New Take on bell curve rolling

[Dairian]

Ive never really been happy with rollign d20s, the range is just too great. But at the same time, the conventional bell curve rolling methods(3d6, 2d10) have too narrow a range, so ive decided in my games to roll a d12 and a d8. this makes rolling a 9-13 equally likely, meaning that you can't just assume an 11 like with the other methods.

I just wanted to se how you guys liked this method, any feedback would be appreciated.

[Samurai Jill]

It's a fresh approach, certainly.

[CarpeGuitarrem]

Very interesting. Lemme see how this plays out...(using "a" for a roll of 10, "b" for a roll of 11, "c" for a roll of 12)

02: 11
03: 12, 21
04: 13, 31, 22
05: 14, 41, 23, 32
06: 15, 51, 24, 42, 33
07: 16, 61, 25, 52, 34, 43
08: 17, 71, 26, 62, 35, 53, 44
09: 18, 81, 27, 72, 36, 63, 45, 54
10: 19, 28, 82, 37, 73, 46, 64, 55
11: 1a, 29, 38, 83, 47, 74, 56, 65
12: 1b, 2a, 39, 48, 84, 57, 75, 66
13: 1c, 2b, 3a, 49, 58, 85, 67, 76
14: 2c, 3b, 4a, 59, 68, 86, 77
15: 3c, 4b, 5a, 69, 78, 87
16: 4c, 5b, 6a, 79, 88
17: 5c, 6b, 7a, 89
18: 6c, 7b, 8a
19: 7c, 8b
20: 8c

Holy cow, that's a nice curve. I like it.

[Catattack]

Or you can use 1d8 + 1d8 + 1d6 - 2.
That has the same range as a d20 but with more curve.
Or use 1d8 + 1d8 + 1d6 + 1d20 - 2 round down.
It gives a curve still a bit random, more like a banana curve than bell.

[Daracaex]

Or you can use 1d8 + 1d8 + 1d6 - 2.
That has the same range as a d20 but with more curve.
Or use 1d8 + 1d8 + 1d6 + 1d20 - 2 round down.
It gives a curve still a bit random, more like a banana curve than bell.

I think the point is to generate a bell curve, not make rolling into a second-grader's math test.

[Khanderas]

Very nice and I like it as this evens out the random quite a bit. Quite intresting as we get the same number of combinations for results 9-13 (not to be confused with the same % chance to get a 9, 10, 11, 12 or 13).

I note there are no "1" rolls, but that is problebly for the best.

[WinterSolstice]

wouldn't a d8 and d12 mean the possibility of a score of 20 before racial modifiers?

[Khanderas]

wouldn't a d8 and d12 mean the possibility of a score of 20 before racial modifiers?

It's 1d12 + 1d8 instead of 1d20
As in attackrolls, skillchecks and so on. Not the 3d6 for starting stats.

a d20 have 5% to land on any number. But the more (and smaller) dice involved the greater the chance of getting a number in the middle. Something I think is good, since that puts more "worth" in getting the skill up as opposed to just have a flat 5% to score a nat 20.


Edit Idonnohowmany: How would you handle a "take 20" here ? since the chance to get a 20 in this system is not 5%, but 1,04%

[averagejoe]

(not to be confused with the same % chance to get a 9, 10, 11, 12 or 13).

Am I missing something. I'm pretty sure that is what it means.

(I should probably be completely sure, but I'm not completely sure about anything when I've been up all night.)

[Khanderas]

Am I missing something. I'm pretty sure that is what it means.

(I should probably be completely sure, but I'm not completely sure about anything when I've been up all night.)

Well I don't feel mathematically fit (slept bad, but drank coffee, so percived reality may differ from actual reality, / feel awake but problebly not at 100%) but I thought that since the chance to roll.... say a 5 on a d12 differs from the chance to roll a 5 on the d8. I thought that it would problebly scewer the results slightly.
For example, 2d10 max result 20.... chance of that is (1/10)*(1/10) = 1%
1d12 + 1d8 max result 20, chance of that 20 is (1/12)*(1/8) = 1,04166666... %

Pointlessly small difference anyways. And I may still be wrong, but that is the reason I thought that while there are the same amount of combinations for the numbers 9-13, those combinations might not have the same chance to pop up. Problebly wrong... because it feels wrong when I look at it again and it problebly is a straight
(nr of possible combinations for nr X / total combinations possible)*100
for the % chance for any given number.

[Studoku]

Well I don't feel mathematically fit (slept bad, but drank coffee, so percived reality may differ from actual reality, / feel awake but problebly not at 100%) but I thought that since the chance to roll.... say a 5 on a d12 differs from the chance to roll a 5 on the d8. I thought that it would problebly scewer the results slightly.
For example, 2d10 max result 20.... chance of that is (1/10)*(1/10) = 1%
1d12 + 1d8 max result 20, chance of that 20 is (1/12)*(1/8) = 1,04166666... %

Pointlessly small difference anyways. And I may still be wrong, but that is the reason I thought that while there are the same amount of combinations for the numbers 9-13, those combinations might not have the same chance to pop up. Problebly wrong... because it feels wrong when I look at it again and it problebly is a straight
(nr of possible combinations for nr X / total combinations possible)*100
for the % chance for any given number.

Each combination is based on a d8 and a d12 roll, so it has a 1/(8*12) : a 1/96 chance of rolling.

[Khanderas]

Each combination is based on a d8 and a d12 roll, so it has a 1/(8*12) : a 1/96 chance of rolling.

Yeah. that is how it works, but my low-on-sleep-high-on-coffee brain tricked me into overcomplicating it.

[Dairian]

How would you handle a "take 20" here ? since the chance to get a 20 in this system is not 5%, but 1,04%

I dont allow taking 10 or 20.

[ShneekeyTheLost]

Reminds me of a dice system I was considering using in a game I am contemplating actually writing up.

Basically, ranks in a skill determine how many dice you get to roll to overcome difficulty. Kind of like the WW system, in that it uses a number of d10's, only instead of each die trying to overcome a number, the total has to overcome a number. So if the DC was a 40, then you'd need at least 5-6 ranks to be able to have a chance in heck of hitting it, but at 8-9 ranks, it wouldn't be too difficult.

Also reminds me of the 'other' way of doing percentiles: 11d10-10.