
![]() |
1 person marked this as a favorite. |
As I've shoveled out heaps of snow in the driveway I've been thinking today about how the different character creation methods compare to each other, that is...
4d6 (drop lowest)
3.5 Point Buy
Pathfinder Point Buy
Going in from out of the cold and Googling I come up with 4d6, on average, yields 13's across the board. I thought, "Hmm... how unimpressive."
So then I go and compare all 13's on the two point buy systems:
3.5 method = 30 points
Pathfinder = 18 points
So that's quite a difference there! Of course the point buy methods scale differently. The standard point buy for 3.5 was 25 points, while for Pathfinder it is 15 points. So in terms of standards, both point buy methods are below the value of the 4d6 average.
In the last couple of days, itching to do some kind of game analysis, I plodded through zillions of iteration of Pathfinder Point buy spreads at the different power scales (10, 15, 20, 25).
After making all of these spreads, I added up the net values of the bonuses for each power scale to get some comparison of how they affect a character. The end result is that the most optimized spreads to yield as many attribute bonuses are:
10pb = +5 total
15pb = +7 total
20pb = +8 total
25pb = +10 total
If you factor in the "dump stat" effect, basically ignoring one dump stat because the player would likely succeed at being able to negate at least one stat's negative effects, the overall comparison becomes:
10pb = +6 total
15pb = +7 total
20pb = +9 total
25pb = +10 total
So a slight increase in optimization with 10 and 20 point buy. It's interesting that with all of the leverage that 25 point buy has, it couldn't increase itself at all even with one dump stat. I couldn't find any combination that would yield +11 ability score points.
The whole exercise came out of some older threads on how APL might be impacted by point buy selection. I was curious what the real numbers would yield, so I went through that laborious process.
The end result are those scores about. I guess it really depends on how you look at the impact those values can have in the game. A 15 point buy, the standard in the game is 3 points lower than a 25 point high fantasy scale. If you optimize for maximum value attribute points you'll be 3 points higher.
How much that is in terms of overall CR value is a bit tough to quantify. Likely just 30% of a 1 CR level. Not really enough to have you bump up the CR on encounters by a level, but still giving some edge.
Because it is kind of a slight boost, it does lend credence to the idea of having lower "tier" classes, like the Monk, be given higher point buy's to help even things out a bit compared to a Wizard. It's a slight nudge that isn't going to throw off the game in any meaningful way, but can help bolster some of the perceived weaknesses of the class.
As for if you are using 4d6 drop the lowest? That I'm not sure about. The probabilities are, inherently, chaotic.
I guess it might help to see the probabilities of the Heroic option of 2d6+6 compared to 4d6, or do the 28 dice pool method. Or perhaps another option where you do 4d6 drop the lowest, but also reroll 1s.
Ultimately, I'm just trying to see how all of these different approaches actually affect things at a more specific level, and not just eyeball it all.

udalrich |
1 person marked this as a favorite. |

The advantage of rolling is slightly better stats on average.
The disadvantages of rolling is that you probably won't get the stats you want, and you might get worse stats.
18-12-12-12-10-10 is a net +7, so in theory compares with a 15 point buy. It's going to be challenging to make a Paladin with that array. If you put the 18 into charisma, you can't hit and barely do any damage unless you are smiting. If you put the 18 into strength, then you lose on smiting, spells and lay on hands.
14-12-12-12-12-12 is also a net +7, but you are unlikely to see a point-buy wizard with an array like that. (Well, perhaps if you are building toward a concept and willing to sacrifice power for the concept. In which case the first array causes problems.)
If you are unlucky, you could roll 16-10-10-10-7-5. You can still make a decent wizard with that array (at least until he gets into melee), but it's rather hopeless as a paladin or a monk. It would not even be that good as a fighter.

