Chargen methods?

[Particle_Man]

For those playing, running or potentially running 3.5 campaigns, what methods do you use for generating ability scores? The default array? Point buy (how many points)? 4d6 six times, in any order? Something else (if so, what)?

[ahyangyi]

Point buy or roll with point buy as a fallback plan.

The points can vary, but if you go below 25 points (or 15 points for Pathfinder 1e), then your PCs are objectively worse than NPCs with class levels, and there needs to be a story or mechanical reason for that.

[Inevitability]

32 pts: low point buys ends up punishing martials and gishes a lot more than it punishes wizards and druids. You could comfortably get 18 in your primary casting stat from 22 pt buy on, but a paladin with a very modest 14/10/14/10/14/14 spread is already demanding 28 points.

[pabelfly]

4d6 drop one, but you get a free 18 as well.

[Jay R]

The two DMs I've played with used the same method:
18, 17, 16, 15, 14, 13, in any order, and then you can swap anything 2 for 1. [A wizard can drop STR by 2 to get +1 INT, for example.]

That seemed a bit too generous to me. [But I didn't turn it down.]

When I started running a game, I let them choose between 2 arrays:
18, 15, 14, 13, 12, and 9; or 17, 16, 14, 13, 11, and 10.
[I didn't tell them that at the end of the first session, they would get +1 in every stat.]

[Crake]

I use 35 point buy to encourage players to pick a 17 instead of an 18, which they can then boost at level 4, and have more even stats across the board.

I say it encourages a 17 because the only odd ability score point buys are either 17, 13,11 or 9, but you dont want to waste levelup boosts on a 13,11 or 9, and you also dont want to waste your point buy on a random odd stat that will never get bumped, so a 17 becomes a natural choice.

So you could do 17, 16, 12, 12, 10, 10 or 17, 14, 14, 12, 12, 10 or 17, 16, 16, 10, 8, 8

[One Step Two]

28 to 35 Point-Buy is my preferred method, depending on the scale of game I am running.

However, when my players ask to roll for stats, I use roll 4d6 drop 1, making them roll a number each in turn and put the results into a 6x6 grid, and pick 1 or 2 different lines for them to choose from. That way they get to roll, but I arbitrate the final results in order to make things fair across the board. I don't enjoy having players who get their ideas for characters torpedoed because of a poor run of dice luck.

[Crake]

I don't enjoy having players who get their ideas for characters torpedoed because of a poor run of dice luck.

I feel like, if you’re rolling dice, then you probably shouldn’t come to the table with a plan. Rolling in dnd is like drafting in card games, you need to be flexible and adapt to what you get

[One Step Two]

I feel like, if you’re rolling dice, then you probably shouldn’t come to the table with a plan. Rolling in dnd is like drafting in card games, you need to be flexible and adapt to what you get

It's a small hazard of dealing with the "old school" players I have at my table. They prefer it because that's how they've always done it, despite the disparity it can cause at lower levels. And they seem to forget the fact our old DM used to replace really bad rolls with better numbers to balance it out anyway making the exercise a little pointless.

I find it's a decent middle ground, they roll dice and get a few high moments as a group when anyone rolls 16+ to add to the grid. Then I usually pick out the array that has a rough analogue of the point levels I want them to start with. It's a win-win.

[Crake]

It's a small hazard of dealing with the "old school" players I have at my table. They prefer it because that's how they've always done it, despite the disparity it can cause at lower levels. And they seem to forget the fact our old DM used to replace really bad rolls with better numbers to balance it out anyway making the exercise a little pointless.

I find it's a decent middle ground, they roll dice and get a few high moments as a group when anyone rolls 16+ to add to the grid. Then I usually pick out the array that has a rough analogue of the point levels I want them to start with. It's a win-win.

I've always found an interesting middle ground is rolling 24d6, dropping the lowest 6, and then arranging the dice into sets of 3 for each of your stats. You can give everyone the same 18 dice to play around with, and everyone can be involved in rolling the initial 24 dice, so everyone's equally participating and on the same playing field, but there's still room for everyone to customize their ability scores within the context of what they have.

