1d20 vs 3d6

Rolled 6, 1, 14, 12, 13, 2 = 48 (6d20)

You're on.

That could be interesting way to play D&D. How advantage and disadvantage would work?

I'm retarded, please be patient with me.

I think storyteller does.

Higher crit chance without the swingy randomness of 1d20. 2d6+1d8 has an interesting curve that works great depending on modifiers.

Lol, had to look this up

Barbarian here i come

Rolled 16, 19, 20, 1, 1, 13 = 70 (6d20)

Git gud.

Your dex is so low you actively jump into the attacks

>said the retarded adolescent barbarian

I give up.

Why not 9d2s?

The reason for 3d6 was to emulate 1d20 with normal dice.

There is no "magical thing that made us pick 3d6"

I have considered it for homebrew. I gues it's good if you want some bell-curve but not as extreme as 3d6. Also, I like how all outcome probabilities are whole percent.

user.

Jesus, user.

No.

People actually wanting a bellcurve probability is a thing.

Advantage is take highest, and disadvantage is take lowest. Just like normal.

>user.
>Jesus, user.
>No.

I never said 3d6 emulate 1d20, stupid people decided to use 3d6 because "urr durr thats a ok substitute"

Assuming you want to NOT use 1d20 because of non linearity and not because of how easy is to find d6, 2d10 -1 would be a good thing.

2d10-1 Its not linear, and there is 1% of getting a 1, 2% of getting a 2, 3% of getting a 3... up to 10% of getting a 10 and then 9% of getting a 11, 8% of getting a 12......

Wrong.

With rolling three dice that sounds pretty rough in both ways, though.

the chance of getting the smallest number is 100/x^3, X is the number X at 3dX
as some example the chance of getting a 1 in 3d6-2 is 100/63=100/216

the second number has + 2*(100/x^3) chances of being selected than the first number
as some example at 3d6-2 the chance of 2 being selected is (100/6^3)+(2*(100/6^3)) or 3*(100/6^3)

After it, the chance of number N being selected compared to N -1, is (100/x^3) plus than the chance of N-1 being selected instead of N -2
as some example at 3d6-2 the chance of 2 being selected is the chance of 1 being selected + (2*(100/6^3)), and the chance of 3 being selected is the chance of 2 being selected + (2*(100/6^3)) +(100/6^3)
This kind of stuff stop after the X digit, at 3d6 -2, the X digit is 6

After it stop, the chance of number n being selected compared to N-1 is (100/x^3)*2 plus than the chance of N-1 being selected instead of N -2
No examples here because laziness
This happens x/2 digits.
After x/2 digits, the chance of the next digit is the same and then it decreases by this same value for more x/2, than it continue decreasing by usual (100/x^3) until the last possible digit

It's pretty brutal, yeah. Much higher than the +/- 3.82 that dis/advantage gives.

Rolled 2, 11, 6, 19, 14, 11 = 63 (6d20)

Rollang

>I never said 3d6 emulate 1d20
>The reason for 3d6 was to emulate 1d20 with normal dice.

Uh huh.

I'm going to take a wild swing and guess you've never taken a probability and statistics course. Or looked at the bonus/penalty behavior of GURPS's 3d6 roll.

>>The reason for 3d6 was to emulate 1d20 with normal dice.
>Uh huh.

People decided "lets use 3d6 to have 1d20 with d6 that are common dices"

I DONT think 3d6 emulate 1d20, and HATE non linear dices (prefer the simplicity to convert real life data to linear dice) and would never use 3d6

see

>19
What? I mean he might anyway because he's pants on head retarded, but low dex is not an issue.

Rolled 6, 2, 3, 6, 5, 6, 5, 3, 5, 6, 6, 5, 3, 3, 1, 3, 5, 6 = 79 (18d6)

Roll

I really like that system.

All the other character building stuff gives you one consistent to hit stat, and then you roll three dice everytime and pick the middle unless you're in your class's favoured position or weak position.

Then skills work off your regular stats but you rank up which skills you're average and then good at. Giving you the middle and then high roll.

>off by one
Wish I could roll like that IRL

Except real life doesn't work on a linear distribution. It works on a gaussian distribution. Quite a lot maps to bell curves, which you will not get in the slightest with a d20.

Are you calling me a faggot?

Rolled 3, 14, 8, 3, 18, 15 = 61 (6d20)

>Except real life doesn't work on a linear distribution
You can convert gaussian distribution to a linear one

And it doesn't work when you do that. On a 3d6, 3 and 18 can occur once: 1, 1, 1 or 6, 6, 6.
4 and 17 can occur 3 times: 1, 1, 2; 1, 2, 1; 2, 1, 1 or 6, 6, 5; 6, 5, 6; 5, 6, 6.
10 and 11 both can occur 27 times. I won't write those out.

