How do you generate stats for new characters at your table?

[pabelfly]

I'm interested in seeing what stat generation methods are in place at other tables for new characters.

At the table I run, I do 4d6 drop lowest, but give the players a free 18 stat to start with. Most common stat methods I've played with were 32 or 36 point buy, and 4d6 drop lowest, and you get 7 stats rolled from that to drop the lowest roll from that.

[Particle_Man]

I'm playing in a game where it is roll 4d6 (drop lowest) 7 times, reroll 1s continually, and if you get quadruplicates (4 2s or whatever) in that process, then it becomes a score of 20 if you accept some unspecified trait from the DM (like the DM decides, usually more weird than bad, so worth it). Then pick the best 6 out of 7.

As DM I am getting rather fond of the default array, but it has been a while since I DM'd.

[arkangel111]

I saw a thing on tik tok I wanted to try. Everyone rolls 6 times and writes it down. Then you take turns and draw a straight line through the stats to choose which set you get. you draw lines vertical and diagonal but no set can be used twice though a single number could be used up to 3 times depending on it's placement. It's recommended you get at least a 6x6 grid like below.

16 12 11 8 6 18
14 13 14 12 10 8
13 13 16 15 5 12
8 15 11 8 18 11
12 11 12 14 8 6
11 16 10 15 5 15

clearly some rows won't be chosen if it can be helped but despite only 2 players rolling an 18 5 players could draw lines to get an 18 in 1 stat, granted a couple are making quite the sacrifice. I rolled these using 3d6 reroll 1s but you can still do 4d6b3 if you want slightly higher or any other method you like.

[Elvensilver]

Roll 4d6, drop lowest, then shuffle around or, depending on the DM, just swap two numbers.

If the total is doesn't come up too a modifier of at least +1 over all stats (such as only 60 points over all stats) you can roll anew.

As a result, there are quite varied stats in one party...

[Maat Mons]

My current DM likes to roll 3d6 (reroll 1s) 5 times, and a free 18. I’m more accustomed 4d6 (drop lowest) 6 times. I like 4d6 (drop lowest) better than 3d6 (rerolls 1s). The former is more likely to give you some high stats, with the caveat of also being more likely to get some low stats. The latter tends to give more average stats, which I find boring. But at least my current DM gives a free 18, so I put that in my class’s key ability and hope for something decent to put in Con.

Ideally, I don’t like to include any randomness in character creation. I prefer predetermined arrays or point buy. I also prefer average HP (after 1st level) or just max HP at all levels. Next time I run a game, I’m planning to give everyone the ability score array 18, 16, 14, 12, 10, 8. That gives each ability score have a different modifier, so you don’t run into the issue that a 10 and an 11 are essentially the same. This makes every ability score assignment into a meaningful choice. You’re actually deciding which things you’re better at, not leaving a bunch of things as a tie. I also plan to give everyone max HP, because bad die rolls should hurt you in the short term, not stick with your character forever.

[Condé]

I don't.

More seriously, I mostly play on Roll20 and my DM had a macro like 4D6, drop any stat under 7. So you end up with not so terrible characters.
I'm a bit more old school and make my players rolls 4D6 six or seven times, I can't remember, and if you are not satisfied by what you get, you can make a new array of stats but you have to take it even if it is worse than the precedent. It's dumb, but it is what it is.


Funnily enough, as a player of 3.5, I would vastly prefer to use a point buy system... Why? Because I spend a lot of time making builds for my characters and want to play very specific things and would be a bit disappointed if I rolled poorly and end up not being able to play what I wanted to just because my stats are bad. It happened to me lastly in a pathfinder game, my stats where horrendous but not unplayable, so I made an oracle of life. I play a heal bot. It was the first time I did it, because I like healing in general but... It was the first and last time in my life I did that.
I just want my character to die.