[Tzardok]

I prefer the standard 4d6, drop the lowest. One variant I've used often is "roll '4d6, drop the lowest' seven times, throw one out".

[lylsyly]

Our table is currently using two 18s, two 16s, and two 14s. But we always play gestalt and need those scores.

[holbita]

One table made of 4d6 drop lowest, if you don't like your results you can choose the standard array.

[One Step Two]

I've always found an interesting middle ground is rolling 24d6, dropping the lowest 6, and then arranging the dice into sets of 3 for each of your stats. You can give everyone the same 18 dice to play around with, and everyone can be involved in rolling the initial 24 dice, so everyone's equally participating and on the same playing field, but there's still room for everyone to customize their ability scores within the context of what they have.

Hah, consider this idea stolen for the next campaign! Thanks Crake :smallbiggrin:

[Powerdork]

I'm partial to an adaptation of the old Method V. Instead of telling people how every class is played in advance, I tell players to rank the ability scores from most important to least important for what they want to do. (So for instance, someone who wants to play a paladin might tell me Strength, Charisma, Constitution, Intelligence, Wisdom, Dexterity.) Then, I tell them for the most important stat to roll 9d6, discard the 6 lowest dice, and repeat that process for the next ability score with 1 fewer die rolled and discarded, and continue like that down the line (so 9d6k3 Str, 8d6k3 Cha, 7d6k3 Con, 6d6k3 Int, 5d5k3 Wis, 4d6k3 Dex). This happening before classes are committed to, as all dice rolls should be, so if they roll something out of expected values, they can repurpose, or commit to playing what nonsense they've been handed with their class of choice. If they're even pushing for rolling dice.

[RexDart]

4d6, drop the lowest, if you get two 1s, re-roll them.

Plus you have the option of asking the GMs' five year old to help you roll the dice, which so far has consistently resulted in results well above average for all.

[D+1]

For those playing, running or potentially running 3.5 campaigns, what methods do you use for generating ability scores? The default array? Point buy (how many points)? 4d6 six times, in any order? Something else (if so, what)?I have collected about 30 fairly distinct different methods for ability score generation - including just letting players choose what they want.

I STILL use 6 rolls of 4d6, dropping the lowest of the 4 dice; arrange as desired; two 15's minimum. I use that across multiple versions of the game. Nobody complains.

[Troacctid]

My big problem with rolling dice for stats is equity, or rather, the lack thereof. I don't like that some players just randomly have less resources than others. So whenever I run a game with rolled stats, I always stipulate that stats are rolled as a group. We roll some number of arrays, and those arrays are then available to all players.

I prefer point buy, though.

[satorian]

I prefer just letting players choose their own stats, if I'm confident they will be reasonable.

[blackwindbears]

I used to do 4d6 and let players have one reroll. Then I did high powered point buy. Then I did 4d6 and everyone gets some bonus points to catch them up towards the person that rolled best.

My games were arms races with the monster manual.

This year I started doing 3d6. IN ORDER.

Players complained about this. If you ask them which system they like best they pick the highest power one.

If you ask them what their favorite session was it's one with the low power characters. If you ask them what their favorite adventure was it's one with the low power characters.

If you ask them what their favorite character has been then it's more split, but it's like 60-40 to be a 3d6 character.

Ultimately, it depends on the game you run, but I really like the randomness for a West Marches style game.

[Particle_Man]

This year I started doing 3d6. IN ORDER.

This interests me. Do you have any "reroll a hopeless character" rule? Or is it truly "Let the dice fall where they may"?

Like if I roll Str 15, Dex 10, Con 7, Int 8, Wis 11, Chr 7 for a 3.5 character using your method, then that is what I go with, period?

Edit: Also are there any particular house rules or setting conventions that you use that might be relevant here?