Old Guy |
I don't know about anyone else, but I rather enjoy having an ability or two that are just horrendous. It also makes character creation a bit more interesting. But that is my two cents.
+1
I always roll for stats because it's more authentic. My method is to roll one hundred complete sets of stats and pick the one that fits my concept. This avoids the cookie-cutter aspect of point buys used by lesser roleplayers.
LOL
(I sincerely hope that was a joke)
![]() |

I always roll for stats because it's more authentic. My method is to roll one hundred complete sets of stats and pick the one that fits my concept. This avoids the cookie-cutter aspect of point buys used by lesser roleplayers.
Thanks, made me chuckle, if only because I recently had a player try that on me.
*WAIT* I need to try someone else' dice!
FWIW in my group everyone in the group rolled 3d6 and we took the best array and the whole group used it so she wasn't shafted with a horrid character.

![]() |
I always roll for stats because it's more authentic. My method is to roll one hundred complete sets of stats and pick the one that fits my concept. This avoids the cookie-cutter aspect of point buys used by lesser roleplayers.
Hey... watch it or the next game will be a 15 Point buy! ;)

Riggler |
1 person marked this as a favorite. |

Kortz wrote:I always roll for stats because it's more authentic. My method is to roll one hundred complete sets of stats and pick the one that fits my concept. This avoids the cookie-cutter aspect of point buys used by lesser roleplayers.Hey... watch it or the next game will be a 15 Point buy! ;)
15 point buy? Hmph. I must be the worst DM E-V-E-R. My poor players are forced to use a 12 point buy. *Gasp* *Shock* *Horror* How ever will they survive?

Blueluck |

FWIW in my group everyone in the group rolled 3d6 and we took the best array and the whole group used it so she wasn't shafted with a horrid character.
I don't like random character generation methods. If my players insist that they prefer to roll, that's the only method I ever allow - everyone rolls one set, I write them all on a piece of paper, and every player gets to use the best set.

Disciple of Sakura |

I used to go 4d6 drop lowest, default to 32 point if it's not what you're wanting. But, after having one player who just had horrendously good luck that routinely out performed everyone else in stats, I decided to combine a tiny bit of random with point buy. I'm currently running with 22 point buy, and then add 1d4 to two different abilities (each rolled independently) or 1d6 to one ability score. It's yielded slightly more high powered characters, but they're also much more rounded.

Ultradan |
1 person marked this as a favorite. |

I usually make my players roll three sets with 4d6 (drop the lowest die). Then they use the best of the three arrays and place the stats where they want as to accomodate the class they choose afterwards.
Yes, I make them choose their classes AFTER rolling. lol... If they don't have what it takes to be a Paladin, then play a cleric, or a fighter.
This method really creates all kinds of heroes... Yeah, your fighter only has a 15 strength score; But he'll persevere, get through those tough fights, kill the overlord at the end, and THAT's what'll make him a hero in the story.
In my games, I don't have a bunch of Conans and Gandalfs running around. They're simply average joes doing what's right.
Ultradan

![]() |

Normally my group uses the heroic method, but when I challenged them to use a 20 point buy they balked. "You just want us to play farmers!" they said. I'm planning on using the carrion crown ap next and I wanted to see if the ap's are really as tough as I hear they are. I'd also like to actually run the adventures without having to completely rewrite the encounters (the group is usually 6 or 7 strong) for an APL+2 group.
Perhaps they've been spoiled by their luck when rolling?
--Vrock n' roll the bones!

Mahorfeus |

Meh, I'm a little weird with the stat generation. Roll ten sets of 4d6, reroll any scores below 10, and keep the best 6 to be arranged any way you want. I find that this creates characters far above the average, but that just means I can throw APL+2 encounters at them without any reservations.
For NPCs: I always point buy the stats. For PCs, I MUCH prefer to roll the dice.
I find that I always end up using the default NPC scores in the back of the Core book. Not quite sure why.

