Tried the 3D6 matrix for char creation

[ZorroGames]

6 sets of 6 stats using 3D6 pick best 6stats.

In descending order (not stat order)

Range of results was 16 to 4.

16,14,14,14,14,14,13

Interesting. Definitely heroic. Not Superman in a robe.

So,

DMs - would you allow it?

Players - how do feel about this character, especially if you used stand array?

I can go into more detail about the rolling if you want the gritty details.

[Jaxxen]

Me personally and this Is just me as a player, I prefer that people use either the point buy system or the standard array. To often the "I totally rolled 4 18s" comes up. Now admittedly if they rolled it in full view of the table and DM that's a different story. The other reason I prefer the point buy/array is because its more likely to make your character have some kind of flaw as opposed to bonuses to everything.

[nickl_2000]

Character creature really needs to depend on the table. I'm playing/Dming at a table where several players work together and has been together for years. If someone makes a character and the stats are good, then we trust that they actually rolled it. Several of us have even docked our owned rolled stats because they are to good (myself included). If I were doing roll 6 sets of 3d6 and choose whatever you want, and trusted the players. Then it would be fine

If you are playing at a table where you don't know the people or you have a new player that you don't trust. Have someone help them roll a character of go with the standard buy.

[ZorroGames]

Okay, note, nobody, not even my wife, gets to play a character not witnessed.

For the paranoid, can we drop the cheating player mantra now?

This is a character creation method discussion, not a "Can I play this really cool character" discussion.

Here are the results of the rolls:

16x1
14x4
13x4
12x6
11x2
10x4
9x3
8x5
7x1
6x3
5x2
4x1

No rolls of 18,17,15, or 3.

Not exactly overwhelming seen in total IMO.

Even if you say "you get one roll from each of the six sets," it only drops one 14 to 13.

Anyone want to see the actual matrix? I can provide it if you are that paranoid.

[Vorok]

You got 7 numbers there


Players - how do feel about this character, especially if you used stand array?

I haven't played any such character yet, but the concept seems bland. Depending on the race and stat allocation, the lowest bonus can be +2 to everything even before proficiency (or +1 for one modifier). Where's the weakness of such character?

When I use stat array or point buy, the character begins with both strengths and weaknesses, which inform the character, and can show the way for character growth. If the character begins as above average on everything, where does he go from there?

[ZorroGames]

So I do. Mea culpa. Small type and bad eyes.drop the 13 then for discussion.

Thanks.

Edit: drop the last 14, keep the 13.

Double darn.

[nickl_2000]

I think I missed the point of the OP. Sorry about that. As a DM and a player I wouldn't choose that method or those stats. I like a character to have flaws that come to a forefront and impact the party from the stats.
Low Charisma?, he is obnoxious in a tavern and starts a brawl, maybe he talks over an NPC trying to give information.
Low strength? then he can't wear the best armor or carry much. He needs help getting around obstacles
Low Wisdow? he has no common sense, he blunders into caves without really thinking it through
Low Int? he sees research as a waste of time and gets pissy when he has to go into a library

Etc...

[ZorroGames]

You got 7 numbers there



I haven't played any such character yet, but the concept seems bland. Depending on the race and stat allocation, the lowest bonus can be +2 to everything even before proficiency (or +1 for one modifier). Where's the weakness of such character?

When I use stat array or point buy, the character begins with both strengths and weaknesses, which inform the character, and can show the way for character growth. If the character begins as above average on everything, where does he go from there?

Okay I was not assuming racial modifiers as being in play else you could get a starting 20. You get your racial modifiers built into your stat selection.

Still adjusting from OD&D origin where 3D6 was just that - your stat.

[JAL_1138]

The more dice you roll with a "pick the best," the higher the stats will be. It's just skewing the whole array upward.

I don't have JMP to do the statistics I don't know how to do longhand anymore, so I'll leave that to the math wizards.

But you can't draw too many conclusions from one test of a dice-rolling method. Statistically it's meaningless. You need to figure up the averages (to see how far it usually skews from the standard array's average or 4d6B3's average) and the probabilities of each result (to see how likely you are to end up with what numbers).

Without knowing the bell curve this method generates, you can't draw any conclusions about the method itself from the one statblock you've generated.

EDIT: Anybody know the probability formula for 3d6 6 times best 3? While I can't say for certain what the bell curve will be, I'd eat my hat if it didn't have a much higher chance of high stats than 4d6b3.

[Vorok]

Okay I was not assuming racial modifiers as being in play else you could get a starting 20. You get your racial modifiers built into your stat selection.

Still adjusting from OD&D origin where 3D6 was just that - your stat.

I'm not sure I follow, no race has more than +2 to an ability score bonus, so with your stats, you're gonna start with something like 18,15,14,14,14,14. 14 gives +2, 18 gives +4, no 20 in anything.

[ZorroGames]

Interesting aspect of changing to 5th from 0D&D - everybody seems to want a weakness.

My First Wizard would be deemed "unplayable" by 5th Edition. And rightly so. I see why but a low charisma, 12 IN , all other stats 9 to 11 wizard made to the teens with help from the other players and surviving 1D4 per level hit points. You want weakness? I was expecting her to die every adventure until one day I realized she could learn third level spells. Still an egg armed with a hammer. But currently a retired egg teaching young wizardlings wannabes. For a price.

I see the benefit of the bland, very bland, everybody is boring the same "standard array". And I like it. I prefer the point buy. I still think it is worthwhile discussing alternatives if, yea, especially if, you see the pitfall from above not the bottom of the pit.

Who did I miss pissing off? :smallwink: :smalltongue: :smallbiggrin::smallbiggrin:

[Sirithhyando]

Well, if someone decide to do standard array and all the others roll for it, of course i'd have stats way lower than the others.

What we usually do, we roll 4d6 best 3 (in plain view all the time).
Then we look the average, if it's an average mod of +9 then everyone must be there plus/minus 2. So between +7 to +11, it keeps a certain balance between the players though of course we're all atronger than with point buy or standard array.

My favorite character i've made with that started with an 18 and 2 16 but with a 5, 8 and 14. So it ended up an half-orc champion and with ASI is now 20 in str-dex-con but still can't see a thing with a 5 in wisdom and 8 in intel. It's funny, he really do everything his companion ask though often too well and too literally.

Ex : Druid say "Make sure no one follow us from the tavern"; Half-Orc take a stance with shield on the door and spear at the ready intimidating everyone who dare open the door.

[ZorroGames]

Well, if someone decide to do standard array and all the others roll for it, of course i'd have stats way lower than the others.

What we usually do, we roll 4d6 best 3 (in plain view all the time).
Then we look the average, if it's an average mod of +9 then everyone must be there plus/minus 2. So between +7 to +11, it keeps a certain balance between the players though of course we're all atronger than with point buy or standard array.

