Proposal for new standard ability generation method

52 posts / 0 new
Last post
Since the first thread got completely overrun with unrelated tangents right from the start, I'm starting a new one, this time with a much clearer post on what my goals are here...

Overview:
1. The goal here is to provide a system for generating ability scores that both caters to the random roll and the point-buy options, and eliminates the need for the array option.  Why eliminate the array option?  Because in a balanced system between random roll and point-buy, it's not necessary.  The modifiers produced are always equal which was the goal of the array option.
2. The system is easy to understand and intuitive to use.  This is the reason that 1 point spent is 1 point of ability score.  Having changing values here is unintuitive, at least to someone that has never used such a system before.  It's also much easier if you can simply shuffle points around without having to worry about the conversion rate.  It's also easy to double check, as the total of the values beyond 10 should equal 12.  It's easy to modify too, by simply changing that 12 to a higher or lower value, as you like.
3. The system produces completely balanced results (thus the requirement that your modifers total +6) whether you use the random or point-buy option, so different players can use the same system in the same game.
4. The system produces heroic characters.  You'll notice that no scores go below 10, as quite simply, how heroic is it to be a total weakling or a clumsy oaf?  It's not, that's correct.

*Also note that even if this is the "standard" method of generating ability scores in 5E, there can still be other options like "classic method" (which is 3d6 either in order or arranged as you like), "boosted classic method" (which is 4d6, either in order or arranged as you like), "array method" (which is one of 3 different arrays, depending on the power-level desired), etc. - I am not proposing there not be other methods available in the books, it's your game once it hits your table so do whatever you want, really.  I'm talking about what's used at organized play sessions, what's the assumed "standard party" in the written scenarios, monster design, etc., which is also a method that can, of course, be used at your table if you like.

So, the player chooses either of the following:

Standard Point-Buy Method:
All of a character's ability scores start with a value of 10.  The player has 12 points to spend to improve their ability scores from 10.  Each point spent is an increase of 1 for the score.  The total of the ability score modifiers must also equal +6 after spending these points, not counting any bonuses from race or class.  The maximum value allowed in any score is 18, including any bonuses from race or class.

Standard Random Roll Method:
There will be a chart, which I'll not provide here - so use your imagination, which lists every possible combination of ways one could spend the 12 points on their 6 ability scores, matching the criteria set above.  The player rolls some dice and uses whichever set of numbers corresponds to that roll, effectively doing the math for him.


What you need to do is:
1. Decide if you can live with this as the standard method, even if you ignore the standard method and use one of the optional methods (or modular methods if you prefer that term) instead.  If not, why not?  And "because I said so," "because I prefer something else," and/or "because I don't like it" sort-of answers really are of no use to the thread, provide real, concrete details, please.
2. Are there glaring flaws with the system?  Or did I miss on any of the goals? Again, real, concrete details, please.
I really quite like it.
"4. The system produces heroic characters.  You'll notice that no scores go below 10, as quite simply, how heroic is it to be a total weakling or a clumsy oaf?  It's not, that's correct."

Sorry, you state that like it's a universal truth. It's not. Ask any 10 pre-4E D&D players about their favourite character ever and I guarantee 3 of them will tell you about one with a gimped stat that made them fun to roleplay.
Ask any 10 pre-4E D&D players about their favourite character ever and I guarantee 3 of them will tell you about one with a gimped stat that made them fun to roleplay.

What makes you think that 4E players will be any different? People did end up with stats below 10 in 4E, and they were exactly as much fun to roleplay.

As for the method, I think that it sort of misses one of the major draws of rolling randomly, the chance that you'll end up with some spectacularly above average. Without that chance, there's no point to rolling, no point in not just using the point buy to put your ability score right where you want them.

Why, yes, as a matter of fact I am the Unfailing Arbiter of All That Is Good Design (Even More So Than The Actual Developers) TM Speaking of things that were badly designed, please check out this thread for my Minotaur fix. What have the critics said, you ask? "If any of my players ask to play a Minotaur, I'm definitely offering this as an alternative to the official version." - EmpactWB "If I ever feel like playing a Minotaur I'll know where to look!" - Undrave "WoTC if you are reading this - please take this guy's advice." - Ferol_Debtor_of_Torm "Really full of win. A minotaur that is actually attractive for more than just melee classes." - Cpt_Micha Also, check out my recent GENASI variant! If you've ever wished that your Fire Genasi could actually set stuff on fire, your Water Genasi could actually swim, or your Wind Genasi could at least glide, then look no further. Finally, check out my OPTIONS FOR EVERYONE article, an effort to give unique support to the races that WotC keeps forgetting about. Includes new racial feature options for the Changeling, Deva, Githzerai, Gnoll, Gnome, Goliath, Half-Orc, Kalashtar, Minotaur, Shadar-Kai, Thri-Kreen, Warforged and more!
"4. The system produces heroic characters.  You'll notice that no scores go below 10, as quite simply, how heroic is it to be a total weakling or a clumsy oaf?  It's not, that's correct."

Sorry, you state that like it's a universal truth. It's not. Ask any 10 pre-4E D&D players about their favourite character ever and I guarantee 3 of them will tell you about one with a gimped stat that made them fun to roleplay.