![]() |
What happened to the "heroic ability score array" from 3rd Edition? ( 15 14 13 12 10 8 iirc)
Thanks for pointing that out, I had intended to add that in to the OP but forgot.
What little bit of interest is that the Elite Array changed between 3.5 and Pathfinder:
3.5 = (15, 14, 13, 12, 10, 8)
PF = (15, 14, 12, 11, 10, 8)
Under 3.5 point buy, the Elite Array had a value of 25, right on the money for standard point buy value for 3.5
For PF's Elite array, the PF value of point buy brings it in at a 13 point buy, just a little below the 15 point standard value. If you were to use the 3.5 array then it would come out to a 15 point buy. I'm not really sure if this is an issue of a typo on the part of Paizo, or if it was intentional that the Elite array for Pathfinder would be slightly weaker than standard point buy.
In terms of the Standard Array (13, 12, 11, 10, 9, 8) they are the same in both 3.5 and Pathfinder. It results in a 3.5 point buy of 15, which is pretty low-fantasy.
For PF, the point buy of the Standard Array is awful, amounting to a 3 point buy spread. Compared to even low-fantasy point buy of 10 points, lowly commoner NPCs just plain stink.

Gambit |

Honestly it depends what we are running, for homemade Forgotten Realms campaigns we usually roll, which ends up with higher than average stats. For AP's set in Golarion I use the following system.
A 30 point buy without being able to drop your stats below 10 to get more points, it works quite well for us, the PC's feel challeneged and the DM doesnt need to ramp things up much at all. Lets examine this by comparing 2 human wizards:
Human Wizard 15 Point Buy: Str 7, Dex 16, Con 12, Int 20, Wis 7, Cha 7
Human Wizard 30 Point Buy: Str 10, Dex 16, Con 13, Int 20, Wis 10, Cha 10
Almost the exact same for the "important" stats of the class, only exception is character number 2 isnt a weak, naive, social pariah. This is just using a SAD character as an example, who it really helps out are the MAD characters (who, lets face it, are the ones who need the most help). Though that approach is about min/maxing and just showing an example of comparing the two, we also have an additional rule that you cant have more than one 10, so our take on the Wizard would probably be something like:
Str 10, Dex 16, Con 14, Int 18, Wis 12, Cha 13
I like well rounded stats and characters.

![]() |

But that's just your building... Given 30 points, as a wizard not allowed dump you should be:
Str: 10 int: 19 wis: 12 dex: 14 con: 16 chr: 10
With all level-ups to int. That character just has no disadvantages. And legalizing dump stats would make the character -2 to Chr-skills (which he can make up with ranks by mid level) for the much-desired 20 int.
So higher point buys help SAD as much or more than MAD; by giving them their Max stat with the survivability of dex/con/wis

Benicio Del Espada |

I like well rounded stats and characters.
Me, too. A base assumption in my games is that the PCs are truly exceptional people, and above average in almost every way. They're the stars of the show, so they need to be smarter/stronger/better-looking than most people. We don't watch TV shows about people who don't have what it takes to overcome their challenges. That would be boring. Heroes are usually able to win their battles, or the show's over.
And, as you said, MAD characters can use some help to shine.

kenada |

I use a hybrid point buy where players roll an initial array and then buy up to a 15 point buy. One thing I've been experimenting with in my current Kingmaker game is giving the PCs an additional point buy point every level (and a bonus one at 4/8/12/etc) instead of the traditional ability score increases. I’ve not rigorously analyzed it, but what math I’ve done suggests that someone trying to focus on increasing a high primary score should only going to get four increases while someone increasing multiple scores should tend to get six or seven by the time they reach level 20. I also like that it allows players to diversify their characters post-creation (e.g., a Fighter with a starting Int of 10 can easily go down the Combat Expertise tree by spending three point buy points to increase his Int to 13).