My favorite character i've made with that started with an 18 and 2 16 but with a 5, 8 and 14. So it ended up an half-orc champion and with ASI is now 20 in str-dex-con but still can't see a thing with a 5 in wisdom and 8 in intel. It's funny, he really do everything his companion ask though often too well and too literally.

Ex : Druid say "Make sure no one follow us from the tavern"; Half-Orc take a stance with shield on the door and spear at the ready intimidating everyone who dare open the door.

OMG, I love it! I know a dwarf not too unlike the that.

[Mortis_Elrod]

My group uses a similar style. We allow point buy and array and 4d6 drop lowest. But generally we all use 4d6 drop lowest. 6x6 box of numbers. Have to pick a set of numbers either left to right; up and down, or diagonal from on corner to the other. All the numbers are rolled like 6 sets in a row, 6 rows. This way you can't pick the best numbers, you have to pick a set. usually works out that an 18 (if any) is tied in with a 6-9.

[JAL_1138]

Interesting aspect of changing to 5th from 0D&D - everybody seems to want a weakness.

My First Wizard would be deemed "unplayable" by 5th Edition. And rightly so. I see why but a low charisma, 12 IN , all other stats 9 to 11 wizard made to the teens with help from the other players and surviving 1D4 per level hit points. You want weakness? I was expecting her to die every adventure until one day I realized she could learn third level spells. Still an egg armed with a hammer. But currently a retired egg teaching young wizardlings wannabes. For a price.

I see the benefit of the bland, very bland, everybody is boring the same "standard array". And I like it. I prefer the point buy. I still think it is worthwhile discussing alternatives if, yea, especially if, you see the pitfall from above not the bottom of the pit.

Who did I miss pissing off? :smallwink: :smalltongue: :smallbiggrin::smallbiggrin:

I played in 2e and we tended to use 4d6b3, but did a few 3d6 straight down (and yes, a lot of my wizards--and other classes--died from one hit, because negative HP were an optional rule and the standard was death at 0). You had a large range before modifiers happened, and those modifiers didn't affect opponents' saving throws. Those went off a class/level table. Every wizard had the same spell save DC--their mod affected the spells they could learn, not the effectiveness of the spells they knew. Strength mod didn't start giving you an accuracy bonus until...can't remember but it was somewhere around 16 or 17.

The game math for 5e is not the game math for AD&D, and your scores matter much, much more for game balance.

EDIT: Again, the fact that you didn't roll a superhuman with one test of the method isn't statistically valuable when discussing whether the method is balanced. When rolling dice it's always statistically possible, if extremely unlikely, to roll nothing but ones, nothing but sixes, nothing but threes, etc., so it's not really useful to look at a single result. The array you've generated could be on the low, side, or the average side, or even high side (but I strongly doubt that it's on the high side) for the method's likely results, for all anybody knows without crunching the numbers. Somebody better at math than me would need to get the bell curve for it and see what it's likely to do to the array to see whether it's a good method or is likely to bork the game balance to heck and gone.

[nickl_2000]

Interesting aspect of changing to 5th from 0D&D - everybody seems to want a weakness.

Guilty as charged. I find that it adds a depth to a character that may otherwise not exist. It makes the player think more through about how you would mitigate something they are not good at compared to being decent at everything. It may seem odd in a place where we cast magic, but it almost makes it more real. How many stories have you heard about the Mathematical Genius whose brain worked on a level that most people can't even comprehend, but would get so tied up in their work that they would forget to eat and shower?

[Thrudd]

Yeah, I found that method generally gave the strongest scores in 1e. Were you using the exact method from 1e AD&D? This method's high scores are tempered a little bit, in the btb method, by being rolled in stat order. You don't just get to pick the best 6 of 36 rolls and plug them in where you want, you roll six times for each stat.

There was another method where you roll 3d6 in order, enough to make 12 characters, and pick the array you want.

Or roll 3d6 twelve times, take the best six and arrange as you want. Bump that up to 18 or 20, and it might give slightly more 5e friendly arrays.

I do like the 6x3d6 for each ability method, when I want the players to all feel good about their characters' prospects.

[JAL_1138]

Yeah, I found that method generally gave the strongest scores in 1e. Were you using the exact method from 1e AD&D? This method's high scores are tempered a little bit, in the btb method, by being rolled in stat order. You don't just get to pick the best 6 of 36 rolls and plug them in where you want, you roll six times for each stat.

There was another method where you roll 3d6 in order, enough to make 12 characters, and pick the array you want.

Or roll 3d6 twelve times, take the best six and arrange as you want. Bump that up to 18 or 20, and it might give slightly more 5e friendly arrays.

I do like the 6x3d6 for each ability method, when I want the players to all feel good about their characters' prospects.

We generally used either 4d6b3 arrange however, 4d6b3 straight down, or 3d6 arrange however, or 3d6 straight down.

Run the numbers. I can't say for sure, but on gut feeling I will eat my hat if your methods don't average WAY higher than the official rolling method for 5e, which is 4d6b3 assign however, and leave standard array in the dust. Which, if so, will mess up the game math for 5e a lot worse than it did in AD&D.

[ZorroGames]

Me personally and this Is just me as a player, I prefer that people use either the point buy system or the standard array. To often the "I totally rolled 4 18s" comes up. Now admittedly if they rolled it in full view of the table and DM that's a different story. The other reason I prefer the point buy/array is because its more likely to make your character have some kind of flaw as opposed to bonuses to everything.

I understand your position and point buy helps me create my character, warts and all.

[ZorroGames]

Here are thevrolls from the matrix, note that in the beginning you chose one set in toto (no racial modifiers at first that changed quickly) and in the order rolled. Not suggesting that today though.

+ = rolls over 12, - = rolls below 9

6,10,12,12,16,5 Die often, die young +1 -2
13,8,9,13,10,5 strong, tough grumpy potential +2 -2
10,8,11,14,10,9 tough but playable +1 -1
12,12,12,12,6,14 talk fast, hire a bodyguard +1 -1
9,11,14,14,4,6 OH My +2 -2
8,13,7,8,8,13 Spoke too soon +2 -4

Not exactly overwhelming characters unles you chose the high 6 stats...

[ZorroGames]

The more dice you roll with a "pick the best," the higher the stats will be. It's just skewing the whole array upward.

I don't have JMP to do the statistics I don't know how to do longhand anymore, so I'll leave that to the math wizards.

But you can't draw too many conclusions from one test of a dice-rolling method. Statistically it's meaningless. You need to figure up the averages (to see how far it usually skews from the standard array's average or 4d6B3's average) and the probabilities of each result (to see how likely you are to end up with what numbers).

