
Bandw2 |

Bandw2 wrote:of course i've yet to have a player do a full caster well.That may be part of it. The wizard puts 18 in Int, plays a human for an additional +2, and dumps everything else except Dex and Con. Using the same point-buy, he's 10x more effective than the poor fighter, rogue, paladin, and monk.
yep, mine thought his first 2nd level spell should be whip of spiders.

Bandw2 |

bookrat wrote:Remember that thread where we all rolled 3d6 in order to see what we'd get?Hey! I remember that!
Quote:Remember how many people got straight 12s or 13s and declared those characters to be unplayable commoners?No, I remember no such thing. I remember the "unplayable commoners" topping out at 12-13, with multiple single-digit stats and a negative total modifier for the set. I remember anything with a double-digit CON score and at least one mental stat of 13 or higher getting a response of "Hey, I could totally play a [insertcasterhere]". Those are the kinds of things I remember from that thread.
I think your "12 CHA is so low" player is an anomaly, not a member of a larger trend.
mmmmm i remember more how he put it. half the people would play things with animal companions or a synth summoner though... :P

bookrat |
2 people marked this as a favorite. |

bookrat wrote:Remember that thread where we all rolled 3d6 in order to see what we'd get?Hey! I remember that!
Quote:Remember how many people got straight 12s or 13s and declared those characters to be unplayable commoners?No, I remember no such thing. I remember the "unplayable commoners" topping out at 12-13, with multiple single-digit stats and a negative total modifier for the set. I remember anything with a double-digit CON score and at least one mental stat of 13 or higher getting a response of "Hey, I could totally play a [insertcasterhere]". Those are the kinds of things I remember from that thread.
I think your "12 CHA is so low" player is an anomaly, not a member of a larger trend.
You're probably right. I did admit my error a few posts after that one.

voska66 |

Randomly rolled stats typically produce a character of average stats. For every super powered character you have gimped one but.
One the nice thing about randomly rolled stats is it kind of determines the character you will play. You end up with an interesting party some times. I saw one game where the players all rolled high INT an CHR and ended up being Wizard, Bard, Oracle, and Sorcerer for the the party.
On the flip side I like point buy because I don't have to roll stats that turn to be awesome when no one was around to verify I didn't cheat. So I can build characters via point build. I like that.

Triune |
3 people marked this as a favorite. |

The answer is pretty simple. People always assume they'll be lucky, and that they'll get to be Mr. Amazopants with the massive stats. It's the same reason people go to casinos.
To paraphrase what I heard someone once say on the subject: When four people roll for stats, you'll usually get Captain Amazing, two average Joes, and Mr. Retardo, and that is a stupid way to start a game.

Kirth Gersen |
6 people marked this as a favorite. |

To paraphrase what I heard someone once say on the subject: When four people roll for stats, you'll usually get Captain Amazing, two average Joes, and Mr. Retardo, and that is a stupid way to start a game.
On the flip side, with a 15-point buy, you'll usually get Dr. Strange and 3 average Joes, which is an equally stupid way to start a game.

Kobold Catgirl |
1 person marked this as a favorite. |

Hah, people calling 20 point buy low and I've made players use 10. HAH.
Anyways, I use point buy and average HP because the video gamer in me likes the act of building a character and making pieces fit together and tweaking little details. I like more levers to play with.
You monster. Did your mother drop you on your soul as a baby?
Point Buy is fun. So is rolling dice. There's this really strong pushback against the latter of late, though. Kinda sad—rolling can make for a lot of fun. Bad rolls aren't going to hold you back that badly in practice (though I do generally allow rerolls for drastically weak results) and you don't get to play characters with really low stats without rolling for it. One or two low rolls (like Rynjin's 5 from above) can really help to flesh a character out.
That's partially why I hate working with arrays, actually. Recently I was building a character who was supposed to be extremely stupid...except the lowest I could get was an 8 Intelligence and a 10 Wisdom. Blegh! Kinda a different story, though. :P
Also, rolled stats tend to be less optimized than point buys, so to some extent, they're usually less effective than a point buy of the same value that would have been needed to make them.
This is true. Not a bad thing, though.
EDIT: I wonder if we could design a "class-based Point Buy" system? Seems like a pretty logical step to take. I guess multiclassing complicates matters, though.

bookrat |
3 people marked this as a favorite. |

Triune wrote:To paraphrase what I heard someone once say on the subject: When four people roll for stats, you'll usually get Captain Amazing, two average Joes, and Mr. Retardo, and that is a stupid way to start a game.On the flip side, with a 15-point buy, you'll usually get Dr. Strange and 3 average Joes, which is an equally stupid way to start a game.
A few months back someone (TOZ, I think) made a suggestion that I really want to try: just have the players pick whatever stats that think they should have based on their character concept.

thejeff |
1 person marked this as a favorite. |
thejeff wrote:Also, rolled stats tend to be less optimized than point buys, so to some extent, they're usually less effective than a point buy of the same value that would have been needed to make them.This is true. Not a bad thing, though.
Not a bad thing, but it makes direct comparisons tricky.