[Powerdork]

I would suddenly like to remind everyone of vanilla 3.5's stipulations that you can't have less than +1 net pre-adjustment ability modifiers, and you can't have your highest score be less than a 14 pre-adjustment (both are circumstances where you're prompted to reroll), not to tell anyone that they're doing anything wrong, but to remind people that's how the game assumes you're playing, the reference point.

[Crake]

My big problem with rolling dice for stats is equity, or rather, the lack thereof. I don't like that some players just randomly have less resources than others. So whenever I run a game with rolled stats, I always stipulate that stats are rolled as a group. We roll some number of arrays, and those arrays are then available to all players.

I prefer point buy, though.

The only time lack of equity appears to be a problem is when players are competing with one another, rather than cooperating in a narrative roleplaying experience. I’ve played characters that have rolled well below the other players, but have still stolen the spotlight with their character and actions, its not all about the raw numbers.

However, if people are gonna be petty about it, point buy does eliminate that perception… to a degree. Some might say point buy favours SAD over MAD, for example, and still feel stiffed at the end of the day. But theres no pleasing those people, so I say just go with what you enjoy.

[blackwindbears]

This interests me. Do you have any "reroll a hopeless character" rule? Or is it truly "Let the dice fall where they may"?

Like if I roll Str 15, Dex 10, Con 7, Int 8, Wis 11, Chr 7 for a 3.5 character using your method, then that is what I go with, period?

Edit: Also are there any particular house rules or setting conventions that you use that might be relevant here?

This has been mostly used in a West Marches style game. Character generation is always level 1, and capped at 20 minutes.

I had a player roll 10, 4, 10, 9, 8, 10

That's the character they played. Made a barbarian.


I would suddenly like to remind everyone of vanilla 3.5's stipulations that you can't have less than +1 net pre-adjustment ability modifiers, and you can't have your highest score be less than a 14 pre-adjustment (both are circumstances where you're prompted to reroll), not to tell anyone that they're doing anything wrong, but to remind people that's how the game assumes you're playing, the reference point.

The game is pretty forgiving. In my personal experience, it tends to break more when you move *above* the elite array than when you move *below* it.

The "too good" statistical equivalent is a net +9 pre-adjustment.

[ahyangyi]

I honestly don't understand how does the barbarian work without a strength of at least 13 (and accessing Power Attack). At that point you might as well play a Commoner (and roleplay the illiteracy and bad mood)

[Crake]

I honestly don't understand how does the barbarian work without a strength of at least 13 (and accessing Power Attack). At that point you might as well play a Commoner (and roleplay the illiteracy and bad mood)

+4 strength while raging, if your DM is generous, theyd let you take PA, and just only be able to use it while raging.

[ahyangyi]

Indeed, that should work (though I can't imagine myself playing that)

[Thurbane]

Our most common is 4d6 (drop lowest) six times, arrange in any order, and can re-roll any one of these dice. If the end results comes out with terrible abilities (DM discretion) a complete re-roll is allowed.

[Cortillaen]

My favorite method is having everyone roll arrays (3d6, 4d6 drop lowest, whatever), pooling them, and everyone can pick whichever array they like.

A friend played in a game with a variation of this where the 5 players and GM all rolled 4d6DL, then arranged them into a 6x6 matrix. Each player could choose any line of 6, including vertical or diagonal, as their array and arrange them how they want. It produced wildly powerful results, of course, but that's not necessarily a bad thing.

[Remuko]

I've never had a chance to try it but I'd really like to play a game where the whole game, not just chargen, uses point buy. Expand the point cost for stats out "infinitely" and have every level up grant some number of points. this way lower stats could be shored up quickly for MAD characters while SAD ones can save up their points to keep pumping a single stat. I think it would be an interesting experiment if nothing else. the problem is figuring out how many points per level should be earned and how it should scale.

[Crake]

I've never had a chance to try it but I'd really like to play a game where the whole game, not just chargen, uses point buy. Expand the point cost for stats out "infinitely" and have every level up grant some number of points. this way lower stats could be shored up quickly for MAD characters while SAD ones can save up their points to keep pumping a single stat. I think it would be an interesting experiment if nothing else. the problem is figuring out how many points per level should be earned and how it should scale.