Without knowing the bell curve this method generates, you can't draw any conclusions about the method itself from the one statblock you've generated.

EDIT: Anybody know the probability formula for 3d6 6 times best 3? While I can't say for certain what the bell curve will be, I'd eat my hat if it didn't have a much higher chance of high stats than 4d6b3.

I believe the term is cherry picking.

[ZorroGames]

We generally used either 4d6b3 arrange however, 4d6b3 straight down, or 3d6 arrange however, or 3d6 straight down.

Run the numbers. I can't say for sure, but on gut feeling I will eat my hat if your methods don't average WAY higher than the official rolling method for 5e, which is 4d6b3 assign however, and leave standard array in the dust. Which, if so, will mess up the game math for 5e a lot worse than it did in AD&D.

Actually I have come to the same non-mathematical conclusion myself.

[ZorroGames]

okay, no cherry picking, no racial pluses, no stat set with a number below 8 (which, BTW, eliminates the 16,) leaves you 14, 11, 10, 10, 9, 8 as the only potential 5th Edition "playable" character.

Scramble those numbers anyway you want - do you have a playable chapter?

Defaulting to my first two PC choice of Cleric:

ST 10, IN 10, WI 14, DE 8, CO 9, CH 11. Playable per se'?

Thoughts on alternative class?

[ZorroGames]

Guilty as charged. I find that it adds a depth to a character that may otherwise not exist. It makes the player think more through about how you would mitigate something they are not good at compared to being decent at everything. It may seem odd in a place where we cast magic, but it almost makes it more real. How many stories have you heard about the Mathematical Genius whose brain worked on a level that most people can't even comprehend, but would get so tied up in their work that they would forget to eat and shower?


BTW, not a slam or complaint. Just adjusting to 5th Edition mind set.

My character's usual weakness is the player (me.)

[Pex]

There's nothing wrong with having an 8, but I object to the absolute requirement of one. It is not requirement to have one to be considered a good player or have a complete character. Having a low score is one way to have a flaw but not the only way. There's the inherent flaw that exists because players are people and people are not perfect. The decisions players make result in consequences. A character's values affects the decisions which can result in interesting complications. There is also the flaw in class game mechanics. The great smiting paladin might not use a bow and have difficulty when peppered with arrows and spells from far away. The low hit point wizard is in trouble after the enemy unleashed a Fireball or two. The wild shaped moon druid is losing hit points fast because he's being attacked by an orc chief barbarian and even using great weapon master pummeling through his low AC.

[Goggalor]

Our table used to do the 6x6 stat array, as well, until one day we decided to make a new standard array of 16, 14, 14, 12, 10, 10. Along with that, each character gets a free feat at first level (variant human is not allowed and usually the first level feat cannot be a "combat" feat) and you are not allowed to raise a stat to 20 until level 8. Additionally, we also make use of Zman's feat changes, though, that is homebrew territory, so your mileage may vary on that. It makes for stronger early levels, but feels like you can focus more on creating the actual character rather than focusing on getting that next ASI.

As for the 16, 14, 14, 14, 14, 13, it is not game-breaking and can let the person try out some the crazier MAD builds. :D

[ZorroGames]

Diverging slightly in response to Pex.

On another thread we were discussing Monk builds for a dwarf.

Me? Fixated on Dwaves?

Ended up with these as an example of a Hill Dwarf Monk.

ST 10
IN 9
WI 16 (15+1)
DE 15
CO 15 (13+2)
CH 9

I think those were the values.

So,

Average strong but not overly strong
Not stupid but might need some time to think
Wise enough to usually know better
Agile, fast in sprints - JK - for a dwarf
Rugged
Perhaps introverted, perhaps jaded on life

Playable and fits my idea of the character.

No 8s, weak areas are IN and CH related, strong in WI, DE, CO, low average to average in 3 of 6.

Better than rolling? POV issues in stating yea or nay.
Like it better than standard array, yes, flavor reasons.

Now to play with a Mountain Dwarf Monk... :smallsmile:

[Laserlight]

DMs - would you allow it?

Players - how do feel about this character, especially if you used stand array?


As a DM, I'd allow it. Lets you multiclass into anything you want, but only has one 16 -- not overpowering.

(I'd also let you play with three 18s, if I saw you roll them).

As a player, this array wouldn't bother me. You may have +1 better INT save than everyone but the wizard, and +1 better STR than everyone but the fighter, and +1 better CHA than everyone except the rogue and the bard and the sorc and the paladin....but who cares? I'm much more concerned with whether you do interesting RP and are tactically competent. If you persistently charge alone into the midst of the enemy mob, or if you have the personality of room temperature oatmeal, or if you go off on a tangent with every other sentence instead of paying attention to the fight, those are vastly more irritating to me than having +1 to hit better than I do.

[Thrudd]

We generally used either 4d6b3 arrange however, 4d6b3 straight down, or 3d6 arrange however, or 3d6 straight down.

Run the numbers. I can't say for sure, but on gut feeling I will eat my hat if your methods don't average WAY higher than the official rolling method for 5e, which is 4d6b3 assign however, and leave standard array in the dust. Which, if so, will mess up the game math for 5e a lot worse than it did in AD&D.

Yes, 3d6x6/in order averages a lot higher than 4d6b3/arrange. Knowing that, you'd choose which one you preferred for the tone and difficulty of your campaign. 3d6x12, arrange isn't all that much higher on average than 4d6 b3, in my experience, but maybe a little. 3d6 in order, pick the best array of 12 doesn't give you any better than 4d6b3, generally. It does give you some interesting choices sometimes, where you have to choose between an array with mostly mediocre but no really weak scores with one that might have an 18 but also a couple of really low scores. It also can be a prompt to decide on a character if you're not sure what to play, since the arrays will speak to you regarding classes that fit them best.

[Zorku]

I was gonna set up a little monte carlo simulation in excel... but then I realized how late it was.

So pre-proper simulation, here are 10 sets
12 12 11 10 10 10
14 12 11 11 11 11
16 14 12 12 12 12
14 14 13 13 13 13
16 15 13 12 12 12
14 12 12 12 11 11
14 14 12 12 11 11
15 13 13 12 12 12
13 13 12 12 11 11
14 13 12 12 11 11

Seems like it skews a little worse than standard array (without any 8s in it,) but I should be able to do a proper 1000 trials sort of thing tomorrow.

[JAL_1138]

I was gonna set up a little monte carlo simulation in excel... but then I realized how late it was.

So pre-proper simulation, here are 10 sets
12 12 11 10 10 10
14 12 11 11 11 11
16 14 12 12 12 12
14 14 13 13 13 13
16 15 13 12 12 12
14 12 12 12 11 11
14 14 12 12 11 11
15 13 13 12 12 12
13 13 12 12 11 11
14 13 12 12 11 11

Seems like it skews a little worse than standard array (without any 8s in it,) but I should be able to do a proper 1000 trials sort of thing tomorrow.