Marius Castille |

In 1E and 2E, negative modifiers for most stats started at 6 or 7. Bonuses started accruing at 16 or higher. Basically, a fighter with Str 8 and a fighter with Str 15 both have the same chance to hit and damage. So even terrible stats were more funny than terrible.
Also, character generation was a lot quicker. Even if your PC perished, with just a few dice rolls, you can be right back in the game.

Scythia |
2 people marked this as a favorite. |

Scythia posted something in a thread in Gamer Talk that I'm going to use.
It's sort of a "semi random" rolling, 2d5+8 for each stat.
Seems neat, though it does still have the "problem" of the minimum stat being 10.
Another method I've used, if an extra bump is wanted: 3d5, drop lowest, +8 gives an even higher average.
I prefer rolled stats, but I know well the issue of poor rolls. So I found a method that preserves randomness and gives distinct stats, while insuring playability.

master arminas |

When I run a game where we roll, I have the players each roll four complete set of stats . . . then they can pick one.
4d6 ⇒ (1, 4, 3, 5) = 13
4d6 ⇒ (6, 6, 4, 4) = 20
4d6 ⇒ (2, 1, 2, 3) = 8
4d6 ⇒ (5, 2, 3, 3) = 13
4d6 ⇒ (4, 3, 4, 2) = 13
4d6 ⇒ (2, 3, 6, 3) = 14
Stat Array #1: 12, 16, 7, 11, 11, 12 (11 point buy)
4d6 ⇒ (1, 3, 2, 5) = 11
4d6 ⇒ (6, 1, 3, 1) = 11
4d6 ⇒ (1, 4, 3, 1) = 9
4d6 ⇒ (2, 2, 2, 6) = 12
4d6 ⇒ (4, 2, 3, 5) = 14
4d6 ⇒ (6, 4, 4, 2) = 16
Stat Array #2: 10, 10, 8, 10, 12, 14 (4 point buy)
4d6 ⇒ (3, 4, 2, 3) = 12
4d6 ⇒ (4, 4, 2, 3) = 13
4d6 ⇒ (2, 6, 1, 4) = 13
4d6 ⇒ (3, 4, 6, 2) = 15
4d6 ⇒ (2, 3, 5, 3) = 13
4d6 ⇒ (3, 3, 3, 4) = 13
Stat Array #3: 10, 11, 12, 13, 11, 10 (7 point buy)
4d6 ⇒ (5, 6, 5, 1) = 17
4d6 ⇒ (6, 4, 2, 1) = 13
4d6 ⇒ (2, 6, 5, 2) = 15
4d6 ⇒ (3, 5, 3, 3) = 14
4d6 ⇒ (6, 1, 3, 1) = 11
4d6 ⇒ (4, 3, 2, 4) = 13
Stat Array #4: 16, 12, 13, 11, 10, 11 (16 point buy)
Usually gives them at least ONE set of (good enough) stats that the player feels like he can play.
MA

Tectorman |

For a one-shot game with a character I'm only going to use then and there, I'm good for whichever, controlled or rolled. Rolled is random, and I'm stuck with whatever came out, but the guarantee of this character and those rolls not mattering for any length of time makes it better.
For a character that I'm expecting to use over multiple sessions, I greatly prefer some manner of point buy. Rolling those stats and keeping those rolls for the entirety of that character's existence makes about as much sense to me as rolling my first attack roll or reflex save and keeping that roll for the rest of the campaign.
"Oh, you rolled a nat 1. Guess you're never going to hit anything ever. At least you rolled a 16 on your Will save, so you'll automatically be shrugging off a lot of enchantments, I guess."
Rolls effect what they effect and the game moves on. Except with rolling stats, because the game never moves on from that, except via character death.
Almost the same thing with rolling hit dice, except I could totally get behind rolling those dice every day. Having different hp totals each day gives the fun of rolling hp and the fun of not being stuck forevermore with those rolls (kind of like how it was fun to roll a 2 on your Greataxe damage die and it was fun to not be stuck with that selfsame 2 the next time your Greataxe damage comes into play).
What would that represent in-universe? Getting various degrees of good bedrest or restless slumber. The lingering tendrils of magic and curses just now taking effect. Old injuries acting up.