Rolling stat is fun... For one shot or short campaign. If you want your players to feel invested and care about their characters or they have a really precise thing they want to play, having bad stats can be crippling and detrimental and it is horrible to have a player asking you to kill their character or if they can roll another character.

But I think I don't really like playing with arrays, like I did back in 2014-2015 when I played at the early stage of 5E. Sure it solve a lot of issue, but I think it's pretty boring. *cough*(Like 5E in general)*cough*

I would like to make a one shot with really oldschool fashion roll dicing, like 3D6 in order and play an intelligent fighter or a strong wizard.

[Biggus]

At the table I run, I do 4d6 drop lowest, but give the players a free 18 stat to start with. Most common stat methods I've played with were 32 or 36 point buy

My experience is very similar, last time I was a player the DM used 4d6 drop lowest, then change your worst stat to an 18. I'm currently using 32PB but had just decided today to increase this to 36 for the gestalt game I'm preparing to run.

I used to use a straight 4d6 drop lowest, but with a fairly generous policy about allowing the players to reroll if their stats were poor. I do miss the way it forces you to make some hard choices with your character sometimes, and means you get some more strange and colorful combinations of traits. Point buy characters get a bit samey after a while.

Conde does have a point about rolling being potentially problematic for a long campaign. If I used it again I'd probably let them reroll anything lower than a 6, or maybe use 5d6 drop lowest instead.

[Drelua]

I don't like dice rolls, can be imbalanced when some roll way better than others. I like a 20 point buy starting with 10s, so equivalent to a 32 point buy but with no scaling costs. I don't like how Pathfinder's default system makes an 18 cost almost all of your points, it encourages people to make characters with attributes that are mostly above average but not remarkable, or to dump all the stats they don't really need down to 7s, which I just find boring.

[Telonius]

I'm interested in seeing what stat generation methods are in place at other tables for new characters.

At the table I run, I do 4d6 drop lowest, but give the players a free 18 stat to start with. Most common stat methods I've played with were 32 or 36 point buy, and 4d6 drop lowest, and you get 7 stats rolled from that to drop the lowest roll from that.

My house rules: One free 18. For the rest, 4d6, drop lowest, reroll 1's, just once. Arrange the results however you want. Mulligan if less than +7 collective bonus.

[Gorthawar]

In the past we've used elite array or 25pt buy quite a bit. Currently we're using 28pt buy. Rolling stats always ends up with someone a bit disappointed IMHO.

[Thurbane]

4d6 six times, best 3 dice, place in any order, and re-roll one die at any point. We usually allow a full re-roll if abilities mods don't total +5 or more by the end.

[lylsyly]

At our table we usually play gestalt so MAD is a thing.

8+1d4+1d6 rolled 8 times. Can take a point off an odd score and add it to another odd score. Drop the lowest two and arrange to suit. As good as it sounds the dice are fickle beasts that should have their own monster entry. If you do not like what you roll then 32 point buy it is

[Quertus]

There’s a lot of ideas I’d love to steal, but the most common ones I’ve used include…
4d6k3 x6, assign as desired
4d6k3, in order, reroll one, swap two
Stat array
Choice [stat array // point buy = value (stat array) - X]
3d20k1, in order


And most roll systems have some minimum, like “net stat bonus +1 or greater”.

So, basically this:

Roll 4d6, drop lowest, then shuffle around or, depending on the DM, just swap two numbers.

If the total is doesn't come up too a modifier of at least +1 over all stats (such as only 60 points over all stats) you can roll anew.

As a result, there are quite varied stats in one party...


I’ll often use a stat array of [18,16,14,12,10,8] for “powerful” (not “high level”) NPCs.


Funnily enough, as a player of 3.5, I would vastly prefer to use a point buy system... Why? Because I spend a lot of time making builds for my characters and want to play very specific things and would be a bit disappointed if I rolled poorly and end up not being able to play what I wanted to just because my stats are bad. It happened to me lastly in a pathfinder game, my stats where horrendous but not unplayable, so I made an oracle of life. I play a heal bot. It was the first time I did it, because I like healing in general but... It was the first and last time in my life I did that.
I just want my character to die.