If I have a hat to eat, so be it, I'll get some barbecue sauce, but just to be clear, this is 3d6 rolled six times, cherry-pick the best three, for each stat?

[Pex]

Diverging slightly in response to Pex.

On another thread we were discussing Monk builds for a dwarf.

Me? Fixated on Dwaves?

Ended up with these as an example of a Hill Dwarf Monk.

ST 10
IN 9
WI 16 (15+1)
DE 15
CO 15 (13+2)
CH 9

I think those were the values.

So,

Average strong but not overly strong
Not stupid but might need some time to think
Wise enough to usually know better
Agile, fast in sprints - JK - for a dwarf
Rugged
Perhaps introverted, perhaps jaded on life

Playable and fits my idea of the character.

No 8s, weak areas are IN and CH related, strong in WI, DE, CO, low average to average in 3 of 6.

Better than rolling? POV issues in stating yea or nay.
Like it better than standard array, yes, flavor reasons.

Now to play with a Mountain Dwarf Monk... :smallsmile:

Nitpick: Character has two 9s which are the same as 8s as far as the game is concerned except for ability score increasing.

That character doesn't bother me. What would bother me is someone telling me if my array was 16 16 14 12 12 10 I'm playing the game wrong.

[JAL_1138]

Nitpick: Character has two 9s which are the same as 8s as far as the game is concerned except for ability score increasing.

That character doesn't bother me. What would bother me is someone telling me if my array was 16 16 14 12 12 10 I'm playing the game wrong.

To clarify my earlier comments, that particular array is fine, although it's definitely on the strong side and somewhat affects multiclassing balance; how the stat generation method works statistically is my question. Rolled stats will always be able to produce that kind of array; the question is more one of frequency/likelihood.

If, say, it skews higher, borking the game math by using it expecting it to work like the standard rolling method for game balance, and expecting it to be usable alongside a 27-point array would be an issue. You could compensate for a higher, you could either drop 27-point buy as an available chargen method or increase the array to something that results in characters of roughly-comparable power. If it skews lower, you could likewise compensate by reducing the point-buy.

In either case, you'd also want to be aware of (although you might not want to adjust much, depending on the tone of the campaign you're going for) the effects on encounter difficulty and check/save DCs.

[Knaight]

I could run a proper Monte Carlo in Octave, but it's a pain. However, a few simplifications makes this easy in AnyDice. Notably:

I'll be looking at the total scores (all six attributes added up), and not the individual set.
Modifier implications with odd and even numbers are going to get lost in the noise.
The standard deviations generated should be understood as standard deviations of a distribution with a pretty significant skew.


AnyDice code.

output [highest 1 of 6d(18d6)]
output 6d[highest 3 of 4d6]
output 18d6

This does run into execution problems complements of the five second timer (at least on my computer and internet). However that first line can be played with by examining a smaller matrix. First, the other two:
Best 3 of 4d6: Mean of 73.47, std of 6.97
3d6: Mean of 63.00, std of 7.25

Now for smaller matrices:
Matrix of 2: Mean of 67.09, std of 5.98
Matrix of 3: Mean of 69.14, std of 5.40
Matrix of 4: Mean of 70.46, std of 5.05

Taking into account that the matrix of 1 data is also there (it's the 3d6 down the line method), there are a few notable things.

The mean progressively grows.
The rate at which the mean grows drops, and it drops quickly (a gap of 4.09, then 2.05, then 1.32)
The standard deviation progressively drops
The rate at which the standard deviation drops progressively drops in magnitude (a gap of -1.27, then -0.58, then -0.35)


These patterns are expected, and can be expected to continue. This tells us two things about the matrix of 6 case.
1) The mean is less than 70.46+1.32+1.32, which is to say it's less than 73.10
2) The standard deviation is less than 5.05.

Best 3 of 4d6 had a mean of 73.47, as previously discussed. That makes it a stronger option than the matrix of 6. For 7 and beyond I can't comment.

[JAL_1138]

I could run a proper Monte Carlo in Octave, but it's a pain. However, a few simplifications makes this easy in AnyDice. Notably:

I'll be looking at the total scores (all six attributes added up), and not the individual set.
Modifier implications with odd and even numbers are going to get lost in the noise.
The standard deviations generated should be understood as standard deviations of a distribution with a pretty significant skew.


AnyDice code.

output [highest 1 of 6d(18d6)]
output 6d[highest 3 of 4d6]
output 18d6

This does run into execution problems complements of the five second timer (at least on my computer and internet). However that first line can be played with by examining a smaller matrix. First, the other two:
Best 3 of 4d6: Mean of 73.47, std of 6.97
3d6: Mean of 63.00, std of 7.25

Now for smaller matrices:
Matrix of 2: Mean of 67.09, std of 5.98
Matrix of 3: Mean of 69.14, std of 5.40
Matrix of 4: Mean of 70.46, std of 5.05

Taking into account that the matrix of 1 data is also there (it's the 3d6 down the line method), there are a few notable things.

The mean progressively grows.
The rate at which the mean grows drops, and it drops quickly (a gap of 4.09, then 2.05, then 1.32)
The standard deviation progressively drops
The rate at which the standard deviation drops progressively drops in magnitude (a gap of -1.27, then -0.58, then -0.35)


These patterns are expected, and can be expected to continue. This tells us two things about the matrix of 6 case.
1) The mean is less than 70.46+1.32+1.32, which is to say it's less than 73.10
2) The standard deviation is less than 5.05.

Best 3 of 4d6 had a mean of 73.47, as previously discussed. That makes it a stronger option than the matrix of 6. For 7 and beyond I can't comment.

I did some looking, and it looks like the folks at rpg.stackexchange did some number crunching (https://rpg.stackexchange.com/questions/1528/what-is-the-statistically-superior-character-creation-method-twelve-3d6-or-six)--for a larger set of 3d6 rolls than OP described. It's getting into statistics beyond my math-challened grasp, but it looks like the answers might not be as simple as they look.

At any rate, however, I'll go ahead and start frying up that hat I said I'd eat.

[Knaight]

I did some looking, and it looks like the folks at rpg.stackexchange did some number crunching (https://rpg.stackexchange.com/questions/1528/what-is-the-statistically-superior-character-creation-method-twelve-3d6-or-six)--for a larger set of 3d6 rolls than OP described. It's getting into statistics beyond my math-challened grasp, but it looks like the answers might not be as simple as they look.

At any rate, however, I'll go ahead and start frying up that hat I said I'd eat.