On a d20, 1, 2, 10, 11, 19, and 20 both occur once. You see how there's no average here? How it doesn't accurately model a gaussian distribution because there's no mode here, the d20 doesn't tend towards anything on an individual roll. Every roll has an equal chance of coming up any number. While it may be statistically likely that after 10,000 rolls, 10.5 comes up as the average, that isn't how it works at the individual (read: game) scale.

at 3d6

3 happens 0.46% of the time
4 1.39%
5 2.78
6 4.63
7 6.94
8 9.72
9 11.57
10 12.50
11 12.50
12 11.57
13 9.72
14 6.94
15 4.63
16 2.78
17 1.39
18 0.46%

How is this can't be represented with a linear dice?

Because a d20 doesn't have a mode. The average on a d20 is not average, as it's as equally likely to occur as the extremes. 9-12 happen 1/5 times, just like 1-4 and 17-20, while on a 3d6 there is a distinct mode where 9-12 happen 104/216 times. Did you even read my post? Do you understand probability?

0-0.46 = 3
0.47-1.85 = 4
1.86-4.63 = 5
4.64-9.26 = 6
9.27-16.20 = 7
16.21-25.93 = 8
25.94-37.50 = 9
37.51-50.00 = 10
50.01-62.50 = 11
62.51-74.07 = 12
74.08-83.80 = 13
83.81-90.74 = 14
90.74-95.37 = 15
95.38-98.15 = 16
98.16-99.54 = 17
99.55-100.00 = 18

Rolled 2, 6, 17, 9, 20, 14 = 68 (6d20)

Let me try

Rolled 2, 11, 8, 13, 19, 4 = 57 (6d20)

I'm either going to be glad I'm not in Vegas, or I'm going to regret not being in Vegas.

Depends how you want probability to work. Games like 5e are notoriously swingy with the dice having a huge say in how things turn out. Some players enjoy the possibility of pulling off something unlikely, others dislike having their half-ogre barbarian fail to kick down a door that the party wizard handles with ease.

I've seen 2d10 proposed as a way to make 5e less random. It's not a terrible idea if your players are on board. Add a third dice to handle advantage/disadvantage and discard the appropriate die. Or still roll two dice but double the higher/lower one to get your result.

>they haven't seen the light that is 1D6 exploding on 3+

why not flip a coin?

>as it's as equally likely to occur as the extremes. 9-12 happen 1/5 times, just like 1-4 and 17-20, while on a 3d6 there is a distinct mode where 9-12 happen 104/216 times

What matter is the % of getting the dice result, not the final number you get at the dice, you wouldnt compare the chance of getting a 9-12 on a 1d20 witht he chance of getting a 9-12 in a 3d6, because 9-12 on 1d20 is 4 out of 20, while 9-12 at a 3d6 is 104 out of 216

a 4 in a 3d6 is not a 4, but (1,1,2), (1,2,1), (2,1,1), its 3 different dice rolls.

While in a 1d20 both the 3 and 4 are just one dice roll, and so they will happens the same amount of time as each other since they are the same amount of different dice rolls.

Yes. I know. What you are explaining to me, what I already know, shows why a gaussian distribution doesn't work on a d20, because the d20 itself does not tend towards any result. Every face has an equal chance of occurring. That is not like a gaussian distribution at all, where results tend towards the average result, 50%. 3d6 tends towards the center by having those center numbers (10 and 11) having the highest number of possible combinations of dice out of any other number, and the further you get from those two number, the fewer possible combinations there are. This is not the case with a single die, because every number has an equal chance of occurring. This is why a gaussian distribution cannot be mapped to a d20. Because the d20 does not discriminate against which number it rolls. All numbers are as likely as any other.

>not 3d20 pick the middle one
>not having a advantage/disadvantage system
Do you even roll?

>Because the d20 does not discriminate against which number it rolls

But you don't roll specific numbers, if you want something to have a higher chance you would group them together.

If you want that something has 3 times more than a 1 on a 1d20, you would group as some example 2, 3 and 4 (on a 1d20) as the same thing

Storyteller is a d10 based dice pool system.

>you would group as some example 2, 3 and 4 (on a 1d20) as the same thing
To explain better it would have no difference b etween rolling a 2, 3 and 4, like there is no difference in 3d6 if you roll a (2,1,1) (1,2,1) or (1,1,2)

Then you can't use modifiers.

Because people don't understand mapping.

>muh distribution.

>dices

>multiple set of die is called dice
>multiple set of dice is called dices
Pick two.

Okay. Remap everything on a d20 that the 3d6 does naturally so you can get a gaussian distribution on it. Why? How is that simpler than just rolling 3d6, which are gaussian by default? Actually, does any system map different distributions to d20s? I've never heard of one that does this, probably because it's a pain in the ass to memorize a table compared to just rolling dice that already give you the distribution you want.

>How is that simpler than just rolling 3d6, which are gaussian by default?