kyrt-ryder |
1 person marked this as a favorite. |
One fun rolling method I saw on these boards once was rolling 6 sets of 4d6 drop the lowest, and arrange in a 6x6 grid pattern for the whole group.
Players can take any consecutive set of 6, be it a Column, a Row, or a Diagonal.
[As a way to slightly boost it, I've also seen it where an 18 is swapped from elsewhere in the grid into the center, providing 4 separate arrays with a guaranteed 18 and possibly more than one of them.]

![]() |
2 people marked this as a favorite. |

I like rolling stats for reasons that have mostly been covered: it adds a little surprise to character creation. We do use a generous method and adjust arrays upwards if they fall below a minimum level, so unusual strengths are more common than unusual weaknesses - like my Cha 13 druid, or a Str 12 summoner in the current game. We've never found the difference between characters' stat arrays to detract from our enjoyment of the game.
Something that I think has been alluded to and not stated outright:
It makes dump stats less controversial.
I've seen a lot of complaints on the forums about min-maxers dumping one or more stats in order to get their high stats higher. When you roll stats, however, those things aren't linked. You have high stats and low stats, you don't decide exactly how high or low they are, and it's expected you do what you can with the array you have. So there's less stress about the fighter with mental stats of 7/10/5.
Shadrayl of the Mountain wrote:Try 4d6, re-roll 1's, drop the lowest. You'll end up with something close to a 40 point buy....Because I hate weighted point buy with a passion, so I just came up with a roll method that tends to make medium to powerful characters.
(Also, rolling the standard 4d6, I tend to get characters that would be 25-35 points. Thus a 20-point buy always feels weak to me.)
That's what my group does. Except for the one time we did 3d6+d8, reroll 1s, drop lowest. Paladin got a pre-racial starting strength of 19 - and 18 Cha.

thegreenteagamer |

I roll 5d6, drop the lowest AND the highest, reroll 1s and 6s...do this nine times, discarding the highest, lowest, and average score, and then have the players pass those numbers clockwise around the table to prevent player cheating.
Then I laugh and tell them "just use point buy" because utilizing Rube Goldberg methods of dice rolling to cause some semblence of balance is ridiculous when point buy is easy, quick, and results in 100% equal potential for all players, leaving chance to the day to day stuff instead of your permanent defining features.
I get enough of chance and random genetics defining who I am in life, TYVM.

kyrt-ryder |
2 people marked this as a favorite. |
Then I laugh and tell them "just use point buy" because utilizing Rube Goldberg methods of dice rolling to cause some semblence of balance is ridiculous when point buy is easy, quick, and results in 100% equal potential for all players no matter what.
MAD and SAD classes coexisting in the same game says otherwise.

thegreenteagamer |

thegreenteagamer wrote:Then I laugh and tell them "just use point buy" because utilizing Rube Goldberg methods of dice rolling to cause some semblence of balance is ridiculous when point buy is easy, quick, and results in 100% equal potential for all players no matter what.MAD and SAD classes coexisting in the same game says otherwise.
You actually caught me mid edit, but...I said players, not characters. All PLAYERS have the same potential to choose any class, and in that method all COULD choose SAD classes.
Anyway, changing the method to rolling of any type doesn't fix that. Only a different standard for SAD and MAD does, whether that's MAD gets more points or better rolling rules, but it's not the overall choice that affects that, merely the choice to separate standards regardless of your method...so it's an unrelated point.

kyrt-ryder |
It sort of does, because rolling creates a distribution of stats that doesn't automatically favor the classes that only value one stat highly and con/dex a bit for survivability.
Not saying rolling is awesomesauce, because it's not, but it doesn't explicitly play to the SAD class's favor.
Myself I tend to prefer to hand out a simple array for my players to arrange as desired. And yes, typically it comes out to some very high point buy. [The last game I ran was 18, 16, 14, 12, 10, 8]

bookrat |
2 people marked this as a favorite. |

I roll 5d6, drop the lowest AND the highest, reroll 1s and 6s...do this nine times, discarding the highest, lowest, and average score, and then have the players pass those numbers clockwise around the table to prevent player cheating.
I'm going to do this, except change "average" to "median." ;)
Just for you.
Let's see here...
5d6 ⇒ (3, 5, 5, 5, 1) = 19 reroll the 1...
1d6 ⇒ 6 reroll the 6...
1d6 ⇒ 1 reroll the 1...
... Ok, I'm done with this.