They're looking at an entirely different question - the matrix here posits entirely separate sets of six, one of which is picked. What they're looking at is the difference between generating more than 6 attributes worth of rolls and then picking the best 6 of that set. In anydice terms, we're looking at:

output [highest 6 of 7d(3d6)]
where that 7 can be messed with (they use 12, but that also broke the 5 second runtime for me). I'd have to run a completely new analysis on this (although the same basic behavioral profile of a narrowing standard deviation and rising mean with a decrease in rise rate should still hold).

[Beelzebubba]

Okay I was not assuming racial modifiers as being in play else you could get a starting 20. You get your racial modifiers built into your stat selection.

Still adjusting from OD&D origin where 3D6 was just that - your stat.

Here's my take - OD&D is built around 3d6 stats. It's 'fantasy Vietnam', where death is more frequent, and like kids in the Middle Ages, characters don't really get a 'name' until they reach 3 or 4. :smallbiggrin:

5E is 'heroic fantasy'. Death is harder, you make the character a fully-realized person early on (Backgrounds, Traits, Ideals, Bonds, Flaws, Trinket) that is much easier to get attached to, and they start off much more powerful. It also seems to be balanced for a *slightly* higher stat array, at least early on (from what I can tell).

That set of stats you got is really reasonable. My table has 2-3 people that are higher rolling 4d6 drop low.

It's a weird method - I don't have the math brain to understand what it *means* - but why not run your table with it, and post how it worked out?

[Dappershire]

Just to throw my opinion in here, I prefer rolling for stats over point buy.
But, caveat, only in front of the DM, or a couple fellow players. I'd never want to see a character rolled at home, allowed in a game im in.

[Beelzebubba]

Just to throw my opinion in here, I prefer rolling for stats over point buy.
But, caveat, only in front of the DM, or a couple fellow players. I'd never want to see a character rolled at home, allowed in a game im in.

Yeah, that's how our group runs. All stat rolls in front of the DM. The only way to make it feel fair.

[JAL_1138]

They're looking at an entirely different question - the matrix here posits entirely separate sets of six, one of which is picked. What they're looking at is the difference between generating more than 6 attributes worth of rolls and then picking the best 6 of that set. In anydice terms, we're looking at:

output [highest 6 of 7d(3d6)]
where that 7 can be messed with (they use 12, but that also broke the 5 second runtime for me). I'd have to run a completely new analysis on this (although the same basic behavioral profile of a narrowing standard deviation and rising mean with a decrease in rise rate should still hold).

The way I understood the first post, rolling "3d6 x 6, pick best set of 3, per stat" was essentially what was being suggested. That was clarified (my reading was incorrect) later on by the OP.

Think I'll beer-batter that headgear :smalltongue:

[Finback]

Ex : Druid say "Make sure no one follow us from the tavern"; Half-Orc take a stance with shield on the door and spear at the ready intimidating everyone who dare open the door.

I have PCs with a higher INT who would read that as code for "set fire to the building, implement Order 66".

[Dappershire]

The way I understood the first post, rolling "3d6 x 6, pick best set of 3, per stat" was essentially what was being suggested. That was clarified (my reading was incorrect) later on by the OP.

Think I'll beer-batter that headgear :smalltongue:

So before I embarrass myself, how -does- everyone roll for stats?
I've always rolled 4d6 per stat, keep the top three. So yeah, odds were slightly more in favor of an 18 over a 3, but norms don't choose to adventure.
I've also known DMs that did the 4d6, choose the top three d6 (for your pick of three stats) but you'd have to take the lowest three d6 for the remaining stats. Which was different, but not unfair.

[JAL_1138]

I have PCs with a higher INT who would read that as code for "set fire to the building, implement Order 66".

...and take their stuff when the flames die down.

[ZorroGames]

Yeah, that's how our group runs. All stat rolls in front of the DM. The only way to make it feel fair.

Yes, had the experience of a 8th grader bringing a home rolled character to play with his friends/classmates who did "solo dungeons" at home alone (of course) that had flat out ungodly stat material and equipped with "magical" everything, mostly +2 or better. And if he died he always had a simulacrum or clone, blah, blah, blah.

I did not participate in the "Great Defeat" but his cohorts finally came to the leadership (adults) of the club and asked how to deal with his character who they felt was a "cheat" in so many ways. The President pointed out the public library prohibited an discrimination and unless he was disruptive or broke the law he could not be prevented from coming. The long story of how they actually killed off his character forever, hung on his own petard, is off topic but after their "funeral" cake he eventually cooled off enough to come back and turn out to be decent role player while still of Chaotic Greedy alignment number crunching min-max player.

So I understand the concern for being "overly biased" in creating a character. Like rolling over any rolls under 15 in OD&D for 3D6 stats. I saw that in action at our club once. The young are so easily tempted here as the old are so easily easily persuaded to sin. :smallwink:

[Beelzebubba]

I understand the concern for being "overly biased" in creating a character. Like rolling over any rolls under 15 in OD&D for 3D6 stats. I saw that in action at our club once. The young are so easily tempted here as the old are so easily easily persuaded to sin. :smallwink:

It's totally that age. There's something in that stage of emotional development that forces them to try to get away with everything they possibly can. I still remember some of the shenanigans I pulled quite vividly...and I'm surprised I didn't get beat up even more.

:smallamused:

[ZorroGames]

As OP :smallwink: I guess I should clarify my version of the matrix (extreme makeover) version I was fiddling with.

Caveat: I never actually used this in my adventuring because it was suggested after I had mid-level characters in multiple games (Dwarf Fighter, Dwarf Cleric, Human Cleric x 2, Human Wizard, Human Fighter x2, no halfling, no elf at that time) so I have no data as to how it worked in large numbers of cases. I took my chances on 3D6 rolled in stat order because that was standard at the time. Like medieval children, making it past being an "infant" (levels 1 to 4) was the "expel the held breath" moment.

Okay here is what I did.
Draw out a "spreadsheet" on a piece of paper.
Roll 6 3D6 writing then in a column
Roll 6 3D6 putting the numbers in the next row
Create 4 more rows.
6 sets of 6.

Here is where it could break down for IMO with ungodly rolling in the matrix.

Pick 6 stats.
Not a set of 6 stats (row, column, or diagonal) but any 6.
Need not be the highest 6 but I did that to see what resulted.

Hence the stats I listed.

Nothing higher than 16 in this case:

16,14,14,14,14,13 (move stat of choice to ability)

No racial bonuses BTW. You get enough latitude in arranging you scores as it is. I see this as a decent starting figure for OD&D and slightly stronger than a point but character with a noted edge over Standard Array characters.

Okay, continue discussing, I find this very enlightening.

[ZorroGames]

It's totally that age. There's something in that stage of emotional development that forces them to try to get away with everything they possibly can. I still remember some of the shenanigans I pulled quite vividly...and I'm surprised I didn't get beat up even more.