Rolling stat is fun... For one shot or short campaign. If you want your players to feel invested and care about their characters or they have a really precise thing they want to play, having bad stats can be crippling and detrimental and it is horrible to have a player asking you to kill their character or if they can roll another character.

I would like to make a one shot with really oldschool fashion roll dicing, like 3D6 in order and play an intelligent fighter or a strong wizard.


Conde does have a point about rolling being potentially problematic for a long campaign. If I used it again I'd probably let them reroll anything lower than a 6, or maybe use 5d6 drop lowest instead.

Hmmm… I’m kinda conflicted here. On the one hand, there’s lots of reasons one might not like the character that they ended up with - why not remove one such reason? OTOH, the design space you get from “designed, and fun” is so much smaller than the space of “random, and fun”, leading to even “parties of all fun characters” to be more samey and boring.

But you certainly can have a long-term, fun character created via random stats. Or an unfun character you designed. So I think that the focus should be more on replacing unfun characters, than any one method of trying to increase the probability of success (unless it’s “run an existing character” - that (plus letting the character become “existing” via one-shots) pretty well solves the problem, from top to bottom).

[Vaern]

We usually just use point buy. It lets us make a bunch of character in our free time and bring whatever we want to play to the table without having to worry about anyone fudging their stats when no one can see their rolls.
When creating fresh characters on the spot at the table, though, we'll often do 4d6 and drop the lowest since it tends to be faster than deliberately distributing points for some people, and sometimes we'll reroll 1s on top of that to give our characters a bit of an extra boost if it seems necessary (for example, if we're starting a game with less than the recommended 4 players and our characters need to be a bit stronger to make up the difference).

[Eldonauran]

Depends what kind of game I am running. Usually its point-buy based on what power level I want to see. Usually 25points, per the standard generation method, or the elite array if people don't want to be bothered to calculate the points. If I'm running a more powerful/chaotic game, I have the players roll 4d4+4 (max 18) and let them reassigning any points over 18 to any other stat at a one to one ratio. I've done that one twice, for the gestalt games I run.

[Melcar]

I'm interested in seeing what stat generation methods are in place at other tables for new characters.

At the table I run, I do 4d6 drop lowest, but give the players a free 18 stat to start with. Most common stat methods I've played with were 32 or 36 point buy, and 4d6 drop lowest, and you get 7 stats rolled from that to drop the lowest roll from that.

We simply let people assign a total amount of points as they like. Usually something like 76-80 points…

[Aegis013]

My table's rule for stat generation is "write in your ability scores, no higher than 18 before modifiers".
This allows us to play gestalt combinations that might've otherwise been discarded due to needing too many points or being too MAD without concern.

[pabelfly]

We simply let people assign a total amount of points as they like. Usually something like 76-80 points…

So you could conceivably start with a stat array like 18/18/18/14/18/14 ?

[Remuko]

So you could conceivably start with a stat array like 18/18/18/14/18/14 ?

18+18+18+18+14+14 = 100 not 76-80.

[Quertus]

We simply let people assign a total amount of points as they like. Usually something like 76-80 points…


So you could conceivably start with a stat array like 18/18/18/14/18/14 ?

Thanks for doing the math in that - that’s what a 76 point point buy would look like, isn’t it?

Now I’ve got this silly idea, to see what players do is given such numbers, but (and here’s the twist) where “18” isn’t the limit - the costs keep on increasing, as high as you want (and can pay) for the stat to go: +4 to 20, +5 to 22, +6 to 24, etc. So a 24 base stat costs 46 points.

[pabelfly]

18+18+18+18+14+14 = 100 not 76-80.

That's a point buy of 76 points.

[pabelfly]

Thanks for doing the math in that - that’s what a 76 point point buy would look like, isn’t it?