Brian Bachman |

The real difference between rolling and point but is one of control.
The former method, even at its most generous, invites random chance into the character generation process. That appeals to those folks who probably aren't all that into optimization, and who don't come to the table with a specific concept in mind that they must play. It has the potential to produce more varied characters in terms of stat, and can produce either much more powerful characters (if you get lucky) or much less powerful ones (if the dice gods hump you). A powergamer would like the possibility of the first, but would probably whine horribly if he ended up with the second. An optimizer likely hates rolling, as it reduces his control over his character, and thus his ability to optimize. If balance between character stats is importna to you, you likely also hate rolling.
Point Buy, on the other hand, appeals heavily to the optimization crowd. The powergamer likes it, but only at higher point buys. In the hands of some players, it will produce pretty standard optimized builds for each character class. In the hands of others it will produce min-maxed monstrosities that are a challenge to DM (my experience with players who do this is that they count on lenient DMs who will avoid taking advantage of low stats early in the game, and will accuse DMs who do take advantage of it of being unfair). Better players will avoid both the cookie cutter and the min-max freak to produce good and varied characters no matter what the method. If everyone is roughly equal in optimization talent and desire, it will also produce characters who are well-balanced against each other, for those to whom that is important.
Me and mine, we roll. Always have and probably always will. But that's just us.

ddgon |
Sure Rolling and Point buy are close on average, but they are not as close as you guys seem to think. Roll 4d6 drop the lowest? what happens when you don't get above 10? Some people are ok playing sub par (as in NPCs with the NPC array of 3 10s and 3 11s are better) with other players at 2 18 and such. Now some of you might like that, that is fine. Some of us actualy want to be more then empty space though.
PointBuy alows for the whole group to have the same starting point. How well they optimize is their own issue, but they have the ability to start out equal at least. And it lowers the chance of cheeting. When rolling any 17 or 18 looks like your cheeting, and if the dice don't like you you might just decide to run into the bar and stab everyone you see till your char dies and reroll. Till you get all 18s. Will take a long long time but evenualy (read Maybe a million rolls) it will happen. With pointbuy you can simply start good.
I love that PFS uses the 20 point buy. Much easer for DMs to quickly check your stats and make sure you did it right, and yes you can get a 20 to start with, but you won't have much else. If you want 2 16 you can have that too. (not together with the 20). So over all I think that 20 is a good start. 25 means more points in dump stats.

![]() |
One other observation:
To just lay out the overall parameters of what is possible, if you had an array of (18, 18, 18, 18, 18, 18) that would be equal to 102 point buy in the Pathfinder system.
To go in the other direction, getting (3, 3, 3, 3, 3, 3) is a bit harder. You can't go below 7 in the point buy system, but if you extrapolate out, assuming each score below 7 is another -4 point buy points, then it would result in a -40 point buy.
So 102 to -40 is the potential for rolling.
If you went with the idea that every +10 worth of Attribute bonuses equaled another CR, then if you compared a 102 point buy to a 15 point buy you'd have:
15 point buy (+7 attribute bonuses) versus 102 point buy (+24 attribute bonuses) = +17 point difference. If you were to round to the nearest CR that would make these 1st level characters roughly like CR 3 creatures. If you wanted to just get a point buy to roughly be equal to CR 2 in comparison to the standard 15 point buy, then you'd have a pool of 55 points.
CR 1: 15
CR 2: 55
CR 3: 102
Of course all of that is just roughly approximations, more for getting some sense of power than being a very reliable set of numbers.

BigNorseWolf |

I do like the variety you get when you roll. I have a sorcerer with a moderate strength score and a decent intelligence: something i would NEVER play under a point-buy.
Pointbuy, especially the lower ones, even steer me away from certain classes. I would never try a monk under point buy, and CHA based classes become less appealing because it forces me to pump a stat that does nothing but help my spellcasting: whereas other classes get either will saves or skill points along with their spellcasting.
Pointbuy makes me look at races even more carefully. Halfling and gnome are right out.. no way i'm wasting a +2 to cha that could be pumping a stat that i need/want.
Druid's pets also become god;s. Since they're not affected by the point buy they'll often have the best stats in the party.

