New idea for stat generation: PB *and* roll

[Eladrinblade]

I like rolling, but I don't like the extremes you get with rolling. I like point buy, but I feel there should be some randomness to it. So I did both:

How many points?
22 for elite array equivalent
25 for MAD or otherwise more comfortable characters
30 for "high power"

score/cost
14 / 10
13 / 8
12 / 6
11 / 5
10 / 4
9 / 3
8 / 2
7 / 1
6 / 0

THEN
add 1d4 to each score

THEN
add racial bonuses

You can control your minimums, but can't just pick high scores (though you still have a chance to get them). Elite array gives you at least one 15; you can guarantee yourself at least one 15 (or more) with this system.

What do you think?

[legomaster00156]

Monk 2/Cleric 18. Score 3+ extra points, get all of the cool Monk stuff, then get out of the class.

[Grod_The_Giant]

I don't like the idea of making rational choice THEN randomizing. I'd rather do something like "d4+6 to determine your starting array, then point buy from there."

[InvisibleBison]

I certainly like the idea of a hybrid roll/point buy system, but I'm not sure that this is the way I'd do it. Mostly because it greatly reduces your ability to control your stats, which is the main strength of point buy. I think I'd prefer a system where people roll to generate their base scores and then spend points to adjust them. That being said, this system seems acceptable to me, and I'd be willing to give it a try.

[Eladrinblade]

Monk 2/Cleric 18. Score 3+ extra points, get all of the cool Monk stuff, then get out of the class.

I wasn't implying that MAD classes get the extra points; everybody gets whatever amount the DM allows.


Mostly because it greatly reduces your ability to control your stats, which is the main strength of point buy. .

Well, you can guarantee yourself a perfectly playable stat where it matters, and you can keep yourself from having bad stats if you're willing to give up higher stats. The idea is you still roll, but nobody can justifiably be a crybaby and quit because they rolled poorly.

[Crake]

I don't like the idea of making rational choice THEN randomizing. I'd rather do something like "d4+6 to determine your starting array, then point buy from there."

I feel the same way, I'd rather know what I'm investing into, rather than investing and then getting lowballed

[Krazzman]

I said it on reddit, I'll say it here too: rolled stat generation for 3.5/Pathfinder should go die in a fire. It should only be used for oneshots where you just have a short little adventure where you are not shafted for 8 Months of real life until your character finally hits the bucket only to roll an even worse garbage array (that you still can't reroll) next to Demi-god McRollHigh.

I have been on both sides of this. I had a character with 3 18s. I had a character with less than Elite Array next to one with 15 as his lowest stat. Rolling for stats hinders you more to play what you want compared to point buy.

[Quertus]

Hmmm... Old-school D&D, you rolled your stats, then picked your class. Sometimes, those stats let you play the class you wanted; sometimes, they didn't.

So I'm thinking you can appeal to the most players by rolling your stats, then picking your point buy. But you can keep the exact same formula: roll a d4 for each stat, then choose how much to spend to add from 6 to 14 to each stat.

So, my suggestion would be


Roll 1d4 for each score

THEN purchase your bonus

bonus/cost
14 / 10
13 / 8
12 / 6
11 / 5
10 / 4
9 / 3
8 / 2
7 / 1
6 / 0

THEN
add racial bonuses

Of course, having already run a Wizard who reached double-digit levels before finally having an 18 int, I don't see this really stopping me from running a Wizard, if that's what I really want to do.


I said it on reddit, I'll say it here too: rolled stat generation for 3.5/Pathfinder should go die in a fire. It should only be used for oneshots where you just have a short little adventure where you are not shafted for 8 Months of real life until your character finally hits the bucket only to roll an even worse garbage array (that you still can't reroll) next to Demi-god McRollHigh.

I have been on both sides of this. I had a character with 3 18s. I had a character with less than Elite Array next to one with 15 as his lowest stat. Rolling for stats hinders you more to play what you want compared to point buy.

You've gotten to experience both sides of this, and still haven't learned to appreciate it? Sadness.

Btw, rolled stats is more, "come to the table with several ideas, and see if you can get any of them to fit with what you've rolled". And, sometimes, you have to come up with something entirely different from what you've planned, because the reality of your stats says otherwise. Consider it training to not be a railroading GM: you have to learn to not just plan for / force a single outcome.