Now I’ve got this silly idea, to see what players do is given such numbers, but (and here’s the twist) where “18” isn’t the limit - the costs keep on increasing, as high as you want (and can pay) for the stat to go: +4 to 20, +5 to 22, +6 to 24, etc. So a 24 base stat costs 46 points.

Come up with a few stat arrays:

24/16/16/12/12/12 - SAD Caster
22/18/18/12/12/10 - SAD Caster version 2
20/20/16/16/12/12 - Two high stats
18/18/18/18/14/14 - The "I want to play a Monk" stat array

[Fizban]

Random stats are unfair by definition, and point buy allows lopsided theorycrafting, so arrays it is. 25 point buy based arrays, because those are standard and I've experienced firsthand how much of a difference a giant pile of extra stats can make turning something that would have been fine into something ridiculous.

Elite Array (15/14/13/12/10/8)
Specialized Array (17/14/11/10/9/8)
"No Weaknesses" Array (13/13/12/12/11/11) -a gimmick, but if someone has a use for it I certainly won't say no.

I think more people should think about how their stat generation methods, which as seen here include many free 18's (seriously? I was not aware so many people were just adding those on top) and point buy values that allow an 18 while ensuring you have no penalties where you don't want them, affect their expectations of what is normal when the game was all but explicitly built on the Elite array.

[pabelfly]

I think more people should think about how their stat generation methods, which as seen here include many free 18's (seriously? I was not aware so many people were just adding those on top) and point buy values that allow an 18 while ensuring you have no penalties where you don't want them, affect their expectations of what is normal when the game was explicitly built on the Elite array.

It's a problem that's easily solved though - if the players are too OP you just give them stronger monsters to fight.

[Quertus]

I think more people should think about how their stat generation methods, which as seen here include many free 18's (seriously? I was not aware so many people were just adding those on top) and point buy values that allow an 18 while ensuring you have no penalties where you don't want them, affect their expectations of what is normal when the game was all but explicitly built on the Elite array.


It's a problem that's easily solved though - if the players are too OP you just give them stronger monsters to fight.

I want PCs to actually be able to do stuff. Giving them stats sufficient to ensure help them do stuff seems like a good step in that direction.

Also, “Balance to the table”. If you roll straight 18’s (or straight 13s, or even straight 3s (EDIT: and choose to keep it, as there’s free rerolls for “unplayable” characters) I guess), it’s on you to build a balanced character from that base.

I haven’t really seen much “free 18” (or “drop one stat to boost another”) in 3e, but both were fairly common in the 2e days, IME.

If Halls of the High King NPCs are any indication, I’ll wager Ed Greenwood’s table uses something like 1d4+14 for stat generation. Serially, 15,15,16,17,17,18? Gah!

[Ashiel]

I'm interested in seeing what stat generation methods are in place at other tables for new characters.

At the table I run, I do 4d6 drop lowest, but give the players a free 18 stat to start with. Most common stat methods I've played with were 32 or 36 point buy, and 4d6 drop lowest, and you get 7 stats rolled from that to drop the lowest roll from that.

Pathfinder Point Buy, 15 Points.

[False God]

I'm gonna be odd, but I let players pick whatever value they want for each stat between 3 and 20 to best represent their character.

[Quertus]

Pathfinder Point Buy, 15 Points.

As someone blissfully unaware of Pathfinder rules…that works out to what now?

[Ashiel]

As someone blissfully unaware of Pathfinder rules…that works out to what now?
Roughly equivalent to what you get with the Elite array.

[Darg]

4d6 drop the lowest. Players can default to the elite array if their rolls were bad.

[Sir Chuckles]

Roughly equivalent to what you get with the Elite array.