GreyWolfLord |

Well, I know why my players sometimes choose to roll.
My standard array is around 8 points with a 13, 13, 12, 11, 10, 9 for the stats.
If I'm nice I'll give you a 13, 13, 12, 11, 10, 10.
They feel they can do better than that. This is for lower powered games though, and if they do choose to roll...it's 3d6 baby!
IT's amazing when they roll lower....their faces fall...and they suddenly realize my array isn't THAT bad.
OF course, if they all roll I make it a rule that they do not have to keep any ability of 8 or lower. In fact, sometimes I make it so they don't have to keep any score of 10 or lower.
IN addition, they can roll as much as they want on their first roll until they get a roll of at least 15 or higher.
Finally, if their total bonuses are +3 (or for higher powered +5) or less, they can reroll stats of their choice, starting with the lowest stat they have.
When you play like that...people tend to like to roll. It makes it so they don't have as much to lose if they roll.
And for lower powered games...well, already covered that with my standard array.

Scythia |

kyrt-ryder wrote:thegreenteagamer wrote:Then I laugh and tell them "just use point buy" because utilizing Rube Goldberg methods of dice rolling to cause some semblence of balance is ridiculous when point buy is easy, quick, and results in 100% equal potential for all players no matter what.MAD and SAD classes coexisting in the same game says otherwise.You actually caught me mid edit, but...I said players, not characters. All PLAYERS have the same potential to choose any class, and in that method all COULD choose SAD classes.
Anyway, changing the method to rolling of any type doesn't fix that. Only a different standard for SAD and MAD does, whether that's MAD gets more points or better rolling rules, but it's not the overall choice that affects that, merely the choice to separate standards regardless of your method...so it's an unrelated point.
Although purely anecdotal, my experiences would disagree with you. I didn't even know that SAD & MAD were a thing until I read these boards. Characters in games I've run have always had stats that allowed any class to be effective. Even monk.

Bluenose |
Kirth Gersen wrote:A few months back someone (TOZ, I think) made a suggestion that I really want to try: just have the players pick whatever stats they think they should have based on their character concept.Triune wrote:To paraphrase what I heard someone once say on the subject: When four people roll for stats, you'll usually get Captain Amazing, two average Joes, and Mr. Retardo, and that is a stupid way to start a game.On the flip side, with a 15-point buy, you'll usually get Dr. Strange and 3 average Joes, which is an equally stupid way to start a game.
That works well in some systems (and I'd include BECM D&D having done exactly that) where stats don't have such a major effect on character effectiveness, but I'm not sure it'd be a good policy for others including D&D 3.x and it's derivatives (perhaps M&M, thinking on it).
Bluenose wrote:Spend a game rolling nothing higher than seven, and tell me again how little part rolling dice plays.Scythia wrote:Except of course it isn't. Overwhelmingly in character creation/advancement the game is based on choosing what you want - which class to take, what skills and feats to have, and so on. Random rolls are the oddity in that process, rather than the norm.What's the point of rolling anything?
To inject randomness and keep everything from feeling boring/predictable.
The entire game is based on rolling dice.
I repeat, in character creation random rolls are an oddity. The rest of the game they're the norm, perhaps rightly so, but that bit of randomness in what's otherwise a process of design sticks out.

Arachnofiend |
1 person marked this as a favorite. |

Triune wrote:To paraphrase what I heard someone once say on the subject: When four people roll for stats, you'll usually get Captain Amazing, two average Joes, and Mr. Retardo, and that is a stupid way to start a game.On the flip side, with a 15-point buy, you'll usually get Dr. Strange and 3 average Joes, which is an equally stupid way to start a game.
I don't think even Paizo still runs a 15-point buy. The way I do it is 25 points, stats are capped at 18 post-racials. Wizards are left with a lot of points they can't put anywhere that's useful to them and Monks have the points they need to be effective.
Point Buy isn't perfect, but it's a hell of a lot better than random rolls.