[Telonius]

I generally take the opposite track. The game is all about playing a character you want to play. Good rules help you not overpower or outshine other characters while you're doing that; bad rules don't. Forcing someone into a role they don't particularly want, or creating a bigger imbalance in an already horribly balanced system, does not sound like a good rule. That's even leaving aside whether or not the party can field someone for all of the major roles.

I think there's nothing wrong with adding randomness to CharGen, as long as it doesn't prevent the person from playing what they want to play. For my houserules, everybody gets a free 18 in a stat of their choice. The Fighter can put it in Strength, the Wizard can put it in Intelligence; if they really want to, the Fighter can put it in Charisma and the Wizard can put it in Wisdom. It's up to the players. After that, each stat gets 4d6, highest 3; reroll 1's once; arrange as desired. I find this splits the difference reasonably. Nobody's bad at their own shtick. While there may be variance, it's in things that aren't as critical to the character concepts and party roles.

(This doesn't fix things for MAD classes like Paladin and Monk, but that's kind of a separate problem).

[Quertus]

I generally take the opposite track. The game is all about playing a character you want to play. Good rules help you not overpower or outshine other characters while you're doing that; bad rules don't. Forcing someone into a role they don't particularly want, or creating a bigger imbalance in an already horribly balanced system, does not sound like a good rule. That's even leaving aside whether or not the party can field someone for all of the major roles.

I think there's nothing wrong with adding randomness to CharGen, as long as it doesn't prevent the person from playing what they want to play. For my houserules, everybody gets a free 18 in a stat of their choice. The Fighter can put it in Strength, the Wizard can put it in Intelligence; if they really want to, the Fighter can put it in Charisma and the Wizard can put it in Wisdom. It's up to the players. After that, each stat gets 4d6, highest 3; reroll 1's once; arrange as desired. I find this splits the difference reasonably. Nobody's bad at their own shtick. While there may be variance, it's in things that aren't as critical to the character concepts and party roles.

(This doesn't fix things for MAD classes like Paladin and Monk, but that's kind of a separate problem).

Oh, I not only agree that this is a perfectly fine approach, but I know quite a few players who would prefer to use this method.

Personally, I find that the stats are a secondary or even tertiary thing. Player > character, obviously. But personality > stats. Personally, I like to "test drive" lots of different characters until I find one that "fits". To use 3e parlance, I need to "take a 20" on character creation. So, to me, using rolled stats is just a great benefit for the system / Arangee helping me to come up with different character ideas while building 19 characters I won't enjoy playing.

There are, of course, other approaches, like knowing what you want to play, or not starting with a personality at all. But those just aren't for me.

[ryu]

I generally take the opposite track. The game is all about playing a character you want to play. Good rules help you not overpower or outshine other characters while you're doing that; bad rules don't. Forcing someone into a role they don't particularly want, or creating a bigger imbalance in an already horribly balanced system, does not sound like a good rule. That's even leaving aside whether or not the party can field someone for all of the major roles.

I think there's nothing wrong with adding randomness to CharGen, as long as it doesn't prevent the person from playing what they want to play. For my houserules, everybody gets a free 18 in a stat of their choice. The Fighter can put it in Strength, the Wizard can put it in Intelligence; if they really want to, the Fighter can put it in Charisma and the Wizard can put it in Wisdom. It's up to the players. After that, each stat gets 4d6, highest 3; reroll 1's once; arrange as desired. I find this splits the difference reasonably. Nobody's bad at their own shtick. While there may be variance, it's in things that aren't as critical to the character concepts and party roles.

(This doesn't fix things for MAD classes like Paladin and Monk, but that's kind of a separate problem).

I mean I'm pretty sure that system will on average present significantly higher stats than normal point buy. How much more has to do with exactly what is meant by the 1s clause. Is it re-roll all 1s that come up in a single stat roll one time? re-roll any one dice that comes up 1 PER stat roll? I can't imagine it's re-roll all ones that come up once per stat roll. That would pretty handily give absurdly high stat generation. Do keep in mind that's before counting the free 18.

[lylsyly]

I said it on reddit, I'll say it here too: rolled stat generation for 3.5/Pathfinder should go die in a fire.