Gambit |

But that's just your building... Given 30 points, as a wizard not allowed dump you should be:
Str: 10 int: 19 wis: 12 dex: 14 con: 16 chr: 10
With all level-ups to int. That character just has no disadvantages. And legalizing dump stats would make the character -2 to Chr-skills (which he can make up with ranks by mid level) for the much-desired 20 int.
So higher point buys help SAD as much or more than MAD; by giving them their Max stat with the survivability of dex/con/wis
Sorry, I didnt clarify the additional rules, noone can have higher than an 18 to start, even after racial adjustments, so if you play an elven wizard, it only costs 10 points for that 18 int, but if you play an elven cleric, it costs you 17 points for your shiny 18 wis. Also you can only have one stat of 10, before racial adjustments, so no playing a Fighter who pumps all his stats into Str/Dex/Con and leaves his Int/Wis/Cha at 10.

Earthbeard |

For NPCs: I always point buy the stats. For PCs, I MUCH prefer to roll the dice.
Always How I've done it too, I find the group always prefer to roll for stats, it's part of the ritual of character creation, if I have to use point buy, I somehow always feel somewhat less than if I rolled, not due to any statistics, just on the chracter as a whole.

![]() |

Players can either use a standard 20-point PF buy or they can roll 4d6 drop low. In either scenario, any stat can receive a +2 bonus if the player buys me dinner. They can get a +2 to TWO different stats AND receive a bonus magic item at level 1 if they buy me a new hardcover rulebook that I don't already own.

brassbaboon |

I should have read this thread before asking my characters to build their characters for the PF campaign I am just starting.
I gave them the choice of a "standard fantasy" 15 point buy or rolling 4d6 drop one. Two players rolled, and two players did the point buy. The two who rolled got better overall stats. One got truly awesome stats.
However, I have long ago decided that starting stats are not nearly as important to game play as people make them out to be. Some of my most entertaining and enjoyable characters have had the most boring stats, and in one case truly awful stats (back in the "roll 3d6 six times in order of S, I, W, D, C, Ch"). My most awesomely fun character ever started with these stats:
S - 10
I - 16
W - 11
D - 12
C - 10
Ch - 13
He was a wizard, of course, or "Magic-User" as they were known at the time. In a point buy system that would be a 16 point buy, but since cha is a dump stat for wizards, it would have been exactly the same if the cha had been a 10, which would have been a 13 point buy.
As the DM of a new PF campaign I've been creating NPCs using the 15 point buy system and they are totally powerful enough to do anything they need to do.
I did have one character who rolled truly awesome stats. He had two 18s, a 17 and nothing lower than a 12. He wasn't nearly as much fun to play. He totally outclassed the rest of the party and to challenge him the DM had to push the envelope to the point that other party members were constantly in jeopardy of being killed.
In general I prefer characters I have to actually think to play.

![]() |

I like the idea of random stats, its just that either every player turns up with a great set of stats as they just happened to be lucky, or we roll them in front of everybody and someone ends up with a poor character.
Point buy is less exciting and certainly stops super sets of stats. Its fair, but you don't get superheroes. This is only good if you don't want superheroes...
Normally the characters develop independently of stats, but they can certainly help in defining characters early on. I had a die roll fighter in 3.5 Forgotten Realms with Jotenbrud. He was treated as large when it benefited him in manuvers, but he was only as strong as the cleric (16)... I had a (3.5) half-orc with good overall stats so no low charisma or intellegence and a 20 strength. Made a great character to play, but he stole a lot of the lime light from the other players.
In pathfinder i have always used the point buy and overall it has been good. The characters are sometimes flawed with 7 in a dump stat, and normally have atleast one good stat.
Its pretty bad turning up at a game with a "fairly" rolled character when every one else has been "lucky". Point buy stops this...