Truth.
Roleplay produces heroic characters.
Higher ability scores produce mechanically-efficient characters.

That being said, I love the work that people are putting into their creative ventures with DDN on the horizon. I have seen some amazing ideas and inventions coming out of the forums lately. I do hope the devs look at these boards from time to time. I wish they were more like the White Wolf devs who frequently hang out on their boards and interact with the members.
"What makes you think that 4E players will be any different? People did end up with stats below 10 in 4E, and they were exactly as much fun to roleplay."

Wasn't meant as a slam to 4E players. It's only that 4E "bottomed out" at a single score of 8. Also, in 4E having bad stats could mechanically hurt you a lot worse, becuase the math is tighter and less forgiving.

Try playing a 6 Wisdom some time. It's remarkably liberating. *glee*


"What makes you think that 4E players will be any different? People did end up with stats below 10 in 4E, and they were exactly as much fun to roleplay."

Wasn't meant as a slam to 4E players. It's only that 4E "bottomed out" at a single score of 8. Also, in 4E having bad stats could mechanically hurt you a lot worse, becuase the math is tighter and less forgiving.

Try playing a 6 Wisdom some time. It's remarkably liberating. *glee*



Any of us that have played for a few years have been there. It is FUN to have a "6 WIS" or whatever. OTOH it gets old pretty fast to play a character that is basically entirely below par or can't support the concept you happen to be interested in. There's merit in letting the dice tell you what to do, but having an eternal -2 to-hit for 30 levels due to one bad die roll on the first day can eventually feel like it sucks. It COULD get you to work out some interesting concept, but it could also just be a big drag.

Here's another variation we used to use back in the day. Start with 3d6 allocated to each stat, and then add another 9d6, spread as you see fit amongst the various stats. Now roll the dice allocated to each stat, and accept those numbers. You can ALMOST guarantee a high number in 1 or 2 stats, or probably get a decent stat in all 6, and with some luck you can have a really above average character. You can also easily end up with a stat or two that are low. You get most of the advantage of the point systems but you leave in a fairly strong random factor. Sometimes the PC you end up with will be totally different from what you were aiming for. Best of all worlds IMHO.
That is not dead which may eternal lie
I generally call low stats Parody play...  you can roleplay as though you were very impaired but remember the paradigm the less competant you are Jarjar the more that providence protects you. In my sig I have a Joxer character he is an incompetant fighter well actually he isnt mechanically incompentant but a bit divers, however his flavor is entirely the incompetant one with almost only accidental successes.
   
  Creative Character Build Collection and The Magic of King's and Heros  also Can Martial Characters Fly? 

Improvisation in 4e: Fave 4E Improvisations - also Wrecans Guides to improvisation beyond page 42
The Non-combatant Adventurer (aka Princess build Warlord or LazyLord)
Reality is unrealistic - and even monkeys protest unfairness
Reflavoring the Fighter : The Wizard : The Swordmage - Creative Character Collection: Bloodwright (Darksun Character) 

At full hit points and still wounded to incapacitation? you are playing 1e.
By virtue of being a player your characters are the protagonists in a heroic fantasy game even at level one
"Wizards and Warriors need abilities with explicit effects for opposite reasons. With the wizard its because you need to create artificial limits on them, they have no natural ones and for the Warrior you need to grant permission to do awesome."

 

I really quite like it.


Thanks, I'm glad someone does and says so...

"4. The system produces heroic characters.  You'll notice that no scores go below 10, as quite simply, how heroic is it to be a total weakling or a clumsy oaf?  It's not, that's correct."

Sorry, you state that like it's a universal truth. It's not. Ask any 10 pre-4E D&D players about their favourite character ever and I guarantee 3 of them will tell you about one with a gimped stat that made them fun to roleplay.


It is a universal truth, or very close to it.  There's nothing heroic about not being able to do what your class pigeon holes you to do, being too stupid to speak, being unable to hold onto any object because of your 3 DEX, etc.

Not that it can't be fun, I'm sure for some people it can be, personally I'd find it funny (not the same as fun) at best.  And it's not the poor stat (the mechanics) that makes the character fun to play, it's the method in which the character is roleplayed (or role assumed) and the experiences had, IME.  Besides, you can always play your Fighter with an 18 INT as not too bright if you want to play that character despite his/her stat.  They'll just succeed more at skill checks and whatnot than someone with a 10 INT is all.

Truth.
Roleplay produces heroic characters.
Higher ability scores produce mechanically-efficient characters.


Indeed, this.

As for the method, I think that it sort of misses one of the major draws of rolling randomly, the chance that you'll end up with some spectacularly above average. Without that chance, there's no point to rolling, no point in not just using the point buy to put your ability score right where you want them.