Lakesidefantasy |
4 people marked this as a favorite. |

Do we have a thread just listing all the different ways to generate scores? I'm really digging a lot of these, especially "reroll lowest and highest scores".
Here are a list of methods I have collected over the years.
From rulebooks.
Classic 3 of 3d6
Standard 3 of 4d6
High Powered 3 of 5d6
Floating Reroll (DMG)
Organic (DMG)
Heroic 2 of 2d6 + 6
Dice Pool (CRB)
From Combined Stat-Rolling and Point-Buying. Where we were comparing hybrid methods.
Loremaster Method (w/o points)
Loremaster Method Variable Base
Loremaster-Spaulding Method
Steelfiredragon d6 Method
Steelfiredragon d8 Method
Spaulding Method (w/o points)
Karui Method 7 Point (w/o points)
Karui Method 6 Point (w/o points)
Purplefixer Method
From Graphing point buy vs. dice comparisons for stat generation. Where we were discussing various methods’ abilities to reduce the disparity problem.
Lurk3r 3d5+3 (reroll 1s)
Lurk3r 2d4+10
Glutton 4 lowest of 7d6
Glutton 5 lowest of 8d6
Kaisoku 3d6+10 (points)
Way of the Wicked 1d10+7
Arrsanguinus 5d4-2
Dice Point
Arengrey 3 of 4d6 27/26/25 (mirror)
Murph 9d2
Scythia 3d4+6? (reroll 1s and 2s)
Lakesidefantasy 2 of 3d6+6
Lakesidefantasy 3 of 4d4+6
Lakesidefantasy 4 of 5d3+6
Azran 6d3
Nobledrake ???
Manimal 1d8+8
Lakesidefantasy 15d2-12
Ilja 1d10+2d4
Ilja 1d12+2d3
Ilja 2d8+1d2
Ilja 3d3-1d10+10
Majuba 5 of 7d4-2
Majuba 5 of 7d6-6
Majuba 5 of 8d6-6
JohnF 3d5+1d4-1
JohnF 2d6+1d4+1d3-1
Ilja 1d8+1d6+1d4
1d8 + 8
1d12 + 6
2d4 + 10
Super Heroic 2 of 3d6 + 6
2nd and 3rd high and 3rd low of 6d6
3 lowest of 4d6
3 lowest of 10d6
1d8 + 1d6 + 1d4
1d10 + 2d4
1d12 + 2d3
2d8 + 1d2
3d4 + 6 reroll 1s and 2s
3d5 +3 reroll 1s
Fake Healer 3d6/2d8 +2/2d6 +6/2d4 +10
3d6 plus 10 point buy
4 lowest of 7d6
5 lowest of 8d6
1d6 + 1d5 + 1d4 + 1d3 + 1d2 – 2
5d3 +3 reroll 1s
5d4 – 2
5 of 7d6 -6
6d3
9d2
Some of these are obviously just wild experiments, but a lot of the crazy ones are actual methods people posted.
I have been using something I call the dice point method. For this system each player gets four dice points to distribute among their abilities. They can apply no more than three points to any single ability. Afterward they roll for each separate ability according to the number of points they spent on it. For instance if they spent 2 dice points on a particular ability they would set two of the dice to sixes and roll the third. If they spent three dice points in an ability then they set all three dice to sixes and get an 18!
I later combined the 1d8 + 1d6 + 1d4 method to the Dice Point method so that 3 points = 18, 2 points = 1d4 + 6 + 8, 1 point = 1d4 + 1d6 + 8, 0 points = 1d4 + 1d6 + 1d8. I like this method because if you spend one point in an ability then you are gauranteed it will not suck because the lowest you can roll 10 which is not below average. If you spend 2 points in an ability then you are guaranteed to excel at that score because the lowest you can roll is a 15.
Another advantage to this method is it is scalable. For higher powered games you can apply 6 or even 7 dice points to your scores. For lower power games you can use 3 dice points.
Finally this method absolutely eliminates dump scores while at the same time allowing a degree of customization because you have no ability to choose which scores will be low, only what will be at least average or at least high.