Not roughly. Exactly. 15/14/13/12/10/8 is a 25pb in 3.5e and a 15pb in PF1e. The Nonelite Array (13/12/11/10/9/8) is 15pb in 3.5e and a 3pb in PF1e. Pathfinder 1e lists 15-point buy as "Standard Fantasy" and 20 as "High Fantasy". Pathfinder Society (organized play) and Modules assume 20 points.

15 is lower than expected and can hurt some MAD classes but is mostly workable. Typically I end up sad about Skill Points or something similar. But it's a perfectly ok amount if the GM isn't throwing anything mechanically mean at you.

---

I've experimented with all manner of rolling, arrays, point buy, and even "3d6 flat in order" for one silly game. But I just keep coming back to point buy for Pathfinder and the Standard Array (which is the same numbers as the elite array) for 5e. It's expected, it's consistent, it allows pretty much any build, and only occasionally results in me getting fussy over skills.

But I've been looking at PF2e's lately and really appreciating it. The added choice and control, yet ultimately leading to an array, is really nice. I think if you added that choice to races rather than set numbers - like what 5.5e seems to be doing in a different way - I think it could end up with some neat results.

[Biggus]

I'm gonna be odd, but I let players pick whatever value they want for each stat between 3 and 20 to best represent their character.

Interesting. How does that work out in practice?

[Mordante]

I dislike dice rolling for stats. Too often I ended up with character with stats between 8 and 14. Having stats below 10 is something I normally want for all my characters. However having multiple stats below 10 sucks

32 points buy seems the best way to go.

[Aegis013]

Interesting. How does that work out in practice?

While I'm not False God, I think I can comment here as my table uses the same method up to 18 before modifiers.

As a DM, when I used to do Point Buy, my players would usually get an 18 and try to devise who to make their character extremely SAD around that 18, which meant a lot of classes and concepts were just out of consideration because there weren't enough stats to make the concept work. This method means those classes/concepts are still in consideration, so as a DM I haven't noticed any significant change in character power, but there's been more experimentation with a broader scope of character concepts, which I like.

[False God]

Interesting. How does that work out in practice?
Fairly well. It tends to give me an idea how powerful a game people want to play (I am usually playing with folks with at least a moderate amount of 3.5/pathfinder experience), and I have far far fewer retires/replacements/"accidental" deaths. Though I let people change out freely anyway, as long as the game is at a reasonable point to do so, or they can give me a good reason why the party would encounter this character at an un-reasonable point. I'm also fairly stingy with magic items, so in some ways the bigger scores make up for that.

I also have fewer issues with people who like MAD classes performing poorly, and people with SAD classes often put effort into areas their character normally wouldn't be real good at. Since SAD classes themselves rarely even use the other stats, it doesn't change them very much.

Even if someone writes down all 20s, the numbers are simply not big enough to cause me any trouble. I just throw higher challenges at the party.

Its not perfect, sometimes a new person takes it as a way to be a jerk to other people with a "You had the option to have all 20's, why didn't you!" But I typically play with adults, and again experienced 3.5 players, so this is rare.


While I'm not False God, I think I can comment here as my table uses the same method up to 18 before modifiers.

As a DM, when I used to do Point Buy, my players would usually get an 18 and try to devise who to make their character extremely SAD around that 18, which meant a lot of classes and concepts were just out of consideration because there weren't enough stats to make the concept work. This method means those classes/concepts are still in consideration, so as a DM I haven't noticed any significant change in character power, but there's been more experimentation with a broader scope of character concepts, which I like.

Yeah a lotta this.

And, in the end, I'd rather kill characters who were themselves actually a challenge than kill characters who really stood no chance simply by the bad luck of dice rolls. That's not fun for me, that's not fun for them.

