price of ability-boosting magic items makes three +4 items more expensive than one +6 item, and the every-four-levels ability boost is great for someone who wants +5 to one stat by 20th level but pretty bad for someone who wants +2 to three stats, and since differences between SAD and MAD characters don't tend to get really bad until the mid levels, it can be more effective to focus on the advancement side than the generation side.

Something I do in most of my games is stick with pretty generic ability generation schemes but then hand out level-based ability boosts much more readily--anything from "+1 to three stats every four levels" in relatively standard games to "+1 to all stats at every even level" in games where I removed ability-boosting items to encourage more magic item variety. SAD characters do benefit from a boost to secondary and tertiary stats but aren't hugely affected, while MAD characters really appreciate the all-around better stats.

I'd considered doing something similar, adding ability boosts as a class feature to classes who are both MAD and tier 4 or below (two things which often coincide). I'd thought maybe +2 to any one stat every 5 levels for the characters who really struggle (like Monks) and a bit less frequent for some others (like Rangers).


I've read all suggestions, and the array one is my favorite so far.


Arrays would be my least favourite of the options as a player, it seems the most boring and restrictive. I'd at least suggest running the idea past your players and getting their reaction before you start using it.

[heavyfuel]

Arrays would be my least favourite of the options as a player, it seems the most boring and restrictive. I'd at least suggest running the idea past your players and getting their reaction before you start using it.

Is a choice between multiple arrays that restrictive? There are 7 options, of which I think 6 of them are perfectly viable (I really don't see anyone going for all 14s).

But, of course, the players will get their say.

[lylsyly]

When I DM I use 8+1d6+1d4 (average 14) and +2 to any score every THREE levels. I've been doing it this way since late in the 2E days

[Xervous]

What if key levels just handed out more points to improve ability scores via continuous point buy? Bank the extra that you don’t spend. Generate one way progress another is something I’ve come to avoid if possible in most TTRPGs.

[upho]

Do these look good?Most of these should work just fine, but unless I've missed that array #6 will come with some kind of special disadvantage, it's clearly superior to especially #3, having 40% more PB and no lower values. So I can't see why a player with "Mary Sue tendencies" looking to play a non-SAD build would pick array #3 instead.


(PF1 game)Bit of an aside, but if this is for PF, I doubt there are objective balance reasons serious enough to motivate replacing the standard PB (and say 20 points) with far less flexible arrays. AFAIK, even something as seemingly MAD as a single-classed Str-based fighter has an op-ceiling high enough to allow for a combatant mechanically superior to any SAD caster during a large majority of levels, and for giving the same fighter enough general adventuring competence to remain at the very least relevant also outside of combat all the way up to 20th. At least if the game includes more challenging combat against more powerful foes than the typical Paizo AP run as written does.

That said, if for example your players have trouble putting together more MAD PCs on par with their their more SAD ones (most players still do IME), if your game will have restricted access to magic items and limited crafting opportunities, or you expect some other game specifics or group preferences to result in a game which plays more to the strengths of SAD casters, you should of course do what you can to mitigate potential balance issues.

Also, in addition to PairO'Dice's suggestions on tweaking ability score advancement, I heartily recommend you use DSP's alternative rules for the standard "Big Six" magic items (https://libraryofmetzofitz.fandom.com/wiki/Steelforge_Items). These definitely help mitigate many of the related issues the standard magic item rules create for MAD or otherwise item dependent PCs.


I agree that it does, but as I said before, I don't like characters like that.FWIW, I definitely don't either, at least not in 3.5/PF.

[Biggus]

Is a choice between multiple arrays that restrictive? There are 7 options, of which I think 6 of them are perfectly viable (I really don't see anyone going for all 14s).


I think if you're going to use arrays you need more choices. For example, Barbarians and melee Fighters generally want two good stats (Str and Con) and don't care much about the rest, there's not an option on your list that would suit them. You can't start with a 17 or 15 in a stat at all (not a common choice, but I've seen it happen; one of the players in my current game chose Wis 17 for their Druid).

More generally, if a character is starting middle-aged, having all but one of the options being all even numbers is going to be really bad for them. I think you probably need at least twice as many arrays to suit all character concepts.