This is actually what I do, though it doesnt expand out infinitely, what I do is say “you can increase your highest stat by 1, or instead increase a number of lower stats by the number of point buy that your highest stat would have costed” since 18 caps out at costing 3 points, anything above 18 always goes up by 1 point, but below that you can maybe turn a 13 into a 14, and a 15 into a 16 instead of turning an 18 into a 19.

[Arkain]

As most here I have seen and tried various things over the years.

Rolling for stats has been wild, overall. In one PF1 group, a player ended up rolling 13, 13, 13, 16, 17, 17 (or point buy 45) which was so ludicrous in comparison that simply everybody got that. In a similar vein, there have also been cases of (usually low-ish) fixed arrays given out that everyone can arrange individually. Another time I rolled a 3.5 psychic warrior whose only stats above 5 were a 13 and a 15 and I had to "explain" for quite some time why I refused to just play that. When we rolled hit points later I ended up with just about 30 on a level 8 frontliner. It wasn't very fun, truth be said. Those were all 4d6 drop lowest if memory serves.
When players insisted on rolling attributes I've tried improving that by allowing 2-3 arrays to choose from, rolled with the usual 4d6 drop lowest. That turned out rather well, all things considered.
Still, I rather dislike rolling for character creation, not because it might be a weird competition for the greatest stats, hp pool or such , but because I tend to find the extremes can end up so absolutely frustrating that it's very tempting to just bend the rules and roll again, revise or whatever, no matter how much it was agreed upon not to do that. Or if push comes to shove, conveniently kill off the character and roll a new one, which isn't much better, really. Same reason I dislike rolling for hit points, the privilege of being worse than even average isn't something I'm very fond of, considering all the time, energy and emotional investment that come with the hobby. No matter how high or low quality, giving everybody the same stats at least circumvents this.
That said, I can be totally on board with it, if the system doesn't punish too harshly, like say Stars Without Number. Minor differences all around and the system assumes most characters won't have superbly awesome modifiers to anything at all. If the entire table is on board with what it entails, it can be immensely fun.

For point buy I've tried and been on the receiving end of many variants. For instance, a 1:1 point buy without increasing costs usually for about 78 total. I once tried 8 minimum 30 points to allocate up to18, for instance. Or 10 standard, 18 points, that sort of thing. Have been in a situation where it was a completely free 78 points (max 18). Last time I went with plain old PF1 point buy 25. My experience so far has been that most every player will create a somewhat balanced character that may not even differ much from traditional point buy, so maybe all these special variants are uncalled for. I did have the one exception to the rule who thought 18/18/18/8/8/8 was the way to go, but that's just something that may happen.
What it all has in common is going for an overall above average power level, which allows for both decently built characters from an optimization point of view and some flexbility along the line, so people don't feel pressured to for instance dump charisma on their fighter because they "really need" that +1 strength. I also wish to have a player character, who do tend to take on heroic roles, feel like they are just that tiny bit "more" than ordinary folks they might end up protecting. Often, seeing a couple higher numbers on a character sheet may help with that illusion (or completely shatter it, depending on how things go). However, I usually play with people who are incredibly un-optimized in 3.X and its variants all the time, so maybe that's all nonsense and high attributes truly break the campaign after all.

I also rather like starting at 3-5 for "low" level campaigns, because level 1 and to a degree 2 are often no fun at all with how swingy things can be. Also serves to reinforce the feeling of being more than the common city guard as an adventurer without trudging through a dozen convenient low level caves first. Max hit points all the time every time, if only because it means I can feel safer in not pulling punches or fudging numbers, as I've often had people who like danger (e.g. other systems where one can die easily), yet aren't fond of actual character deaths. Taking a significant chunk of hp off a character without worrying to maybe outright kill another who rolled poorly in the hit point department with the same attack tends to instill a sense of danger and seriousness that is usually just right.