Preston Poulter |
Sure Rolling and Point buy are close on average, but they are not as close as you guys seem to think. Roll 4d6 drop the lowest? what happens when you don't get above 10? Some people are ok playing sub par (as in NPCs with the NPC array of 3 10s and 3 11s are better) with other players at 2 18 and such. Now some of you might like that, that is fine. Some of us actualy want to be more then empty space though.
As an advocate of 3D6 keep the order (termed Hand of God) by others, I am prepared to play a pathetic character. If a 10 is your highest stat, then put the racial +2 there and go with it. But I must say that it's extraordinarily unlikely to get an array of 6 stats with 10 being the highest for 4D6DL. I can't run the numbers right now, but it's probably less than 1% occurence.

brassbaboon |

As for stats not being important at low levels, I'd disagree.
Give a rogue extra con and he has the average hp of a fighter.
Give a fighter extra int and he has the skills of a cavalier, barbarian or ranger.
Strength can make up for base attack.In a point buy you make the choice.
In the end, unless you somehow manage to get truly miserable rolls (in which case the DM should let you roll again) or truly awesome rolls (two 18s, nothing less than a 12) then the vast middle range allows you to create completely viable, effective characters who do their assigned job just fine. DMs will adjust the campaign anyway, so if everyone has awesome stats, then you'll just fight stronger monsters or be bored.
To me the interesting part of the game has always been the role playing part, and in some cases its more fun to role play a reasonable character with strengths and weaknesses than an uber character that waltzes through the game.

brassbaboon |

Point buy is about balance.
Rolling dice is about variance.
They are two diametrically opposed approaches to creating characters. When the game was first introduced character variance was considered a strength of the game. If you read the original Player's Handbook it explains that some characters are supposed to be more gifted than others.
That's a concept that has been largely repudiated, for very valid reasons.

Bill Dunn |

Point buy is about balance.
Rolling dice is about variance.
They are two diametrically opposed approaches to creating characters. When the game was first introduced character variance was considered a strength of the game. If you read the original Player's Handbook it explains that some characters are supposed to be more gifted than others.
That's a concept that has been largely repudiated, for very valid reasons.
I'd say that rolling is about balance as well. I find it makes for a better balanced game when playing with both optimizers and non-optimizers. With optimizers, you can usually count on a wider difference between padded offensive stats and less useful dump stats than with dice rolling. Dice rolling can lead to it, but far less frequently.
In general, my experiences over 30 years, have convinced me that a rolled game ends up being fairer in the long run with diverse player types than a point buy one.

Pendagast |

It turns out that if you roll 4d6s, drop the low score, six times, you will generate an array where no number is higher than 10 4 times out of 10000.
For 3d6s, you will get an array where 10 is the highest score 1 time in 64.
I find 4d6 rolls turn out too many super characters, 3d6 reroll ones usually (9 out of 10) makes you a reasonable and playable character and it also usually means you do not have a 20 stat at 1st level, which is also desirable.

Valandil Ancalime |

I dislike rolling and am not a big fan of Point buy. The rolling process is fair, but the outcome can be anything but fair. Point buy is fair, but not very organic. I do like the organic nature of some randomness in stat creation. Stat arrays are ok, but when was the last time you saw a wizard not put their highest number in INT?
My current favorite method
1- Take 20 cards from 1-6
(totaling whatever power level the DM/group can agree on.
example: elite array of 15, 14, 13, 12, 10, 8 = 72; Could be (1-6)x3, 6, 3 = 72)
2- Set aside a 1 and a 6
3- randomly deal out 3 cards, IN ORDER, for str, dex...
4- add the 6 to a stat and then 1 to another stat
(18 maximum, lose any excess;
Before or after revealing the random cards depending on how cruel the DM wants to be)
5- add up the cards and adjust for race
It's organic (random, in order), but fair. The +6 will let you make a modest roll good if you really want to play a (insert class here). The +1 lets you turn an odd number even.