:smallamused:

LOL in East L. A. growing up the gang bangers referred to me as "a smart ass set of lips asking to be busted." My vocabulary and my sharp tongue was both my strength and my weakness.

[Zorku]

If I have a hat to eat, so be it, I'll get some barbecue sauce, but just to be clear, this is 3d6 rolled six times, cherry-pick the best three, for each stat?

No, I've probably misunderstood the method if that's what we're supposed to be doing. This is 3d6 rolled 36 times, picking the best 6 sets.

Ive got the spreadsheet working now so my apparently wildly mistaken method produces about about 150 16s, 340 15s, 320 14s, 130 13s, 30 12s, and the odd 11 once in awhile, out of 1000 trials (1 27 146 355 336 135 last run.) Where an 18 should have a 1 in 216 chance of happening I worry that I have done something terribly wrong in the early steps of the spreadsheet :/
Best guess? Maybe the Randbetween function secretly keeps track of the decimal place and I'm adding a bunch of 5.3s together. I really didn't expect to be hunting bugs in such a simple excel sheet, so I'll have to shelf this for awhile.

e: Redid it from scratch in a new file with entirely the same formulas and this time it works differently. I don't have any idea why.

Most recent run gave me:
3 20 161 335 317 151
Where this is the best number in the array, starting from 13 and going up to 18. 36/213 = 16%, so I'm getting about as many lead 18s as you would expect on this attempt. I should also roll 3x as many 17s, which doesn't quite match this run as well as with the 18s, but it's still reasonably close when you account for how many should be obscured by the 18s. You've got a <<1% chance of having a 12 as your best stat, a 0.3% chance of a 13, a 2% chance of a 14, 16% 33% 31% and 15% chances for the remaining numbers.

Here's the first 10 arrays
17 16 16 15 15 14
18 16 15 15 15 15
17 16 15 15 14 14
17 17 16 13 13 13
18 17 16 16 16 15
17 16 16 13 13 13
17 16 14 14 14 13
17 17 16 15 15 15
16 16 15 13 13 13
16 16 14 14 14 14

Overall this looks more than a little bit on the heroic end of the scale, but in exactly the same way as the OPs example.

[Sir cryosin]

Wow I hate that load of crap " I use standard array because I like my character to have a weakness. Like he trips over a pebble because he has a 8 for Dex or she can read a map because she as a 8 for int." Cut that crap out a character with all 20's + can have weakness's like "my character will not attack anything that looks like a dog. Because he had to kill his own dog after it was being controlled by a druid that had it kill villagers. Or I had a character that was afrade of highest and we had to go up onto a tower on a cloud so I had to roll con checks and Wis checks going up and down the cloud stares to the tower. It just a excuse for people to with annoying an unoriginal character's.

Now with that rant out of the way. To the OP do what you like man. Your not going to hear anything you haven't before. I like X because ........ or X is better because..... Me as a player I injoy play a hero so I tend to like higher stats for my character. But as a DM I'm coming to likeing my players use standard array because I can as a DM effect there characters instead of them just steam rolling everything. It also helps players to cooperate with each other to cover stats that others are low in.

[ZorroGames]

No, I've probably misunderstood the method if that's what we're supposed to be doing. This is 3d6 rolled 36 times, picking the best 6 sets.

Ive got the spreadsheet working now so my apparently wildly mistaken method produces about about 150 16s, 340 15s, 320 14s, 130 13s, 30 12s, and the odd 11 once in awhile, out of 1000 trials (1 27 146 355 336 135 last run.) Where an 18 should have a 1 in 216 chance of happening I worry that I have done something terribly wrong in the early steps of the spreadsheet :/
Best guess? Maybe the Randbetween function secretly keeps track of the decimal place and I'm adding a bunch of 5.3s together. I really didn't expect to be hunting bugs in such a simple excel sheet, so I'll have to shelf this for awhile.

A campaign I never played in used that method. They never had more one natural 18 in a character's stats but I have Hot Die rolling get two or even three for a character just using 3D6 so my concern to keep those characters my friends had very uncommon. Why ruin such a great and true story?

[Lombra]

Rant:

What I don't like about stat rolling is the fact that some players start unnecessairly higher than other players. I play a PHB tiefling shadow monk (because who cares about optimization) and started the campaign at level 1 with 14 in DEX and 14 in wis, while the sorcerer and the wizard of the party started with 19s (post racial) in their spellcasting stat. Needless to say that I feel "nerfed", I'd like to pick so many flavour feats but I won't because stuning strikes don't hit by themselves and I need both WIS and DEX to land them. Not to mention the sub-par AC that the character has. (The sorcerer also has 16s in constitution and dexterity, at least the wizard has an 8 in strength to compensate... somehow...) I know the sorcerer sounds fishy but all the rolls have been made with everyone watching (it was painful). Ironically my character would be better if I just point-buyed him in.

[Knaight]

The way I understood the first post, rolling "3d6 x 6, pick best set of 3, per stat" was essentially what was being suggested. That was clarified (my reading was incorrect) later on by the OP.

Think I'll beer-batter that headgear :smalltongue:

Fortunately, AnyDice should also be able to handle this (although that five second runtime limit and I have not been getting along this thread, so we'll see). On the bright side, with this as a per stat method it compares to 4d6 best 3 pretty easily without looking at the whole array, so I can drop a 6d at the beginning.

output [highest 1 of 3d(3d6)]
output [highest 3 of 4d6]

The mean (per stat) of highest 1 of 6d(3d6) is 14.23 with a standard deviation of 1.77.
The mean (per stat) of highest 3 of 4d6 is 12.24, with a standard deviation of 2.85.

That actually is a fair bit stronger, and is roughly comparable to highest 3 of 6d6 (at 14.27 mean and 2.36 std).

[Snails]

Method we use:
1. Take the A-2-3-4-5-6 cards from a standard deck of playing cards.
2. Shuffle the 24 cards and deal into 6 piles
3. Name a stat, choose a pile, pick 3 of the 4 cards (i.e. pick "in order" in the classic tradition, which order does not matter)
4. You may swap two stats
5. If you do not like what you get, you may use the standard array instead

You get the fun of wondering what your stats will be.
You get the fun of improvising off of stats that are significantly random.
You are likely to get something very slightly better than standard array.
You are not stuck with bad rolls -- the standard array is available if the numbers are gelling into a concept you like.
Really high and really low stats across the board can never happen, so it keeps the stat disparities under control.
Since the stats are effectively boxed, the DM does not have to worry about power levels that are significantly different from standard array.
You are not automatically destined to have a dump stat.
You might get a 17 or 18, but you will definitely have a low stat or two in the bargain.