Indeed, and that's what the other "non-standard" methods are for.  The reason for rolling here is really just for that feel or rolling for stats (even without the pros/cons), and/or because you want inspiration from the dice for the character and you also don't want to be gimped or godded by the randomness - basically, you're coming to the table having no idea what you want to play, so you just roll it and do whatever.
I'ts not bad but I would suggest making it so your prime stats cant go below 10 and the rest is fair game. (or make the minimum stat optional)
Have one attribute at 8 and keep everything else the same, and I believe it would work well.
Since DDN is so modular, why not just keep the character creation method modular as well. If a DM doesn't like his players having to deal with characters having single digit abilty scores, then he can homebrew a rule so that they can reroll or something. However, I dont think ALL campaigns need to be held to that. As others have pointed out, there is a challenge and great amount of fun to be had playing characters with a low stat or two. As a DM and as a player, I prefer random die rolls for stats and believe in a "let the dice fall where they may" style. It isn't for everybody, and I support others being able to choose THEIR prefered method.
While I like the simplicity, the reason for sliding values in most point buy systems is that the top 2 or 3 stat scores tend to be much more important than the lower 2 or 3.  An 18, 14, 10, 10, 10, 10 is likely to be more effective than a 12, 12, 12, 12, 12, 12.  Beyond this, the system (as I understand it) effectively procludes all humans since their minimum net modifier (after the 12 point buy, racial and class bonuses) is +8 and you state it must be +6.

Like others, I don't agree with your criteria (esp. #4).  To me, a heroic character must have some foibles.  A hero is a hero not because they are spectacular, but because they spectacularly overcome their limitations.

Finally, I think it is right to consider a "standard" stat generation method for purposes of module design, officially sanctioned events, etc.  But this method serves very different purposes than a method I would want for day to day play.  Factors like balance and creativity get weighted very differently depending on the purpose for which the characters are generated.   
perhaps you could give more points to the Fighter, Rogue and less points for the Wizard...
Here is a pure rolling system that I believe is a good approach.

1. Roll 4d6 drop lowest.
2. Total score and compare to a threshhold.
3. If equal or greater than threshhold goto 4.
    If less than threshhold, reroll the lowest score and repeat step 3.
4. Arrange scores as desired. 

So essentially you keep rerolling the lowest score until you hit the threshhold.  DMs can set the threshhold based upon how powerful they want their PCs to be.

Another option for mix/maxers is just give them one 18 and then do the rest of the scores as I suggested.

 
perhaps you could give more points to the Fighter, Rogue and less points for the Wizard...



All you need to do is make the primary then secondary abilities cost (exponentially) more than the rest of the abilities. Then the single-ability classes spend more to get a really effective primary score. The multiple ability classes can sacrifice a bit on the primary score, but have lots of points left over for nearly as high secondary and remaining ability scores.
This is the exactly same as a linear point buy table with a budget of 12, no matter how you try to present it differently the math is the same except the point buy table pre-counts the 18 to be worth 8 rather than you subtracting 10 to get 8.

10 0
11 1
12 2
13 3
14 4
15 5
16 6
17 7
18 8

The reason for the non-linear point buy table is as previous poster mentioned, it is to model the die rolls that the odds are against getting the high roll, and that higher rolls provide more benefit in the game.

The prior thread went off on tangents of people proclaiming flaws with other ways than their way before it settled into the idea that random rolls can be modified to conform to the point buy budget.  Doing this eliminates the percieved negatives (min-maxers vs. bad rolls) of both, which is an important idea that you have not considered in your revision.   Those who want to min-max or use bad rolls just opt out of using the combined rule.  The prior thread also lead to research in other systems, finding that 5e standard array was the 3e standard array which was an average 4d6 drop average roll rounded down, rather than normally rounded which means it is a few points lower that 'standard'

If doing budgeted rolls then putting higher cost 18's into the table is indeed not necessary, because they are already constrained by the roll and you budget your roll in the assigned order (preferred, random or straight) so the guy that always rolls three 18's has to lose all but one to the budget.  

So I would be OK with the point buy table being linear, the argument is the min-maxer is not going to constrain their odds with a roll anyways so let them do what they want, and that includes adding below average scores to the table and not costing the high 18's.  With the 4d6 drop method the high and low rolls are unlikely so adding them to the table is not much of a problem.  

However you cannot go too far with negative as they become required to meet the budget, that was the issue with the zero point buy idea is that you had to pair up highs with lows.   This is not based on my ideas being different, rather I implemented it into a random point buy spreadsheet to see if it would work, and ended up agreeing with those who said they did not like the deep negatives on principle.   In practice it did not work.


8 -2
9 -1
10 0
11 1
12 2
13 3
14 4
15 5
16 6
17 7
18 8

The problem then becomes how many negatives do you allow is 18 18 11 9 8 8 OK?   If you budget the roll it is very unlikely to get two 18's, you are basically rolling for your highest numbers, flipping them down if they are over budget, and the lower numbers become 10's or modified to fit the budget, so such a role is really only possible from the min-maxer that bypasses the roll.

And if people really have a problem with negatives, simply saying you subtract 10 from your ability works just as well if not better than saying subtract 8 if you increase the budget by 12.  This may work better with 5e since it seems more tolerant of 8's than 4e so I think they are presuming it is the base rather than a 10.

8 0
9 1
10 2
11 3
12 4
13 5
14 6
15 7
16 8
17 9
18 10

The table of standard arrays need to remain, these have always been there for quickly getting rolled up without messing with the die and math.   It should just be a d20 list of already rolled, budgeted arrays with the standard, a min-maxed, and a balanced one highlited.