Depends on how you set up your rolls, 4d6 drop lowest isn't the only way. My home group uses 2d4+10. High powered? Yes. Get what you want? Yes. You character concept MAD? So what.

[Zombimode]

I like rolling, but I don't like the extremes you get with rolling. I like point buy, but I feel there should be some randomness to it. So I did both:

[...]

What do you think?

The thing is, people how don't like rolling for stats (or any other kind of randomness in character generation) won't like your method either because it still contains the elemenent they don't like.

[AnimeTheCat]

I like the idea of determining your minimums and then purchasing/rolling for stats. This is just a variation on what racial modifiers already do with point buy really, it just increases the likelihood that you'll really get what you want. That's good if that's what you want and I'm all for that.

I don't think that rolling stats is the absolute worst. I also am not saying that the only way to play is 3d6 in order. I personally like 4d6b3 in order and then pick class, race, etc. I like building a character from instant inspiration as opposed to approaching a setting/game with an idea in mind. I also just make characters using point buy on a whim all the time so if I ever want to try something I usually mechanically test run it as a DM before I play it as a player.

To each his/her own. I enjoy long running games with under statted characters because it makes me really get in to the character, but that's just how my brain works.

Good stat idea system, so long as it meets your desired outcomes. There isn't a "bad" system for determining stats, there are just systems that suite your desires and systems that don't.

[Fouredged Sword]

One thing i have done in the past is to have all the players roll scores and convert them to point buy totals. The player with the highest total sets the baseline. All other players add points to their rolls until they match him.

All the fun of random rolls without the problem of one or more people feeling slighted by bad rolls.

[Telonius]

I mean I'm pretty sure that system will on average present significantly higher stats than normal point buy. How much more has to do with exactly what is meant by the 1s clause. Is it re-roll all 1s that come up in a single stat roll one time? re-roll any one dice that comes up 1 PER stat roll? I can't imagine it's re-roll all ones that come up once per stat roll. That would pretty handily give absurdly high stat generation. Do keep in mind that's before counting the free 18.

Re-roll each one, once. So if you get 2-3-1-1, you re-roll the two ones; keep the result even if you roll another 1. I haven't worked out the stats, but it usually seems to get a result around 12-14 most of the time, with the odd 8 or 16. So yes, it generally does result higher than standard point buy. (Major villains get the same treatment, so it usually works out evenly).

[Psyren]

Roll for your Point Buy value: d10+15

[ryu]

Roll for your Point Buy value: d10+15

A method of rolling for points where the best possible result is equivalent to one of the lowest versions of point buy in somewhat common use? Wow. That's a sad state of affairs.

[Psyren]

A method of rolling for points where the best possible result is equivalent to one of the lowest versions of point buy in somewhat common use? Wow. That's a sad state of affairs.

I was referring to Pathfinder actually, 15 PB isn't bad there and 25 is considered epic. (PFS uses 20.)

[lylsyly]

Roll for your Point Buy value: d10+15

I could almost buythat if it was more along the lines of say 5d4+15 giving a range from 20 to 35, then everyone at the table gets the same value. It would kind of set the tone for the campaign IMHO.

[Xuldarinar]

You know, it occurs to me: What game version are we even discussing? I'd assume 3.5 based on the number totals, but it could be 3e, it could be Pathfinder...

[Psyren]

I could almost buythat if it was more along the lines of say 5d4+15 giving a range from 20 to 35, then everyone at the table gets the same value. It would kind of set the tone for the campaign IMHO.

Again, talking about PF here. 35 PB is way too much there.

Maybe 10+4d4 could work too though for less variation.

[Quertus]

One thing i have done in the past is to have all the players roll scores and convert them to point buy totals. The player with the highest total sets the baseline. All other players add points to their rolls until they match him.

All the fun of random rolls without the problem of one or more people feeling slighted by bad rolls.

But I wanted to play a Wizard! I rolled a 3 for Intelligence, and got the highest total in the party. :smallfrown:

[ryu]

But I wanted to play a Wizard! I rolled a 3 for Intelligence, and got the highest total in the party. :smallfrown:

I was kinda assuming they were including arrangement.

[Telonius]

You know, it occurs to me: What game version are we even discussing? I'd assume 3.5 based on the number totals, but it could be 3e, it could be Pathfinder...

3.5 here. (And more text so I can post a response).