[ciopo]

Thursday table : 32 Point buy with a quirk : odd values cost 0, but next even value costs double, basically giving those pesky "prerequisite X 13" the middle finger, and discouraging starting 18 "a lot" compared to starting 17, and with 15 costing the same as 14, it promotes spreading out the ASI

Tuesday long running campaign :"use this array the GM rolled for every PC ever, and some DMPC", she used 4d6 drop lowest, and the array is a lovely 16 15 15 14 14 13

Monday : pf point buy, I forgot if it's 20 or 25

Friday : 4d6 drop lowest OR point buy, roll first and if you don't like the array then point buy

[Particle_Man]

Has anyone experimented with a bidding system, or alternatively a system that gives different advantages/disadvantages for low/high stats?

[Goaty14]

Personally, I don't like systems that vary as much as 4d6b3 x6, as usual. I don't like trying to roll a monk with 12/12/10/10/9/9 (or whatever), nor do I like being the luckiest man alive with 17/15/15/15/15/15 (has actually happened) whilst I look down over my fellow player with ugly stats. That said, variation is good! Arrays rub me wrong because they feel a little too static for me, and it's a little toooo easy to predict where the other players put their prized 18 :smalltongue:

That said, yesterday my DM used the following system:
- The DM rolls 4d6b3 x6 per usual
- The DM puts all the "kept" dice into a pool, and players can draw from that pool to make their stats
- Small Rule: No more than 3 dice may be used for a single stat

Thus everybody has the same dice total with their stats (like an array), but can use it to total up in wildly different ways (unlike an array), and the table as a whole has varying levels of starting power from game-to-game (much like rolling).
Thoughts?

[Remuko]

That's a point buy of 76 points.

The guy he was quoting wasnt using point buy. he said they just assign scores with 76-80 points. meaning starting from 0 not 8 or 10. thats 100 points.

[Darg]

Has anyone experimented with a bidding system, or alternatively a system that gives different advantages/disadvantages for low/high stats?

You mean like targeting low saves/AoE damage to plink low health/enforcing carry weight rules/employing skill checks on characters using skills verbally trying to get around skill checks? No, I would never.

[Quertus]

Personally, I don't like systems that vary as much as 4d6b3 x6, as usual. I don't like trying to roll a monk with 12/12/10/10/9/9 (or whatever), nor do I like being the luckiest man alive with 17/15/15/15/15/15 (has actually happened) whilst I look down over my fellow player with ugly stats. That said, variation is good! Arrays rub me wrong because they feel a little too static for me, and it's a little toooo easy to predict where the other players put their prized 18 :smalltongue:

That said, yesterday my DM used the following system:
- The DM rolls 4d6b3 x6 per usual
- The DM puts all the "kept" dice into a pool, and players can draw from that pool to make their stats
- Small Rule: No more than 3 dice may be used for a single stat

Thus everybody has the same dice total with their stats (like an array), but can use it to total up in wildly different ways (unlike an array), and the table as a whole has varying levels of starting power from game-to-game (much like rolling).
Thoughts?

Dagnabbit, you included the “ No more than 3 dice may be used for a single stat” line. Ah, well. No challenging myself to see how many dice I can shovel into a single stat, then.

[Thurbane]

I think because a fair chunk of my group started on 1E, where rolling was the norm and no point-buy existed at the time (not officially, anyway), is a big part of why we prefer rolling.

I have seen the occasional extremely lucky roll (on a Warlock, of all things - from memory, lowest stat was 14, and there was a couple of 17s and an 18); I've also seen people roll total garbage. The latter is why all the DMs at my table allow scrapping awful roll sets, and starting over. No one would expect you to play 12/12/10/10/9/9 from the example above.

[Maat Mons]

Has anyone tried some sort of system where your ability scores are derived from your class? So, for example, taking a level in Wizard gives you a good Int score, taking a level in Fighter gives you a good Str score, et cetera? I think it might work if every class had a set of baseline ability scores, and multiclass characters took the best value for each one. Customization of ability scores would then consist of buying them up from the base values.

[Thurbane]

Has anyone tried some sort of system where your ability scores are derived from your class? So, for example, taking a level in Wizard gives you a good Int score, taking a level in Fighter gives you a good Str score, et cetera? I think it might work if every class had a set of baseline ability scores, and multiclass characters took the best value for each one. Customization of ability scores would then consist of buying them up from the base values.