If you are just posting for approval of your ideas rather than looking for community feedback improvements and considering other ideas you did not originally think of, then you best use the blog format rather than the forum.

I think a good point-buy system would cap stats at 16. So you can buy exactly what you want fit your character but spread out the points more, or you could roll and hope for a 17 or 18. 
Before posting, ask yourself WWWS: 
What Would Wrecan Say?

5 Minute Workday

My Webcomic
Updated Tue & Thur

 

@Jester. Im comfortable with no one starting level 1 with a score higher than 16. It is better for the overall math of the game.



A difficulty with the scores themselves is, the odd numbers are worth significantly less than the even scores.

Score: Base Value
18: 4
17: 3.2
16: 3
15: 2.2
14: 2
13: 1.2
12: 1
11: 0.2
10: 0
9: −0.8
8: −1
7: −1.8
6: −2


Note, conceivably one can keep the same ratio as above without the decimal points, but I would stick with the more transparent values above. The values above also work better with exponents. For the sake of comparison:

18: 20
17: 16
16: 15
15: 11 
14: 10
13: 6
12: 5
11: 1
10: 0
9: −4
8: −5
7: −9
6: −10



Again when using these numbers, the primary ability (namely the highest score) and the seconday ability (namely the second highest score) cost significantly more.

So, if an 18 normally costs 4 points, since this will be the highest score, it will cost 4 squared points: 4×4 = 16 points. Likewise, a 13 costs 1.2 points, but if the highest or second highest score, it costs 1.2×1.2 = 1.44 ≈ 1.4 points



Actually, the ideal values for the primary turns out to be to the power of ^2.3, while the secondary is ^2. But simply squaring both is probably workable enough, but if generating a choice of arrays, its better to use the more precise value for the primary.   
The argument that a 17 is worth less than a 18 is true because of the modifier step, but I am not sure how you determined that it is worth only .2 more than a 16.   Some value odds more than others, as they see the values it adds in skills and defense, or getting ability plusses later.

The budget is set by the values in the table based on the standard array, so what point buy is doing is saying how you can more or less vary the standard array and have the same character power.

linear has something going for it which is you don't need the table lookup at hand, you can just subtract from ability.  Of course you would provide the table for those who can't subtract without their calculator, but they still need to add.
The argument that a 17 is worth less than a 18 is true because of the modifier step, but I am not sure how you determined that it is worth only .2 more than a 16.   Some value odds more than others, as they see the values it adds in skills and defense, or getting ability plusses later.

The budget is set by the values in the table based on the standard array, so what point buy is doing is saying how you can more or less vary the standard array and have the same character power.

linear has something going for it which is you don't need the table lookup at hand, you can just subtract from ability.  Of course you would provide the table for those who can't subtract without their calculator, but they still need to add.



I did alota math in another thread, and the “.2” value for odd scores produced excellent arrays that really felt equal to eachother. Also unlike where I *always* buy even numbers with the 4e point-buy method, when instead using the .2, I found myself selecting odd scores about 50% of the time. I can say with confidence, to me, the odd scores seem to be worth exactly .2.

Indeed, in one sequence of same-cost arrays, I noticed one array did feel more powerful than the others, but when I doublechecked it I realized I had made a typo with a wrong number. When I corrected it, the array felt equal to the others.
I think a good point-buy system would cap stats at 16. So you can buy exactly what you want fit your character but spread out the points more, or you could roll and hope for a 17 or 18. 


That is what the standard array should be, WOTC rounded down rather than rounded thus the 16 average high score of 4d6drop 1 became a 15.  

So the average 4d6drop ability set rounded













16141312109

became this rounded down.













15141312108

So you have to decide which version you are using to establish the budget for your table.  The entire point of point buy is to ensure you can tilt or flatten the standard array and end up with a character of the same power.   4e had specialist, dual specialist and balanced as standard arrays.  

If you combine random rolls with point buy budgets, then you can only get the 18's if you roll it and it is not at the end of your budgeting order.  So it still makes it hard to get so you should leave them in the table.  Of course that depends on which dice method you use, if you want to use 2d6+6 18's are easier to get and so are 8's.

You should not rule that 18 is impossible to get, because then the one who does min/maxing of race/class bumps effectively gets it, so the table needs to allow it so you can get it without min/maxing race/class.   Yes it means the min-maxer can do both, but if you craft the system to completely contain them then you are making it unoptimal for those who do not min-max.
Actually, the ideal values for the primary turns out to be to the power of ^2.3, while the secondary is ^2. But simply squaring both is probably workable enough, but if generating a choice of arrays, its better to use the more precise value for the primary.   



ideal as in ... feels subjectively right to you?   Why be so objectively precise in the decimal points if it is based on subjective feels right?  I have no idea how you are justifying that power series on any other basis.  You could just as well use an accumulated array of modiifer scores, which is what the prior point buy systems are, that is what felt right to the devs (well they nerfed the highest scores a bit because it felt right).


For what its worth:

If players are using the Playtest Array, then the following three arrays feel about equally good.