For most players/DMs, this method hits a sweet spot of gaining most of advantages and mitigating most of the disadvantages of rolling.

Re: #3: If you are a Real Loonie, you are allowed to pick the three low numbers, discarding the high. I recommend you check with the DM and others at the table if your gaming style will fit. YMMV. :smallbiggrin:

[Sir cryosin]

Rant:

What I don't like about stat rolling is the fact that some players start unnecessairly higher than other players. I play a PHB tiefling shadow monk (because who cares about optimization) and started the campaign at level 1 with 14 in DEX and 14 in wis, while the sorcerer and the wizard of the party started with 19s (post racial) in their spellcasting stat. Needless to say that I feel "nerfed", I'd like to pick so many flavour feats but I won't because stuning strikes don't hit by themselves and I need both WIS and DEX to land them. Not to mention the sub-par AC that the character has. (The sorcerer also has 16s in constitution and dexterity, at least the wizard has an 8 in strength to compensate... somehow...) I know the sorcerer sounds fishy but all the rolls have been made with everyone watching (it was painful). Ironically my character would be better if I just point-buyed him in.

Are you kidding me you are bitching about feeling nerfed because you choose the sub optimal choices.

[Lombra]

Are you kidding me you are bitching about feeling nerfed because you choose the sub optimal choices.

That's not the point, "nerfed" is a bad word, please try to understand what I actually meant (sub-par? Not effective?) and forgive my poor use of it. Still, if I rolled the same stats of the sorcerer for example, I'd have an 18 in dexterity, 15 in wisdom, 15 in constitution (edit: at level 1), and the rest of the stats would be all above 12. Now I know that this isn't actually an issue, but it makes me feel like my character is unfairly inferior to the other members of the party.

[Sir cryosin]

That's not the point, "nerfed" is a bad word, please try to understand what I actually meant (sub-par? Not effective?) and forgive my poor use of it. Still, if I rolled the same stats of the sorcerer for example, I'd have an 18 in dexterity, 15 in wisdom, 15 in constitution (edit: at level 1), and the rest of the stats would be all above 12. Now I know that this isn't actually an issue, but it makes me feel like my character is unfairly inferior to the other members of the party.

Dude that's the nature of the Beast you take what you get and roll with it. You don't playing a ball game and expect to get the trophy just for playing. Yes this is a co-op game. But rolling for stats is ment to give you random stats. Do you get mad at the barbarian for rolling high on his weapon damage dice. When you are rolling low on your weapon damage dice. Do you?

[Knaight]

Dude that's the nature of the Beast you take what you get and roll with it. You don't playing a ball game and expect to get the trophy just for playing. Yes this is a co-op game. But rolling for stats is ment to give you random stats. Do you get mad at the barbarian for rolling high on his weapon damage dice. When you are rolling low on your weapon damage dice. Do you?

"That's just the way it is" is roughly the single most useless statement that can be made in a discussion about whether something should have been designed the way it was. On top of that, comparing the results of an individual roll to the results of a set of rolls that systematically alters the individual fleeting rolls is fundamentally disingenuous.

[ZorroGames]

Rant:

What I don't like about stat rolling is the fact that some players start unnecessairly higher than other players. I play a PHB tiefling shadow monk (because who cares about optimization) and started the campaign at level 1 with 14 in DEX and 14 in wis, while the sorcerer and the wizard of the party started with 19s (post racial) in their spellcasting stat. Needless to say that I feel "nerfed", I'd like to pick so many flavour feats but I won't because stuning strikes don't hit by themselves and I need both WIS and DEX to land them. Not to mention the sub-par AC that the character has. (The sorcerer also has 16s in constitution and dexterity, at least the wizard has an 8 in strength to compensate... somehow...) I know the sorcerer sounds fishy but all the rolls have been made with everyone watching (it was painful). Ironically my character would be better if I just point-buyed him in.

You "... play a PHB tiefling shadow monk (because who cares about optimization) and started the campaign at level 1 with 14 in DEX and 14 in wis, " but you are unhappy with your stats. Hmm.

Was Die rolling... aka gambling... required? Chance is a fickle queen/paramour.

If you chose to die roll versus buying stats or standard array, so be it. Lady Luck urinated on your offering? It happens. Get a new character opportunity from your DM or change classes fast.

What were your pre-racial numbers? All of them. What about those numbers, all of them, for the sorcerer and wizard? Was there opportunity to ask for a binding opportunity to retool your numbers? Did you ask?

Optimization or flavor, only rarely can you have all of both that you want said the AD&D 1st Edition Halfing Fighter.

What level are you now? Have you hit your first ASI? I hope I have that acronym right. That might help and can't hurt.j

You feel "nerf'ed" as in life (the dice) treated your character unfairly. I don't want to sound unsympathetic to your plight but dice have a life of their own. Yes, in 5th it appears you can only do so much to fix a low stat situation but have you had opportunity to try and fix your challenges (starting with +1 De and +1 WI upgrades?)

:smalleek:

The question now is, do you start a new character, change class, or try and role play and gut it out? There may be other better options but I am still learning 5th Edition. I will remember your dilemma and keep it in mind when I get to start playing regularly.

[ZorroGames]

"That's just the way it is" is roughly the single most useless statement that can be made in a discussion about whether something should have been designed the way it was. On top of that, comparing the results of an individual roll to the results of a set of rolls that systematically alters the individual fleeting rolls is fundamentally disingenuous.

Uh, wait, how is stating the truth about dice rolls (random chance) fundamentally disingenuous?

"... should have been designed that way..." ? I thought the discussion was about the OP methodology tried. Whether or not it "should" have been designed that way is ludicrous to me. More important to me is whether it creates viable, playable characters that are neither demigods or glorified children with sticks. I have concerns about how viable it is, not how it "should" (by what criteria "should") be designed.

I have so far, in my limited experiments found ithe OP method slightly higher skewed than standard methods 5th but any chance 3D6 technique can produce high scores overall or in part (seen 2x18 and 3x18) sooner or later. Also rolled a 3D6 character with no stat higher than an 8. DM looked and said, "Died of plague in childhood," than handed me an eraser.

Random or pseudo-random (electronic dice) is just that, "Luck of the roll." It IS just the way it is. Not "fair" in outcomes at all.

[ZorroGames]

No, I've probably misunderstood the method if that's what we're supposed to be doing. This is 3d6 rolled 36 times, picking the best 6 sets

Snip


Here's the first 10 arrays
17 16 16 15 15 14
18 16 15 15 15 15
17 16 15 15 14 14
17 17 16 13 13 13
18 17 16 16 16 15
17 16 16 13 13 13
17 16 14 14 14 13
17 17 16 15 15 15
16 16 15 13 13 13
16 16 14 14 14 14

Overall this looks more than a little bit on the heroic end of the scale, but in exactly the same way as the OPs example.