There was something like this in the 1E UA, from memory. You rolled more dice, and kept best 3, for stats important to your class. 9d6 for you most important (as decided by the devs), down to 4d6 for least.

Yep, looked it up: Fighter, for instance, rolled Str 9d6, Int 3d6, Wis 5d6, Dex 7d6, Con 8d6, Cha 6d6 (and Comeliness 4d6, which was an optional ability score). I was forgetting the optional 7th ability was included in this system.

This would potentially produce some very powerful 3.5 characters. I'd probably just let the player set how many dice he wanted for each ability score, 9 though 4, and roll them in order.

[pabelfly]

Has anyone tried some sort of system where your ability scores are derived from your class? So, for example, taking a level in Wizard gives you a good Int score, taking a level in Fighter gives you a good Str score, et cetera? I think it might work if every class had a set of baseline ability scores, and multiclass characters took the best value for each one. Customization of ability scores would then consist of buying them up from the base values.

Pathfinder 2e has you generate a stat array based on your race and class. So a Dwarf gets +2 to WIS, CON and a stat of choice and Fighter gives a +2 to STR or DEX. Was a neat system.

[MesiDoomstalker]

I've been strongly considering just letting my players play with straight 18s (pre-racial). Mainly because in 3.P, you are going to get 16-18 starting in your main stat anyways and you are only going to use resources to bump that value higher. Every other stat is, at most, going to get +2 or +1 throughout the whole career. Maybe more in the mid to high teens.

Basically, for the 4 or 5 stats you aren't using, the value of a +4 is only relevant at levels 1-3 or so. After that, they aren't contributing enough to your overall numbers without further investment to be noticeable.

[Crake]

I'm a fan of doing 24d6 drop lowest 6, and arrange dice in sets of 3 as you desire. Gives you more freedom in how your abilities are calculated, slightly more powerful than 4d6k3 x6 because you get to drop the lowest 6 dice overall, rather than the lowest 1 per set of 4, which can sometimes result in feelsbad moments where you have to drop a 4+, while another set is like, three 1s and a 2.

[Biggus]

As a DM, when I used to do Point Buy, my players would usually get an 18 and try to devise who to make their character extremely SAD around that 18, which meant a lot of classes and concepts were just out of consideration because there weren't enough stats to make the concept work.

I haven't experienced the "must have an 18" thing so much, but I did move from up to 32PB for precisely this reason: SAD classes are perfectly workable as long as they can afford one good stat and one half-decent (usually for Con) but MAD classes suck horribly with low scores.



This method means those classes/concepts are still in consideration, so as a DM I haven't noticed any significant change in character power, but there's been more experimentation with a broader scope of character concepts, which I like.


Fairly well. It tends to give me an idea how powerful a game people want to play (I am usually playing with folks with at least a moderate amount of 3.5/pathfinder experience), and I have far far fewer retires/replacements/"accidental" deaths. Though I let people change out freely anyway, as long as the game is at a reasonable point to do so, or they can give me a good reason why the party would encounter this character at an un-reasonable point. I'm also fairly stingy with magic items, so in some ways the bigger scores make up for that.


Cool, I might give it a try. I like the idea of giving better scores but less items: many fantasy characters are highly dangerous without any items at all, and D&D as normally played doesn't really support that. I've occasionally contemplated a sort of half-way house between normal and Vow of Poverty where you get better starting stats and more stat increases as you level, but stat-boosting items and spells are rare or nonexistent, while still leaving in things which give you new abilities (flight, invisibility etc).


I've been strongly considering just letting my players play with straight 18s (pre-racial). Mainly because in 3.P, you are going to get 16-18 starting in your main stat anyways and you are only going to use resources to bump that value higher. Every other stat is, at most, going to get +2 or +1 throughout the whole career. Maybe more in the mid to high teens.