16 11 11 11 11 7

15 14 13 12 10 8

14 14 14 11 11 10


Even for a Wizard character, I would be comfortable with using any of the three arrays.
Haldrik

That is a simple idea +/- on the standard array using ordered zero budgeting.  Use the rounded 4d6 drop standard array rather than the rounded down version, since it should be that anyways, that way it starts with 16.  You have to allow 18's because you are otherwise denying it to those who do not min-max their race/class.  But with rolling before point buy balancing you have to roll an 18 to maybe keep it.













16141312109


Lets try this I rolled using 4d6drop, which turns out to be an overpowered roll, sums to a +9.  The DM says we are not opting out of point buy budgeting, so I can't keep this roll.













16169111317












+2-4-1+3+8


The idea of balancing is do it in order this could be straight, random or preferred.   This first example is straight order, which means I am going to lose the 17 that I rolled last.   But I could do it preferred order and put the 17 up front.

1st step leave the 16 alone, I can't increase it as my budget started at 0, 2nd step the 16 is over budget so I drop it to the 14.  While I could have kept it, I don't want to drop a later number to make up for it.













16149111317












-4-1+3+8


next step the 9,11 put me at -5













16149111317












-4-1+3+8


Adding in the 13 I am at -2, so I have the running budget to increase it













16149111517












-4-1+5+8


But that means I have to lose the 17 to the default.













1614911159












-4-1+5



Same thing with preferred order

Again the standard array













16141312109


This time I am going to keep the first three













17161613119












+1+2+3+1+1


By dropping the 13 and 11 to bring it into budget, keeping the last one at 9.













171616889












+1+2+3-4-2


Note I cannot bump the 17 up to an 18 as since I budgeted my roll order, I can't change the 9 and 17 to be 8 and 18.   You can only increase a score if you kept/dropped a lower score that appeared earlier in the order.      Of course if I was opting out of the ordered rolls, I could min-max all I want, just like I could keep good/bad rolls by opting out of the budget step.


As others have pointed out, there is a challenge and great amount of fun to be had playing characters with a low stat or two.


And as I have pointed out, I completely disagree with the notion.  The challenge and fun doesn't come from the stats, it comes from the manner in which you roleplay the character and the experiences you have when you do so.  Your character could have all 18's and you could still play him the same way and have the same experiences - the stats don't figure into the equation.

While I like the simplicity, the reason for sliding values in most point buy systems is that the top 2 or 3 stat scores tend to be much more important than the lower 2 or 3.  An 18, 14, 10, 10, 10, 10 is likely to be more effective than a 12, 12, 12, 12, 12, 12.  Beyond this, the system (as I understand it) effectively procludes all humans since their minimum net modifier (after the 12 point buy, racial and class bonuses) is +8 and you state it must be +6.


I'm not trying to model the bell curve of the old dice rolls, so that and the reasons I listed are why I don't want sliding values.  Also, the +6 should be without the bonus from race/class, just with the pure 12 points spent.  I should have clarified that, you simply can't have more than an 18 with those bonuses figured in.

Like others, I don't agree with your criteria (esp. #4).  To me, a heroic character must have some foibles.  A hero is a hero not because they are spectacular, but because they spectacularly overcome their limitations.


See the first reply I made in this post and:
You can still have character flaws in your hero, there just won't be any mechanical flaws.  A score of 6 isn't heroic, hell it's barely competent for a normal human in D&D, and there's nothing heroic about being incompetent - character flaws sure, that's fun, but not incompetence.  If you really, really enjoy missing your rolls when you need to make them, just tell the GM you failed, and have at it.

Remember, this isn't about the way you like/want to play, it's about the standard method presented in the books - you still have other options.

Finally, I think it is right to consider a "standard" stat generation method for purposes of module design, officially sanctioned events, etc.  But this method serves very different purposes than a method I would want for day to day play.  Factors like balance and creativity get weighted very differently depending on the purpose for which the characters are generated.


What he said...
So I updated the OP with this:

Standard Point-Buy Method:
All of a character's ability scores start with a value of 10.  The player has 12 points to spend to improve their ability scores from 10.  Each point spent is an increase of 1 for the score.  The total of the ability score modifiers must also equal +6 after spending these points, not counting any bonuses from race or class.  The maximum value allowed in any score is 18, including any bonuses from race or class.
Personally, I don't like that point buy method. No scaling costs encourages a single high attribute with a bunch of dump stats, rather than a more even spread.



As an aside, one of my favorite ability generation methods is: The group rolls 4d6 drop lowest 36 times. Place them in a 6x6 grid in order. Each member of the group gets to pick one array from this grid, going horizontally, vertically, or diagonally.

For example, rolling a quick matrix using an online roller I get: invisiblecastle.com/roller/view/3693788/

13 | 07 | 08 | 05 | 09 | 13
12 | 07 | 15 | 10 | 16 | 16
11 | 08 | 09 | 11 | 14 | 14
15 | 15 | 13 | 12 | 10 | 17
13 | 10 | 08 | 13 | 10 | 16
07 | 10 | 14 | 13 | 09 | 16

If you were just doing normal 4d6 drop lowest, you'd have several characters from this that were unplayably bad, and some that were amazingly good. As it is, this matrix ends up with a handful of good arrays to choose from, though look at it the far right column from this is pretty much head and shoulders above the rest, so anyone interested in just high stats will go for that. Usually when I do this there's 3-4 different arrays that are fairly balanced against each other that the players will have a tough choice between. In my last campaign there was an array with an 18 and 15, but otherwise a lot of low scores, one with a 16 a 17, and some low-mid scores, and a relatively well balanced array, that most of the players picked between depending on what they were going for.