Because if you want simple percentages at at any moment , you can easily use it

Can you give an example of how a gaussian distribution on a d20, d1000, whatever gives a simple percentage at any moment compared to 3d6? I can't think of any.

I'm pretty sure he means "simple percentages" as in "not gaussian".

Rolled 11, 15, 7, 1, 8, 9 = 51 (6d20)

Fuck ye.

Fuck no

SLA Industries

>trying to argue with D20tards
user...

Rolled 10, 5, 9, 6, 5, 4 = 39 (6d10)

Let's go

Rolled 6, 2, 4, 8, 5, 11 = 36 (6d20)

SDCIWC

2d10 is my favorite roll and it's so rarely used.

I really don't know what that means. If he means rolling a d1000 where 1-46 is 3, 47-139 is 4, 140-278 is 5, etc., I don't see how that's in any way, shape, or form better for play than just rolling 3d6.

How is he converting 0.46% chance to a d20, or a d100? The smallest % you can get on either is 5% and 1%, respectively. He said you can convert a gaussian distribution to a linear one, and the only way I'm seeing of doing that is creating ranges on a d100 which requires you to know the percentages of a 3d6 to know your result, or look up unintuitive result bands on a table, when you could just roll 3d6 and be done with it.

Please tell me I'm horribly wrong here and am completely misunderstanding him, because I really want to be misunderstanding him right now.

How are either of us to improve if we don't challenge each other's ideas? Am I to let each fester in their respective echo chambers?

I much prefer to use 2d10 to emulate a d100
FFG Warhammer 40k, oh yeah.

Because 3d20 take middle is master race.

>1d20 and 3d6
>average roll of 10.5
>2d10
>average roll of 11
Other acceptable roll types include:

1d12+1d4+1d2
1d10+2d4
1d10+1d6+1d2
1d8+4d2
1d8+1d6+1d4
2d8+1d2
1d6+1d4+3d2
3d4+2d2
7d2

>7d2
MASTER RACE

Doesn't the Fantasy Flight 40k RPGs like Only War and Rogue Trader use 2d10?

I can be, just need a dice with enough faces

All d100 games tell you to emulate d100 with 2d10, anima, rolemaster, warhams, etc...

D10 are not platonic solids, they shouldn't exist to begin with

4d6-4

Rolled 19, 16, 7, 3, 16, 4 = 65 (6d20)

1d20 gives a flat distribution and equal chance of all number.
3d6 gives a bell curve and is the most common die type that people will have access to.

2d10 gives you a very angular distribution, not something you want to see if you are attempting a bell curve. And not something you want to see if you are looking for even distribution, either. It also uses unusual dice, so unless you regularly play World of Darkness/Legend of the Five Rings, most people probably won't have two d10 dice readily available.

That said, if you want to see a system which uses it: Eclipse Phase. That system uses a d% system, but it also uses 1d10 and 2d10 for non-skill based rolls, such as weapon damage.

I've seen 5d4-2 proposed, as a way to replicate the 3-18 curve.

>platonic solid
Do I look like someone who cares?

>not being a 3d20m patrician
PLEBEIANS ALL OF YOU

>most people probably won't have two d10 dice readily available.
Or any d100% system.
Or ever bought your dice in a set

Rolled 19, 16, 14, 20, 9, 10 = 88 (6d20)

let's do dis

>19 str
>16 dex
>14 con
>20 fucking int
I have become an ubermensch

I wouldn't be opposed to it. 1d20 is best if you want the concept of a critical to be able to exist well, but 2d10 works fine. 3d6 is pussy shit, 5d4 or get out of my face with you're middle of the road. If you were so concerned with "muh bell curve" you'd roll 10d3-10 every time.

Reminder that a bell curve doesn't do anything except make modifiers inconsistently powerful. +5 is always +20% with 1d20. It's a moving target with any multiple dice roll. If you want a less han readable system for no justifiable reason, fine. There are far better ways of making a system more consistent. Like taking 10. Or...? Not rolling when the result is obvious.

these guys know what's up

Indeed. It gives more stability than 1d20 but has more swinginess than 3d6. Seems like a good compromise.

Rolled 17, 14, 12, 20, 2, 3 = 68 (6d20)

inb4 a character with sub-5 and borderline autistic stats.

>20 int
>2 wis
Well, you got the "autistic" part down pat.

Rolled 15, 10, 6, 15, 4, 6 = 56 (6d20)

I'm a human fighter.

>Do I look like someone who cares?
You probably should.
youtube.com/watch?v=8UUPlImm0dM

In short: the surface a die lands out changes how it tumbles. This doesn't matter much for a "fair" die since all sides are equal, but your 5-sided or 7-sided die is ultimately going to have some sides favored due to the surface you roll them on.

You might want to look into barrel dice or something like a top, if you want a fair 5-sided object.

Looks like a good spread for a ranger.

How would you do adv/dis with 3d20m? Just high/low?