[Fouredged Sword]

But I wanted to play a Wizard! I rolled a 3 for Intelligence, and got the highest total in the party. :smallfrown:

We generally roll 7 r4b3d6, drop lowest total, and let players asign scores in the order they want. Rolls under 8 count as 8. That 3 ends up ether strength or charisma as an 8.

My party also dislikes dump stats and enjoys high ability scores.

[CharonsHelper]

Hmmm... Old-school D&D, you rolled your stats, then picked your class. Sometimes, those stats let you play the class you wanted; sometimes, they didn't.

In older editions of D&D your ability scores had a much smaller impact upon play than they do in 3.x, so lower stats didn't totally gimp you.

Ex 1: you didn't even get any bonuses to STR until you had a 15-16. The biggest difference that 8-14 made was the armor you could wear, so for many characters it didn't matter much.

Ex 2: DCs didn't change with your casting stats

So random stats in the various iterations of D&D aren't really comparable.

Note: I'm a fan of stat arrays myself. They put everyone on equal footing & benefit MAD classes over SAD, and MAD classes need all the help they can get, while SAD classes (mostly 9-level casters) losing a couple DCs on their spells won't make me cry.

[Quertus]

We generally roll 7 r4b3d6, drop lowest total, and let players asign scores in the order they want. Rolls under 8 count as 8. That 3 ends up ether strength or charisma as an 8.

My party also dislikes dump stats and enjoys high ability scores.

Ah, ok. So, absolute worst case scenario, everyone's dice hate them, and everyone has straight 8s. If no-one beats me to it, I'll calculate the odds later tonight, either before or after I player this party. I think I'll call them the crazy eights. :smallwink:

[PhantasyPen]

I think I can say right now that this probably wouldn't work for my groups, given that most of us universally hate how point-buy systems force you to have at least one extremely low stat if you want to be competent at your job. Of course, the fact that our first DM was a killer DM has already been acknowledged to have colored our world-view of D&D pretty heavily.

[ryu]

I think I can say right now that this probably wouldn't work for my groups, given that most of us universally hate how point-buy systems force you to have at least one extremely low stat if you want to be competent at your job. Of course, the fact that our first DM was a killer DM has already been acknowledged to have colored our world-view of D&D pretty heavily.

Ah but on the other hand rolling makes NO guarantee of any non low stats. At least with a reasonable point buy you can force the ability to function. Hell lets assume you care to varying degrees about three stats with one being primary. 32 PB can get you two 18s if you dump everything else. Alternatively you can have one 18 and split the other into two very respectable but not insane scores. Now lets assume you only care about going as high as 16 for your primary competency which is still a high mark for rolling. That's plenty to be competent without dumping anything as hard as you can.

[PhantasyPen]

Ah but on the other hand rolling makes NO guarantee of any non low stats.

That depends on how you roll your ability scores, and in practice, I've never seen anyone in my usual group who rolled less than a 10.

[ryu]

That depends on how you roll your ability scores, and in practice, I've never seen anyone in my usual group who rolled less than a 10.

Depending on how many times you've rolled and you're talking about the standard method? That's silly.

[rs2excelsior]

I like the general idea, but I do think it'd be better to reverse the process--roll first then have a pool of points to edit those scores. Possibly depending on how your initial roll went.

Personally I like rolling stats, PB just seems too... detached, I guess. It feels like the character is more organic or natural that way. I've definitely gone into character creation without a set plan, rolling for my stats and seeing what kind of character fit. Then again, my group usually doesn't do super high OP/killer DM type games where bad rolls make -that- much of a difference. Plus, we will sometimes roll two sets (4d6b3 any order, like usual) and let you pick one or the other--gives you some control (do you take the one with the 18 for making a caster or the one with several 15/16 for a martial?) as well as some degree of insulation against bad rolls.

[PhantasyPen]

Depending on how many times you've rolled and you're talking about the standard method? That's silly.

4d6, reroll any 1's, arrange the result as you see fit. I've also gone with 1d10+8, rearrange results as you see fit.

[ryu]

4d6, reroll any 1's, arrange the result as you see fit. I've also gone with 1d10+8, rearrange results as you see fit.

And as we see here the second method suggests you'll get below ten literally ten percent of the time. That's six chances per player per character generated this way, and you should see a 9 roughly every ten or so stats rolled.