But on the bright side, anyone who cares will have that array, rather than having one player with a great set of stats and another with 2 scores above 10, and none above 14. Any players who don't care and just want a set of stats that will provide more interesting opportunities or better fits their character concept, might go with a lower array (because yes, I have seen people argue seriously in threads about die rolling that they like doing so not for the high stats, but because they'll occasionally get a 6 and figure out what to do with it from there. Those types of players are still free to grab one of the lower arrays if they prefer).
It is a universal truth, or very close to it.  There's nothing heroic about not being able to do what your class pigeon holes you to do, being too stupid to speak, being unable to hold onto any object because of your 3 DEX, etc.

Everything from 3-18 is within the normal range for a healthy adult human.  A stat of 8 is so close to average that it would barely be perceptible, unless you had a reason to look for it.  It's within one standard deviation, certainly.  Someone with 3 DEX is hardly incapable of holding onto an object - she's just slow to react against threats in combat (which can be entirely overcome with heavy armor in this edition) and she's a whopping 20% more likely to fall down when walking over an icy surface.

As odd as it is to agree with Garthanos, most people dramatically exaggerate the effects of any below average stat, mostly for the sake of levity.

The main (psychological) benefit of including stats below 10 is that it lets people feel better about the other four stats, the ones that aren't an 18 but at least they aren't a 5.  You shouldn't feel penalized when you have an 11, but that's exactly what happens when your average value is at least 12.  When you do make a check against your 5 stat, though, you expect to fail and you don't curse the dice for enforcing that. The actual mechanical difference isn't so huge, unless it's something that comes up very frequently.
The metagame is not the game.
I'm a big fan of the 6x6 grid, and everyone picks an array they like. However I find it too dificult to generate at the actual table.
I'm a big fan of the 6x6 grid, and everyone picks an array they like. However I find it too dificult to generate at the actual table.



Really? I find it pretty easy to use. If you want everyone making a character before a session, the DM providing a set of numbers via an online roller works fine (as demonstrated above). If making characters during a session, just divide up the 36 die rolls among the players. Have 4 players? Each player generates 9 stats and fills it into the grid. Everyone gets to have fun rolling dice and contributing but everyone still gets to pick from the same things.
I think a good point-buy system would cap stats at 16. So you can buy exactly what you want fit your character but spread out the points more, or you could roll and hope for a 17 or 18. 


That is what the standard array should be, WOTC rounded down rather than rounded thus the 16 average high score of 4d6drop 1 became a 15.  

So the average 4d6drop ability set rounded













16141312109

became this rounded down.













15141312108

So you have to decide which version you are using to establish the budget for your table.  The entire point of point buy is to ensure you can tilt or flatten the standard array and end up with a character of the same power.


Makes sense given the higher power level of 4e: start higher and assume more even numbers. 
Before posting, ask yourself WWWS: 
What Would Wrecan Say?

5 Minute Workday

My Webcomic
Updated Tue & Thur

 

I think a good point-buy system would cap stats at 16. So you can buy exactly what you want fit your character but spread out the points more, or you could roll and hope for a 17 or 18. 


That is what the standard array should be, WOTC rounded down rather than rounded thus the 16 average high score of 4d6drop 1 became a 15.  

So the average 4d6drop ability set rounded













16141312109

became this rounded down.













15141312108

So you have to decide which version you are using to establish the budget for your table.  The entire point of point buy is to ensure you can tilt or flatten the standard array and end up with a character of the same power.


Makes sense given the higher power level of 4e: start higher and assume more even numbers. 



I got that information from a thread in 2006 talking about 3e point buy, that is where they got the 5e numbers from.   You can compute the standard array using the dice roll statisitics/probablities is how that array was created, nothing to do with any edition of the game at all, everything to do with what the 4d6drop results are on average.

community.wizards.com/go/thread/view/758...

For 4e they used a different strategy for the standard arrays, they made the specialist, dual specialist and balanced array that conformed to the point buy, in that edition the point buy was the cumulative modifier cost nerfed down for the highest numbers. With a seperate buy to bump the one 8 to a 9 or 10 before the regular point buy budget could be applied, this was confusing and cumbersome, you get the same result by saying default is 10, so 9 is -1, and 8 is -2 and you can only have one 8 or two 9.

The point buy should not be capped at the round down or even round up average of the 4d6drop.  You should be able to go up to 18, and basically see-saw around the average.   The intent of the point buy is to balance out the parties scores against each other, it just gets min-max abuse.  By combining point buy with the random roll, you limit the min-maxing while still being random yet balanced.   And if you prefer random or point buy by itself then don't do both.
Personally, I don't like that point buy method. No scaling costs encourages a single high attribute with a bunch of dump stats, rather than a more even spread.