Basically, for the 4 or 5 stats you aren't using, the value of a +4 is only relevant at levels 1-3 or so. After that, they aren't contributing enough to your overall numbers without further investment to be noticeable.

I can see that's the case with some things like skills, but with others it seems like it would make a difference well beyond that. Having an 18 Con mean means you're going to have many more HPs all the way up, and with saving throws an extra 3 or 4 points would still be quite noticeable well past level 3 I would have thought?

[RNightstalker]

My dad came up with rolling a d10+8 back when you did the 3d6 rerolling 1's 2's deal. My current standard is d8+10, roll 7 keep 6. That is adaptable for what kind of game you want to have/join.

[Drakevarg]

Typically? 2d6+6, reroll 1s, one free 18. I know that leaves them on the high end, but I was kind of a killer DM in my earlier days, and a lot of my adaptions over the years have been focused on directly countering that.

Have been contemplating experimenting with something along the line of 2d6+4 and handing out ability score boosts every level instead of once every 4. I never liked how 3.5e treats basic physical and mental conditioning (outside of your backstory) as a literally superhuman effort, and I'd like to see what it's like when ability scores aren't essentially static outside of your primary stat.

[False God]

Cool, I might give it a try. I like the idea of giving better scores but less items: many fantasy characters are highly dangerous without any items at all, and D&D as normally played doesn't really support that. I've occasionally contemplated a sort of half-way house between normal and Vow of Poverty where you get better starting stats and more stat increases as you level, but stat-boosting items and spells are rare or nonexistent, while still leaving in things which give you new abilities (flight, invisibility etc).

I use a mix of Pathfinder and 5E for level progression:

Every odd level grants you a feat, but you may take a single +1 ASI instead.
Every 4th level grants you 2 ASI points to spend, or two feats.

So people who want big numbers can big numbers their hearts out, and people who want tricks and feats can do that too.

[Drakevarg]

I use a mix of Pathfinder and 5E for level progression:

Every odd level grants you a feat, but you may take a single +1 ASI instead.
Every 4th level grants you 2 ASI points to spend, or two feats.

So people who want big numbers can big numbers their hearts out, and people who want tricks and feats can do that too.

Ooh, I like that. Might steal it, at least partially, since the game does seem to value feats below attribute increases (plenty of classes have bonus feats, very few permanently increase ability scores). Either way it's a way to demonstrate self-improvement in parallel with class levels.

[Thurbane]

This would potentially produce some very powerful 3.5 characters. I'd probably just let the player set how many dice he wanted for each ability score, 9 though 4, and roll them in order.

Question for the math wizzes: if I did use this method (players nominate 9d6b3, 8d6b3, 7d6b3, 6d6b3, 5d6b3 and 4d6b3, and assign them to specific ability scores before rolling, what kind of number ranges would you expect this to produce?



I just rolled up a Beguiler using this method (with an online dice roller) just for giggles:

STR (4) 6 + 3 + 2 + 6 (-2) = 15
DEX (7) 6 + 2 + 5 + 3 + 5 + 2 + 1 (-1) (-2) (-2) (-3) = 16
CON (8) 2 + 6 + 1 + 2 + 5 + 6 + 5 + 6 (-1) (-2) (-2) (-5) (-5) = 18
INT (9) 6 + 3 + 2 + 2 + 2 + 5 + 4 + 2 + 2 (-2) (-2) (-2) (-2) (-2) (-3) = 15
WIS (5) 4 + 4 + 4 + 6 + 5 (-4) (-4) = 15
CHA (6) 5 + 5 + 3 + 1 + 3 + 5 (-1) (-3) (-3) = 15

[Rleonardh]

15
14
13
12
10
8
Roll 3d6 add those any way you please

[Maat Mons]

@Thurbane: Anydice is a great website. (https://anydice.com/program/24503)