Assuming the 4D6 reroll 1s once is drop lowest... pretty sure you still see 9 or less ten percent or more of the time.

[PhantasyPen]

And as we see here the second method suggests you'll get below ten literally ten percent of the time. That's six chances per player per character generated this way, and you should see a 9 roughly every ten or so stats rolled.

Assuming the 4D6 reroll 1s once is drop lowest... pretty sure you still see 9 or less ten percent or more of the time.


This is D&D, statistics rarely affect the Will of the Dice.

[GameMaster_Phil]

How about sitting your group together and having the first player roll 4d6, drop lowest. Then his left neighbour rolls 4d6 drop lowest. Go around until you have 7 ability scores, drop the lowest of these as well. Everyone gets the same 6 rolled ability scores to freely assign to their character.

Avoids dump stats and perceived player imbalance. You can replace this rolling method with any other you like (2d4+10 or whatever).

Opinions?

[Pex]

It's better to have rolling first then the point buy so as not to have the inherent luck factor of dice rolling make the point buy irrelevant.

With bias, I like 27-25-23 method to combine point buy with dice rolling.

1) Roll 4d6b3 three times, minimum roll of 7 so if you roll a 5 it's a 7. These are your first three scores.

2) Choose one roll and subtract from 27, max 18 so no 27 - 7 = 20. It's an 18. 27 - 9 = 18. This is your 4th score.

3) Choose a second roll and subtract it from 25. 25 - 7 =18. This is your fifth score.

4) The third roll is subtracted from 23. This is your sixth score.

5) Arrange scores as desired and add +2 to one score, max 18.

6) Add racial modifiers (DM choice to allow above 18.)

5) & 6) can be done simultaneously or reverse order.

For step 1) you almost don't want to roll an 18.

Suppose I want to play a human paladin in Pathfinder

1) Rolls 12, 10, 12

2) 27 - 10 = 17

3) 25 - 12 - 13

4) 23 - 12 - 11

5) 17 12 12 10 11 13, add +2 to the 13 for 17 12 12 10 11 15

6) Add racial modifier +2 to CO

My final array is ST 17 DX 12 CO 14 IN 10 WI 11 CH 15.

Level 4 CH goes to 16. Level 8 ST goes to 18.

Now let's try a human wizard

1) Rolled a 4 becomes 7, 13, 10

2) 27 - 10 = 17

3) 25 - 7 = 18

4) 23 - 13 = 10

5) 7 17 13 18 10 10, not caring for the 7 so it gets the +2 for a 9

5) Apply racial modifier +2 to CO for a final array of ST 9 DX 17 CO 15 IN 18 WI 10 CH 10

18 Intelligence is good to last for a while, so level 4 CO goes to 16 and level 8 DX goes to 18.

Other people may have different choices for one or both characters.

[PhantasyPen]

How about sitting your group together and having the first player roll 4d6, drop lowest. Then his left neighbour rolls 4d6 drop lowest. Go around until you have 7 ability scores, drop the lowest of these as well. Everyone gets the same 6 rolled ability scores to freely assign to their character.

Avoids dump stats and perceived player imbalance. You can replace this rolling method with any other you like (2d4+10 or whatever).

Opinions?

This could be interesting, and considering how my group rolls I suspect that there'd always be at least one 18 in the mix there... Huh, I might try this for my upcoming game.

[martixy]

I said it on reddit, I'll say it here too: rolled stat generation for 3.5/Pathfinder should go die in a fire. It should only be used for oneshots where you just have a short little adventure where you are not shafted for 8 Months of real life until your character finally hits the bucket only to roll an even worse garbage array (that you still can't reroll) next to Demi-god McRollHigh.

I have been on both sides of this. I had a character with 3 18s. I had a character with less than Elite Array next to one with 15 as his lowest stat. Rolling for stats hinders you more to play what you want compared to point buy.

Spoken as a true probability layman.

Make chance work for you, not the other way around.

You are not bound to EVER ONLY use the method the game designers came up with. This isn't Harry Potter. If the generic method produces undesirable results, modify to taste.

Me, if I was to make things interesting I'd go with 6+1d4 base stats IN ORDER, then some standard point buy value.