OH MY!

If those were physical dice rather than virtual I would be making an offer to buy them.

:smallcool:

[Lombra]

You "... play a PHB tiefling shadow monk (because who cares about optimization) and started the campaign at level 1 with 14 in DEX and 14 in wis, " but you are unhappy with your stats. Hmm.

Was Die rolling... aka gambling... required? Chance is a fickle queen/paramour.

If you chose to die roll versus buying stats or standard array, so be it. Lady Luck urinated on your offering? It happens. Get a new character opportunity from your DM or change classes fast.

What were your pre-racial numbers? All of them. What about those numbers, all of them, for the sorcerer and wizard? Was there opportunity to ask for a binding opportunity to retool your numbers? Did you ask?

Optimization or flavor, only rarely can you have all of both that you want said the AD&D 1st Edition Halfing Fighter.

What level are you now? Have you hit your first ASI? I hope I have that acronym right. That might help and can't hurt.j

You feel "nerf'ed" as in life (the dice) treated your character unfairly. I don't want to sound unsympathetic to your plight but dice have a life of their own. Yes, in 5th it appears you can only do so much to fix a low stat situation but have you had opportunity to try and fix your challenges (starting with +1 De and +1 WI upgrades?)

:smalleek:

The question now is, do you start a new character, change class, or try and role play and gut it out? There may be other better options but I am still learning 5th Edition. I will remember your dilemma and keep it in mind when I get to start playing regularly.

That was just a rant, the character is level 6 now and at level 4 I pumped DEX to 16. I am glad that the sorcerer and the wizard now got their 20 in their spellcasting stat, the wizard picked keen mind, so we never get lost in the wilderness either. My character is actually pretty strong now (no magic items for him yet), but the numbers hurt my eyes when I see them next to the sorcerer's character sheet. Proficiency bonus is a godsend, it works brilliantly whenever it's applied. I did get the chance to reroll, but the lowest number rolled was a 9, so I figured that I might as well just keep it since no negative modifiers felt like a good choice (on top of that I didn't even knew point buy existed at that time, and we always rolled for stats in past campaigns in past editions). The point is that I'll never be able to max dexterity and wisdom, and that there basically is no place for feats. I would have loved to pick magic initiate, mobile, picking dual wielder to raise the AC beyond 20/20 dex/wis, but I'll never get to pick those things, and since I'm a completionist kind of guy, it bothers me. The character flows wonderfully, it's just my OCD that makes me suffer.

[Kurt Kurageous]

To often the "I totally rolled 4 18s" comes up.

As DM I as so sick of this I insist on STD array. I once had a group who seemed to insist on doing it their own way. It lasted two sessions.

I now insist on standard array and give a feat at 1st, or you can start at third with no feat. They'll all be the same level soon enough.

If I ever encounter "I swear, I rolled it" again, I will let it in, then I will paint a bullseye on that character until they are dead. And I'd call it "The Curse of Mechanus."

[ZorroGames]

That was just a rant, the character is level 6 now and at level 4 I pumped DEX to 16. I am glad that the sorcerer and the wizard now got their 20 in their spellcasting stat, the wizard picked keen mind, so we never get lost in the wilderness either. My character is actually pretty strong now (no magic items for him yet), but the numbers hurt my eyes when I see them next to the sorcerer's character sheet. Proficiency bonus is a godsend, it works brilliantly whenever it's applied. I did get the chance to reroll, but the lowest number rolled was a 9, so I figured that I might as well just keep it since no negative modifiers felt like a good choice (on top of that I didn't even knew point buy existed at that time, and we always rolled for stats in past campaigns in past editions). The point is that I'll never be able to max dexterity and wisdom, and that there basically is no place for feats. I would have loved to pick magic initiate, mobile, picking dual wielder to raise the AC beyond 20/20 dex/wis, but I'll never get to pick those things, and since I'm a completionist kind of guy, it bothers me. The character flows wonderfully, it's just my OCD that makes me suffer.

Well maybe OCD (should that be CDO?) but FDS (F****** Dice Syndrome) had a huge impact.

This thread and the Monk on es are really opening my eyes/minds to how 5th differs from OD&D/AD&D 1st!

Must adjust thinking radically before my first game...

BTW, I am typing from my new Hard Drive/New Logic Board iMac. Bookmarks? What Bookmarks? Sigh.

[Pex]

As DM I as so sick of this I insist on STD array. I once had a group who seemed to insist on doing it their own way. It lasted two sessions.

I now insist on standard array and give a feat at 1st, or you can start at third with no feat. They'll all be the same level soon enough.

If I ever encounter "I swear, I rolled it" again, I will let it in, then I will paint a bullseye on that character until they are dead. And I'd call it "The Curse of Mechanus."

The solution is to have all rolls made with everyone together not don't use dice rolling at all. You can keep the baby while throwing out the bathwater.

[JAL_1138]

Fortunately, AnyDice should also be able to handle this (although that five second runtime limit and I have not been getting along this thread, so we'll see). On the bright side, with this as a per stat method it compares to 4d6 best 3 pretty easily without looking at the whole array, so I can drop a 6d at the beginning.

output [highest 1 of 3d(3d6)]
output [highest 3 of 4d6]

The mean (per stat) of highest 1 of 6d(3d6) is 14.23 with a standard deviation of 1.77.
The mean (per stat) of highest 3 of 4d6 is 12.24, with a standard deviation of 2.85.

That actually is a fair bit stronger, and is roughly comparable to highest 3 of 6d6 (at 14.27 mean and 2.36 std).

It seems my hat survives another day! :smalltongue: Also, thank you for running the number crunching; I'm rubbish with probability math.


The solution is to have all rolls made with everyone together not don't use dice rolling at all. You can keep the baby while throwing out the bathwater.

Right. Just require the rolling to be witnessed by the DM. And preferably don't touch the dice after rolling until the DM has seen and logged the rolls. There's no need to ban rolling outright.

[ZorroGames]

So, what have I learned from this thread?

Lots of stuff but the metadata (for lack of a better word) is:

Want a stable good starting character? Standard array.

Want to craft a character towards a class optimized by racial adds? Point buying.

Feeling lucky and willing to accept a possibly suboptimal character? Talk to you DM, decide what method is to be used, and roll the dice in front of him/her and the other players ideally.

Thanks to all crunched numbers, spoke about character creation philosophy, shared their experiences. So very worth you efforts in my mind.

[Knaight]

It seems my hat survives another day! :smalltongue: Also, thank you for running the number crunching; I'm rubbish with probability math.

I assure you, the only skills displayed on my end is that I've used Anydice enough to figure out the Xd(YdZ) trick - crunching these numbers by hand would have sucked.