I think it is OK for incremental costs, as long as it is being used to budget the random roll into balance.  Then 18's are hard to get because rolling 3 18's and 3 8's is extremely unlikely even though it is easy to do with point buy used by itself.   And it avoids the balance issues of random rolls that people (dis)like.   By combining the two it becomes random yet balanced, a middle ground between random rolls and point buy by themselves, but you don't need to use the middle if you like the extremese.

Basically you end up rolling for your high stats, then your low stats are balanced into place.  And by keeping the die odds out of the point buy budget, you can use the point buy with any roll method.

> Standard Random Roll Method:

To avoid the giant chart and to keep the basic premise - same stat total but random distribution - this could be handled in a far simpler fashion.

Start with 10 in each stat.

Roll the d6 six times. If you get the same number more than four times, reroll the excess.

Each 1 gives +2 Strength
Each 2 gives +2 Constitution
Each 3 gives +2 Dexterity
Each 4 gives +2 Intelligence
Each 5 gives +2 Wisdom
Each 6 gives +2 Charisma.

The cap is still 18, you get a total of +12, everything is an even value, etc.
Everything from 3-18 is within the normal range for a healthy adult human.  A stat of 8 is so close to average that it would barely be perceptible, unless you had a reason to look for it. 


Normally 3 deviations on either side covers 95% population, only a 3d6 is a normal distribution and in that distribution, 3, 6, 17, 18 are abnormal as they are in the outlier 5%, so they are NOT in that normal human range.

The argument for 4d6 drop is that adventurers are self selected from a population of normals, rarely does the below average decides to be a heroic adventure, they are instead the NPCs that do not have what it takes.  For those that want to play a NPC that finds out they don't have what it takes to be a PC, the 3d6 option is always there.   The normal range for heroic adventurers is higher.  Only the above average (those 25% better) have what it takes, thus the 4d6 drop with its higher average.   Thus 7-17 almost 18 is the 95% range for a normal adventure, and 3,4, 5, 6 become outliers to normal.  The 7-17 range only gets bumped to 8-18 for D&D norm because people prefer evens especially if it means rounding up rather than down...

The reason for the point buy is that this statistics are for many thousands and thousands of rolls.   If you only take a few samples, even a 6x6 matrix will have very high variance in what the average and standard deviation is.    First row are the averages of the posted matrix, second row are its deviations.

































9.29.510.711.211.211.311.511.711.812.713.715.3
1.42.32.32.52.62.72.72.82.93.03.13.3

So rather than doing that many rolls and hoping it is enough to find one in the middle, instead run your roll thru the point buy budget in the roll order.  That gives you a random result that is unlikely to min-maxed, and unlikely to be too high or too low on the averages/deviation.
Normally 3 deviations on either side covers 95% population, only a 3d6 is a normal distribution and in that distribution, 3, 6, 17, 18 are abnormal as they are in the outlier 5%, so they are NOT in that normal human range.



This may be a nitpick, but if the data generally takes the form of a bellcurve (as truly randonly generated dice rolls should), 2 SDs on either side will cover 95% of the data.
There are a great many problems that can be circumvented by players and DMs having a mature discussion about what the game is going to be like before they ever sit down together to play.

 

The answer really does lie in more options, not in confining and segregating certain options.

 

You really shouldn't speak for others.  You can't hear what someone else is saying when you try to put your words in their mouth.

 

Fencing & Swashbuckling as Armor.

D20 Modern Toon PC Race.

Mecha Pilot's Skill Challenge Emporium.

 

Save the breasts.

Normally 3 deviations on either side covers 95% population, only a 3d6 is a normal distribution and in that distribution, 3, 6, 17, 18 are abnormal as they are in the outlier 5%, so they are NOT in that normal human range.



This may be a nitpick, but if the data generally takes the form of a bellcurve (as truly randonly generated dice rolls should), 2 SDs on either side will cover 95% of the data.


Oops my quality manager at work is going to demerit me.  You are correct.  Does not change my point since I used 95%.   She has software that can calculate stats for non-normal distributions, for example the common manufacturing practice of dipping into the larger low quality bin to smaller higher quality bin, which is the similar process of self selection for heroic adventurers....

> Standard Random Roll Method: To avoid the giant chart and to keep the basic premise - same stat total but random distribution - this could be handled in a far simpler fashion. Start with 10 in each stat. Roll the d6 six times. If you get the same number more than four times, reroll the excess. Each 1 gives +2 Strength Each 2 gives +2 Constitution Each 3 gives +2 Dexterity Each 4 gives +2 Intelligence Each 5 gives +2 Wisdom Each 6 gives +2 Charisma. The cap is still 18, you get a total of +12, everything is an even value, etc.



Interesting idea, or 1d12 for the +1 granularity with wraparound (1,7 is STR +1 etc) or just a second pass of d6.   Easier to say if any ability exceeds the max then reroll rather than tracking the repeats.   And you could also do it preferred, random or straight by saying you are rolling random array which you will then assign later.   And it is easily adjusted for the point buy budget variations for the risky/easy campaign, just use more/less d6.