Lakesidefantasy |
1 person marked this as a favorite. |

Then I laugh and tell them "just use point buy" because utilizing Rube Goldberg methods of dice rolling to cause some semblence of balance is ridiculous when point buy is easy, quick, and results in 100% equal potential for all players, leaving chance to the day to day stuff instead of your permanent defining features.
Actually Point Buy is a pretty convoluted method of generating ability scores. Its not intuitive and pretty confusing to the uninitiated. It's also one of the factors why it takes hours to make character since you spend so much time weighing the consequences of placing each point.

chaoseffect |

Actually Point Buy is a pretty convoluted method of generating ability scores. Its not intuitive and pretty confusing to the uninitiated. It's also one of the factors why it takes hours to make character since you spend so much time weighing the consequences of placing each point.
It is basic math and uses a small and easily understandable table. If you know what the important ability scores are for your class it will take you maybe 5 minutes. Making Pathfinder characters is a labor intensive process for sure, but if you are spending most of it on generating ability scores then I find myself questioning your life choices.

thejeff |
1 person marked this as a favorite. |
Lakesidefantasy wrote:But still simpler than rolling x dice, dropping or rerolling all that meet conditions x, y, and z and then just having the DM adjust them up or down anyway if the RNG still dun goof'd.
It's certainly more complicated than just rolling three dice and adding them up.
That's certainly true. Some of these suggested methods really do get ridiculously complex. And often, in their attempt to avoid the low outliers, push the average much farther up than seems obvious at first glance.
3d6, in order. Simplest way.

chaoseffect |

"Convoluted"? Hardly.
It's a simple ratio you should have figured out by the time you've done it a couple of times.
The only one I don't know the value of off the top of my head is a 17, because who buys a 17?
Anything about a 16 doesn't give enough bang for your buck and a 16 shouldn't be purchased all willy nilly either because its cost effectiveness is borderline.

thorin001 |

I like watching many of the people who defend rolling lie about their motivations. Usually what they mean is that their pseudo-random rolling method gives them higher stats than what they think they will get from a point buy.
The justification of 'rolling prevents min-maxing' is hogwash unless you keep the stats in the order you rolled them. Putting your low roll in Cha is no less min-maxing than buying a low Cha.
The justification of 'you can't play your concept with a point buy' is purely stating that you want higher stats, but lack the honesty to say it outright.
If you like to roll everything, great, enjoy the game the way that is the most fun for you. If you like high stats (and who doesn't) then be honest about it.

bookrat |

Scythia wrote:Sure, and why not go the extra step like Warhammer and roll for class too.What's the point of rolling anything?
To inject randomness and keep everything from feeling boring/predictable.
The entire game is based on rolling dice.
I've done that. And race, too. It's how I ended up with my dwarf samurai (which worked out really well) and it's why I'm playing a paladin in one of my current games.
Now, I've never combined rolling for stats with rolling for race/class. That would certainly bring out some interesting results.
Edit: one of my group's campaigns started off with everyone rolling for random race and class. Turned out a lot of fun (we only played that campaign for a few months). Unfortunately, I don't remember what everyone got as this was about 4 years ago.

Barathos |

Because I hate weighted point buy with a passion, so I just came up with a roll method that tends to make medium to powerful characters.
(Also, rolling the standard 4d6, I tend to get characters that would be 25-35 points. Thus a 20-point buy always feels weak to me.)
Average 4d6-lowest is worth approximately 18.85 point buy, m80.

![]() |

thorin001 wrote:Scythia wrote:Sure, and why not go the extra step like Warhammer and roll for class too.What's the point of rolling anything?
To inject randomness and keep everything from feeling boring/predictable.
The entire game is based on rolling dice.
I've done that. And race, too. It's how I ended up with my dwarf samurai (which worked out really well) and it's why I'm playing a paladin in one of my current games.
Now, I've never combined rolling for stats with rolling for race/class. That would certainly bring out some interesting results.
Edit: one of my group's campaigns started off with everyone rolling for random race and class. Turned out a lot of fun (we only played that campaign for a few months). Unfortunately, I don't remember what everyone got as this was about 4 years ago.
Rolling for race and class helps me get over my urge to only choose either a) Human or b) the most absolutely optimal race for a class (often Aasimar or Tiefling), and I find it a lot of fun to try and challenge myself to play to the strengths of a particular race/class combo.

Saldiven |
Then I laugh and tell them "just use point buy" because utilizing Rube Goldberg methods of dice rolling to cause some semblence of balance is ridiculous when point buy is easy, quick, and results in 100% equal potential for all players, leaving chance to the day to day stuff instead of your permanent defining features.
Um...modified die rolling methods have nothing to do with "balance." I'm not sure where you got that assumption.
The point of modified rolling methods is to skew the average rolls higher. As pointed out before, it's really easy to get a modified die rolling that results in average numbers that equate to a point buy far higher than anything used by Pathfinder players.
And, in point of fact, point buy was developed to "cause some semblance of balance." (You even reference this yourself in your own comment, "results in 100% equal potential for all players.") Everyone has the same points to spend, while not everyone has the same luck with their dice.