[heavyfuel]

Most of these should work just fine, but unless I've missed that array #6 will come with some kind of special disadvantage, it's clearly superior to especially #3, having 40% more PB and no lower values. So I can't see why a player with "Mary Sue tendencies" looking to play a non-SAD build would pick array #3 instead.

Bit of an aside, but if this is for PF, I doubt there are objective balance reasons serious enough to motivate replacing the standard PB (and say 20 points) with far less flexible arrays. AFAIK, even something as seemingly MAD as a single-classed Str-based fighter has an op-ceiling high enough to allow for a combatant mechanically superior to any SAD caster during a large majority of levels, and for giving the same fighter enough general adventuring competence to remain at the very least relevant also outside of combat all the way up to 20th. At least if the game includes more challenging combat against more powerful foes than the typical Paizo AP run as written does.

That said, if for example your players have trouble putting together more MAD PCs on par with their their more SAD ones (most players still do IME), if your game will have restricted access to magic items and limited crafting opportunities, or you expect some other game specifics or group preferences to result in a game which plays more to the strengths of SAD casters, you should of course do what you can to mitigate potential balance issues.

Also, in addition to PairO'Dice's suggestions on tweaking ability score advancement, I heartily recommend you use DSP's alternative rules for the standard "Big Six" magic items (https://libraryofmetzofitz.fandom.com/wiki/Steelforge_Items). These definitely help mitigate many of the related issues the standard magic item rules create for MAD or otherwise item dependent PCs.

FWIW, I definitely don't either, at least not in 3.5/PF.

You didn't miss anything. I did.

#6 was supposed to sport a 15 instead of a 16. Will edit that now.

This isn't so much an effort to curb SAD characters as an effort to help MAD ones.

I skimmed the Steelforged Items rule, and it looks like the 3.5 rule from the MIC where you can add bonuses to existing items, right? I actually do like that rule. Will have to read it properly when I have the time. Thanks!

[heavyfuel]

I think if you're going to use arrays you need more choices. For example, Barbarians and melee Fighters generally want two good stats (Str and Con) and don't care much about the rest, there's not an option on your list that would suit them. You can't start with a 17 or 15 in a stat at all (not a common choice, but I've seen it happen; one of the players in my current game chose Wis 17 for their Druid).

More generally, if a character is starting middle-aged, having all but one of the options being all even numbers is going to be really bad for them. I think you probably need at least twice as many arrays to suit all character concepts.

Fair enough. These ones I put were just from the top of my head. Will come up with more later.

[upho]

#6 was supposed to sport a 15 instead of a 16.Ah, that makes a lot more sense.


This isn't so much an effort to curb SAD characters as an effort to help MAD ones.I think I got that. I mentioned SAD/caster builds as a comparison to illustrate just how much greater the potential of MAD/martial builds have become, mostly thanks to the many strong related options published during the latter half of the PF1 era. It seems quite a lot of posters even here on GitP haven't really understood this yet, and for example still believe the fighter to be a mechanically mediocre class with a low op ceiling, or even that the differences between the Tiers are almost as great as they are in 3.5. IOW: a significant portion of the related issues and associated "truths" PF inherited from 3.5 are gone, and for example MAD builds may have very little to no need of additional stat boosts to catch up with the SAD builds in many games.


I skimmed the Steelforged Items rule, and it looks like the 3.5 rule from the MIC where you can add bonuses to existing items, right? I actually do like that rule. Will have to read it properly when I have the time. Thanks!They're basically the MIC rules, yes. Although IIRC, those restrict each type of enchant to a few specific item slots, while the Steelforge ones have no such restrictions and added armors, shields and weapons as possible target items, along with skill competence enchants.

[Psyren]

I have a slightly different suggestion that might prove helpful. Instead of giving your PCs big point buy or generous arrays to start with, why not just let them raise two stats instead of one as they level?

In my experience, what kills MAD characters isn't where they start, but by falling behind. Starting the game with two 16s in their key stats, or even with a 16 and a 14, isn't actually the problem for these classes; rather, the problem is one of those scores staying at 14 or 16 (or at best, 20-22 with gear) when the monsters they're fighting need those stats to be in the high 20s/low 30s.