I made a spreadsheet for random point buy because budgeting point buy with lookup table is a pain to do manually, and it was a pain to even get working in excel, so easier ways of promoting random point buy I am all for.

Now I need to spreadsheet this...


Everything from 3-18 is within the normal range for a healthy adult human.  A stat of 8 is so close to average that it would barely be perceptible, unless you had a reason to look for it. 

Normally 3 deviations on either side covers 95% population, only a 3d6 is a normal distribution and in that distribution, 3, 6, 17, 18 are abnormal as they are in the outlier 5%, so they are NOT in that normal human range.

That's not exactly what I was getting at.  I guess you could say that what I was saying cannot be represented with statistics (sounds like a cop out, but bear with me on this).

What I was trying to say is that every stat, from 3 to 18, is normal (in the colloquial sense) for a healthy adult human.  Those other stat values, which I claim to be abnormal for a healthy, functional adult, are the ones from 1-2 or 19+.

Obviously, the specific values vary from edition to edition.  In some older editions, your stats would go down with age, so it wasn't unheard of for a venerable character to have a 1 or 2 STR or DEX or CON.  There were also some spells and effects (diseases, etc) which could injure stats.  Since 3 is a valid outcome for a normally-rolled character, then (I was saying) you are still operating as pretty much a normal human for as long as your ability score was at least 3; you don't start losing the ability to speak or becoming incapable of grasping objects until you drop down to a 2.  

Of course, some of what happens when your stat gets that low has to be extrapolated from context (animals in 3.x would have an INT in the 1-2 range).  By the book, the only effects of having a 1 or 2 is an additional -1 to relevant checks; every evel of critically debilitating stat is compressed between the values of 0 and 1, because the system just isn't designed to handle it.
The metagame is not the game.
> Standard Random Roll Method: To avoid the giant chart and to keep the basic premise - same stat total but random distribution - this could be handled in a far simpler fashion. Start with 10 in each stat. Roll the d6 six times. If you get the same number more than four times, reroll the excess. Each 1 gives +2 Strength Each 2 gives +2 Constitution Each 3 gives +2 Dexterity Each 4 gives +2 Intelligence Each 5 gives +2 Wisdom Each 6 gives +2 Charisma. The cap is still 18, you get a total of +12, everything is an even value, etc.


Here is the unlikely odds chart for 1000 12d6 sets randomly allocating the modifiers (+2 ability), you notice I started at 8, with this method the average is dead on base+4, since 4d6drop is 12.25 average starting at 10 is too high .   And actually you need a bakers dozen(13) to get the similar odds (12.3) as 4d6 drop1

I also include 20's because it was too hard to program Excell to reroll, since that rarely happens it would not change the numbers as much as running a larger sample to get the probability rather than statistical odds.























81012141618
2.61.11.01.53.49.8

That means a 8 and 16 char with middle numbers of 10 12 or 14 is likely (an even version of the 5e standard array) with every other character getting an 18.

For similar spread as 4d6 drop with the only difference being balanced and bounded, do 19d6 starting with 6.  























681012141618
8.22.01.21.01.32.14.5

If do it as starting at 10 then average is 14 and it is easier to get 18's (no need to reroll, just shift the unlikely odds chart)























101214161820
2.61.11.01.53.49.8

But I think the start from 8 version is better for 5e which is assuming lower numbers, I would do the start from 10 in 4e though, buying a bump on one number for getting one 8.   But that is the good thing about this idea, it is very easily adjustable average and min/max bound.

Starting at 8 does not work for odd numbers since you roll twice as much allocating the +1's you get a tighter distribution around the average.  



































89101112131415161718
16.53.61.61.11.01.31.93.89.525.978.2

So if you want odd numbers certainly start at 10, but even then 18's and 10's are hard to get, offset with 11, 17 more feasible, and your average is 14.



































101112131415161718
16.53.61.61.11.01.31.93.89.5

So for standard play rule I think start from 8 and do the 12d6 to randomly buy modifiers is the way to go.

Here it is for start with 10 and do 6d6 buying modifers as OP.   I can't get behind that one as it pretty much does very little variation on 10 10 12 12 14 and either 14 or 16 prime.



















1012141618
112751


Want a 4-20 range balanced for 12?   Then do 24d6 starting with 4.



























468101214161820
16.63.31.51.01.01.21.93.88.1


2-18 range balanced for 10? Then do 24d6 starting with 2.  Same odds as above but shifted.



























24681012141618
16.33.41.61.01.01.22.03.98.6



The rule is simply mathematically stated in a very general way, to balance around X/3+Y average with min bound Y, max bound Z

roll X  d6 with each die determining which ability gets the +1 modifier stack (+2 ability) for 1st thru 6th ability.  

The completed ability array is then assigned in random, preferred or straight order.  

Reroll any ability modifier that would increase ability over Z.

If you prefer odd abilities then pair the even abilities increasing/decreasing by 1.

If you prefer not to roll at all, then you are free to buy X modifiers starting at Y ability and stopping at Z ability.

If you prefer not to budget then X must be divisible by 3, simply do (X/3)d6 for each ability, dropping worst die to keep a 3d6.

Standard play uses the d6 budgeting method with X = 13, Y = 8, Z = 18 and preferred ordering.