I figure extraordinarily low power levels could work when there's a decent spin to it. For instance, starting with characters as children or apprentices working their way up to NPC and ultimately PC (class) status. If that's a goal, sure, why not. Or maybe for short stories/oneshots, if only to experiment a bit. Outside of wanting the players to feel "weak", I can't think of many good long term applications though. If anything, low attributes may invite more SAD characters, which is what I usually try to discourage, but that's a personal preference.
To borrow a bit from Remuko's idea, it might be interesting to have characters start at low power and maybe when they get their attribute boost every couple levels, they also get an actual point buy increase, to further cement certain milestones in their adventuring career. Like, start with a (in PF1 terms) low point buy of 10 or maybe average 15, then not just add a +1 at 4th, but a whole slew representing an uprgade to 20 point buy. I've experimented a bit with handing out attributes every 2 levels for instance (but no upgrading the same one twice in a row), in order to encourage a feeling of accomplishment beyond pumping everything into the "main" stat. Some players took the chance to have a secondary focus, others went all over the place for whatever they thought was cool at the time, mostly gettiong all attributes to even values and such, both of which I thought valid and enjoyable. However, this is going well beyond just char gen at this point.

[Xervous]

I've never had a chance to try it but I'd really like to play a game where the whole game, not just chargen, uses point buy. Expand the point cost for stats out "infinitely" and have every level up grant some number of points. this way lower stats could be shored up quickly for MAD characters while SAD ones can save up their points to keep pumping a single stat. I think it would be an interesting experiment if nothing else. the problem is figuring out how many points per level should be earned and how it should scale.

I’ve already done this to great success, though keep in mind it works best when you delete +stat items from existence and roll the values into progression. Racial bonuses become character creation cap adjustments, with the net delta of scores being applied as a modifier on point buy budget. Admittedly I never got around to solving proper valuation of racial bonuses from monstrous races so that’s a spot where this stub fails.

Pardon the lack of tables.

Point buy values at level 1,4,7,10,13,16 are 40,60,90,135,200,300.

Point buy costs progress as follows

8: 0
10: 2
12: 4
14: 6
16: 10
18: 16
20: 24
22: 36
24: 52
26: 72
28: 98
30: 130
32: 168
34: 214
36: 268

I remember fine tuning it for some breakpoints but I don’t have that campaign bible on hand. The biggest complaint was the feels-bad of just being shy of making some combination of scores work, but it definitely helped with diversifying investments beyond Main Stat Go Up. Just be ready for everyone having well rounded saves.

[ngilop]

. I don't enjoy having players who get their ideas for characters torpedoed because of a poor run of dice luck.

I think it is weird that people come to the game with just a single idea for a character so that a die roll can make or break a character being playable.


I have always done 4d6 drop the lowest, repeat 7 times.
I do feel the PCs should be a cut above the random barmaid or blacksmith but not overly so.

I have never liked point buy system, they invariably lead to min-maxing in my opinion (the above 35 point buy explanation from Crake says it better than I can) and to me a roll feels more organic. for lack of better term.

[blackwindbears]

I have always done 4d6 drop the lowest, repeat 7 times.
I do feel the PCs should be a cut above the random barmaid or blacksmith but not overly so.


I see this claim a lot, but there are so many things that make PCs "special" that feel more earned than the DM being generous with ability score points.

There's having a PC class, there's your level, there's the +1 sword you were cleverly able to notice under the brown mold.

It's also interesting to me the the 4d6 drop lowest 6 times is already a major jump over random NPCs (see the NPC array), but the inclination is always to be a little more generous than standard.

Character creation is one spot where system expectations get treated as a starting point rather than a target. Nobody thinks to be generous with weapon damage die sizes, or BAB. Sometimes it happens with treasure, but I generally see about as many PCs wind up with too little loot as too much in 3.5.

(Nothing against you in particular ngilop, most of my campaigns have biased generous with ability scores as well and 4d6 7x is not that much of a bump.)