This solves all the problems raised in the thread - it gives a big power boost to your MAD characters, it doesn't buff the SAD characters nearly as much (sure they can throw that extra point into Dex or Wis or Con, but doing so doesn't cause them to gain the same raw power that the MAD guy gets), and it allows you to have interesting/flawed characters at low levels who aren't prodigies with16s or 18s in everything.

Both 5e and Starfinder have taken this approach (with some minor differences), and I think it's a much more elegant solution.

[Malachite]

Starfinder takes it even further - Choose four stats to raise. If 16 or below gain +2, if 17 or above +1 (if I've remembered correctly)

Now your soldier can be strong, fast, tough and still have some time to read a book or two! =p

[Gruftzwerg]

Since the intent is the balance between builds, I would go for a different method:

1. Ask players what they intent to play and try to find out which Tier level the specific build has

2. T1-2 builds should get something like 18/16, 14, 12, 12, 10, 10

3. Other builds should get better stat arrays like 18/16, 16, 14, 14, 12, 10

To give an example why I think that this is a better approach:

Imagine an Ubercharger. He can be viewed as SAD, while other attributes still make him better/more playable. But if you go for tiers, an ubercharger is T3 at best and thus would get the better stat array. SAD and MAD aren't good standards to determine if a build needs DM fiat or not imho. And any attempt to approach this with any other new point buy method has still the same flaw that those build who are already strong still profit from it.
I mean, just because your are MAD doesn't mean that you don't like to max your main stat as a SAD character does. As such, point buy solutions tend to still favor SAD characters.

[ezekielraiden]

You could make a bunch of arrays you're fine with and tell players to pick one. e.g. some games at my table let players choose between

[18, 16, 14, 12, 10, 08]
[16, 16, 16, 14, 12, 10]
[16, 16, 16, 12, 12, 12]
[16, 16, 14, 14, 14, 14]

Folks typically chose the 1st or 2nd one

Y'know what, I really like this approach. It makes me very sad that, by standard 3.5e rules, the first array is the "weakest" of the set (being "only" 38 PB equivalent) whereas the last is the "strongest" of the set (being 44 PB equivalent). But I very much appreciate that this "penalizes" the person who min-maxes by giving them a net modifier of only +9, while the two middle ones are +12 and the last is +14.

Ironically, if I were choosing, I'd almost always take either the 3rd or 4th options for that very reason; I like well-rounded characters, and racial stat bonuses or other small optimization tricks can easily make a starting 16 work just fine. Ten levels later, you've got (at least) 26 extra skill points, 20 extra HP, and any saving throw you roll is just as good unless they've really pushed their min-maxing to the brink. It lets you relax about all the miscellany of optimization and just focus on the one or two things you want to be really good at.

Hence why my second favorite approach is what JNAProductions proposed: let people pick their stats. Ultimately, 2-3 points in secondary or tertiary stuff will rarely turn a character into a living legend, but it will often be more fun for the player. A Paladin doesn't need Int 18, but a Paladin that has Int 18 might be really interesting to play if it doesn't come at the cost of crappy Charisma. Etc.

[Fitz10019]

...In my experience, what kills MAD characters isn't where they start, but by falling behind.

I agree.

If you extend the point buy costs in the original pattern,


stat score
point buy cost
relative increase


16
10
+2


17
13
+3


18
16
+3


19
20
+4


20
24
+4


21
29
+5


22
34
+5


23
40
+6


24
46
+6



... you can see that with level 4's the free stat-boost, when a Wizard changes his Int 18 to 19, and a Paladin changes her Str to 17, the two characters now have different point-buy values. This disparity grows after every level-based stat boost.

In my houserules, level 4 is not +1 to any stat. It's a raise to the point-buy pool of +4. Anyone changing an 18 to a 19 is exhausting the entire increase and gets the same as without the houserule. Anyone changing a 16 to a 17 has a leftover point they can use on a low stat.

That said, Psyren's suggestion is more elegent, and math-free.

[Calthropstu]

Roll 4d6 drop lowest, reroll 1s.

Roll 6 sets of stats, lined up side by side, as the dice are rolled.

You get a 6x6 square of decent stats. Take diagonal, left-right or up-down. Each set of stats can only be used once.

Change the 4 stats directly in the center to 16. You have at least 6 different stat sets that contain at least 2 16s.

[RNightstalker]

Roll X D8's + 10. Roll 7, drop 1, Roll 8, drop 2, etc.