The current method I use is (4d6b3, out of order)x2, pick one (with an additional set if you are ever as unlucky to roll truly abysmally).

This makes for a very good, organic spread spread of abilities between players and amongst dozens or so characters, we've had only one instance where a third set has been required(as is the nature of all probability - it isn't impossible, just improbable, but still bound to happen sooner or later).

[rigsmal]

Spoken as a true probability layman.

This is an odd comment. I research probability and still prefer PB.

What distinguishes preference is risk-aversion versus risk-acceptance, in microeconomic terms. These are personal properties w.r.t. one's own utility, so there's no such thing as 'rational choice' here.

[ryu]

This is an odd comment. I research probability and still prefer PB.

What distinguishes preference is risk-aversion versus risk-acceptance, in microeconomic terms. These are personal properties w.r.t. one's own utility, so there's no such thing as 'rational choice' here.

There's also the fact that in most cases it's directly more advantageous to have one 18 and perhaps one other good stat while nothing else matters than it is to have solid 14s or even 15s. It's not risk aversion to say that focused high scores are directly a better outcome in terms of relevant advantages in play.

[rigsmal]

There's also the fact that in most cases it's directly more advantageous to have one 18 and perhaps one other good stat while nothing else matters than it is to have solid 14s or even 15s. It's not risk aversion to say that focused high scores are directly a better outcome in terms of relevant advantages in play.

You're right, I misspoke. There is absolutely rational choice, but just because the expected net modifier on rolls is higher than PB doesn't mean that it's rational to choose rolls over PB. This is risk-averse utility, which is not by any means a bad thing.

[Pex]

I don't mind the concept of Point Buy; I care about its implementation. I like Pathfinder Point Buy though I will never player a paladin or monk at the 15 point level. 3E Point Buy has to be at least 28, 32 preferably. Anything lower I'm more likely to play a spellcaster only. Point Buy inherently hurts MAD classes; the lower the value the worse it gets. I loathe 5E Point Buy with a passion. I use it because that is what everyone I play with uses, but I still hate it.

Dice rolling does not guarantee a subjectively decent array for a MAD class, but at least you have a chance where as low value or poor implementation Point Buy guarantees you'll never get it. Funny thing I've advocated for dice rolling a long time, but for my current Pathfinder game I convinced the DM to use Point Buy. Pathfinder Point Buy really works for me.

[Krazzman]

You've gotten to experience both sides of this, and still haven't learned to appreciate it? Sadness.

Btw, rolled stats is more, "come to the table with several ideas, and see if you can get any of them to fit with what you've rolled". And, sometimes, you have to come up with something entirely different from what you've planned, because the reality of your stats says otherwise. Consider it training to not be a railroading GM: you have to learn to not just plan for / force a single outcome.

Having had the "joy" to play mainly 3.5 oneshots over the 6 years we played it... the ones I enjoyed most were the pb ones. In general I like to play the character I want to play for long stretches of time.


Ah, ok. So, absolute worst case scenario, everyone's dice hate them, and everyone has straight 8s. If no-one beats me to it, I'll calculate the odds later tonight, either before or after I player this party. I think I'll call them the crazy eights. :smallwink:

Should be 0 since you have no positive modifier and then have to reroll... but the chance of this happening is going to be slim


Spoken as a true probability layman.

Make chance work for you, not the other way around.

You are not bound to EVER ONLY use the method the game designers came up with. This isn't Harry Potter. If the generic method produces undesirable results, modify to taste..

Ok next time the dm says 4d6b3 I'll roll 8d6b3 because I can change it, right?

Rolling 36 times 4d6b3 in a matrix yields 14 possible arrays. One of them should be usable.

[martixy]

This is an odd comment. I research probability and still prefer PB.

What distinguishes preference is risk-aversion versus risk-acceptance, in microeconomic terms. These are personal properties w.r.t. one's own utility, so there's no such thing as 'rational choice' here.

I think of it like this: You can always tweak your rolling method to get the results you wish. Given that balance is equally served by either method, the question of PB vs rolling falls to subjective preference(or perception).

[Roog]

I think of it like this: You can always tweak your rolling method to get the results you wish.

The result I want includes the criteria that:
"a good or bad roll at a single point in time shouldn't determine how powerful one character is relative to the party for an entire campaign".

Please tell me how to tweak my rollling method to get this result.