[QuadraticGish]

I use roll 4d6 drop 1, making them roll a number each in turn and put the results into a 6x6 grid, and pick 1 or 2 different lines for them to choose from..

This is what I do, but I always have a point buy option set aside that's just slightly below average of the grid in the case that there are no good arrays.

[Crake]

I see this claim a lot, but there are so many things that make PCs "special" that feel more earned than the DM being generous with ability score points.

There's having a PC class, there's your level, there's the +1 sword you were cleverly able to notice under the brown mold.

It's also interesting to me the the 4d6 drop lowest 6 times is already a major jump over random NPCs (see the NPC array), but the inclination is always to be a little more generous than standard.

Character creation is one spot where system expectations get treated as a starting point rather than a target. Nobody thinks to be generous with weapon damage die sizes, or BAB. Sometimes it happens with treasure, but I generally see about as many PCs wind up with too little loot as too much in 3.5.

(Nothing against you in particular ngilop, most of my campaigns have biased generous with ability scores as well and 4d6 7x is not that much of a bump.)

I think this partially stems from DMs making NPCs (or players THINKING their DM is making NPCs) using the same creation methods as PCs are expected to be made with, and thus they erroneously think that the PC generation method should be stronger, when, in reality, NPCs only use the elite array when they have PC class levels, and NPC classed NPCs should have the standard or commoner array.

By the standard expectations, the PCs are already a good cut above more than 99% of the world by standard ability score generation methods, assuming you don't get atrocious luck. The re-roll criteria ensure this, as the standard array has no score above 13, and it's modifiers total up to +0, so by allowing re-rolls if none of your scores are above 14, or your total modifiers add up to less than +1, you're quite literally ensuring that players are better than the a regular commoner.

[ngilop]

So.. here is one that i just thought of.. Not sure if it will be any good or not.


The DM rolls 3d6 for each attribute.

The players they arrange the 3d6 roll however they want to, and have the option to roll 1 die for each attribute to swap out with.


That way there is a similar 'base' of sorts, for all players, but it allows for a nice difference in the attributes for the players.


I am not sure how that would work in actual use though.

[Gruftzwerg]

I like to use 18/16/14/14/12/10 for tables with people who have no clue about optimization (yeah I do play at such tables. Don't get mislead by my forum showcases^^).

It helps to compensate for their lack of system mastery when it comes to building characters.

And it helps MAD (multiple attribute dependency) builds to be more reliable while doing less for SAD (single attribute dependency) builds.

Imho being generous with ability scores helps mundanes and gishes more and may even give your players reasons to not go for a (much stronger) full caster build.

So, imho there is even an argument to give this super strong ability score array on tables who like to play optimized chars. Because on tables with "optimization" anything that makes your players less likely to play a T1 full caster build is a good thing imho...^^

[Crichton]

I like to use 18/16/14/14/12/10 for tables with people who have no clue about optimization (yeah I do play at such tables. Don't get mislead by my forum showcases^^).

It helps to compensate for their lack of system mastery when it comes to building characters.

And it helps MAD (multiple attribute dependency) builds to be more reliable while doing less for SAD (single attribute dependency) builds.

Imho being generous with ability scores helps mundanes and gishes more and may even give your players reasons to not go for a (much stronger) full caster build.

So, imho there is even an argument to give this super strong ability score array on tables who like to play optimized chars. Because on tables with "optimization" anything that makes your players less likely to play a T1 full caster build is a good thing imho...^^

I've used this exact array for this exact reason, and it tends to work out well. Alternately I've used some variation of the 4d6 drop lowest system, but usually with either/both of rolling 7 results and dropping the lowest, or rerolling 1s and 2s, or both.


As you say, being generous with a few extra points on chargen stats helps martials and non PO/TO optimizing players out way more than it does optimized T1 builds.


Lately, I often just say 'Pre-adjustments, you get one free 18, and you have to have one 8. The other 4 stats you roll 4d6 drop lowest for'