06-07-2013, 03:52 PM | #1 |
Join Date: Jan 2006
|
Bell Curve on 1D6 [math]
One of the things that bugs me about GURPS is that the damage probability distributions vary so much with how many dice you roll.
The worst offender is the 1d6 - if it was just a bit more bell-shaped, it would make me happier. So here's my idea: When rolling 1d6 damage, actually roll 3d6 and take the middle value. I don't think this is too hard. Code:
1 16 7.4% 2 40 18.5% 3 52 24.1% 4 52 24.1% 5 40 18.5% 6 16 7.4% It's not as sharply belled as 3d6, but it is belled. Last edited by martinl; 06-07-2013 at 03:56 PM. Reason: premature postulation |
06-07-2013, 03:53 PM | #2 |
Wielder of Smart Pants
Join Date: Aug 2004
Location: Ventura CA
|
Re: Belling a 1D6 [math]
That the damage probability distributions, what?
Have you tried Armor as Dice? Penetration is fixed but injury is randomized. |
06-07-2013, 03:57 PM | #3 |
Join Date: Jan 2006
|
Re: Belling a 1D6 [math]
|
06-07-2013, 04:01 PM | #4 |
Join Date: Jan 2006
|
Re: Belling a 1D6 [math]
|
06-07-2013, 04:28 PM | #5 |
Join Date: Dec 2008
Location: Cumberland, ME
|
Re: Bell Curve on 1D6 [math]
Another way to approach it would just be to roll 3d6, divide the result by 3, and round to the nearest integer. That yields the following distribution:
Code:
1 1.85% 2 14.35% 3 33.80% 4 33.80% 5 14.35% 6 1.85% 2d6 also has a bit of odd behavior, namely that it's a linearly-changing distribution (each level of probability is the same roughly-2.78% change from those on either side of it). Unfortunately, taking the "multiply by 2, divide by 3" approach yields weird results because of how the rounding works (for example, you are less likely to get a result of 4 than you are a result of 3 or a result of 5... the whole "curve" is a bit spiky). If you're bothered by 2d distributions, you can take a similar approach to your original suggestion: Roll 3d6, and instead of taking only the middle result, omit only the middle result. That yields: Code:
2 0.46% 3 2.78% 4 6.02% 5 11.11% 6 17.13% 7 25.00% 8 17.13% 9 11.11% 10 6.02% 11 2.78% 12 0.46% |
06-07-2013, 05:03 PM | #6 |
Hero of Democracy
Join Date: Mar 2012
Location: far from the ocean
|
Re: Bell Curve on 1D6 [math]
I'm enough of a math nerd that the changing of the curve shape even from 3d to 5d bugs me. I've tried a variant where damage is read as an "average" amount (1d becomes 4) and everything is multiplied by (2d-2)*.2. It requires a bit of math, but most of my play is by post, so it yields really nice results, if you don't mind worrying about tenths of points of damage (I don't, but your milage may vary).
__________________
Be helpful, not pedantic Worlds Beyond Earth -- my blog Check out the PbP forum! If you don't see a game you'd like, ask me about making one! |
06-07-2013, 05:40 PM | #7 |
Join Date: Feb 2005
Location: Berkeley, CA
|
Re: Bell Curve on 1D6 [math]
I've pondered going to a single log-based roll system -- say, make a skill check and look up the margin of success on a table, or roll 3d6+n and look up the results. This allows you to have granularity between 1d6 and 1d6+1, which is nice, but the table tends to be ugly, because you need fairly large numbers unless you want damage to be really really random. For example, this MoS system works out to variability comparable to 3d6:
Code:
MoS F 0 5 8 11 14 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 +30 Dam 1 2 3 4 5 6 7 7 8 9 10 11 12 13 14 15 16 17 18 20 21 23 25 27 29 31 34 37 40 43 46 50 54 58 63 68 74 80 86 93 x10 |
06-07-2013, 06:41 PM | #8 |
Join Date: Feb 2013
|
Re: Bell Curve on 1D6 [math]
Say more about log rolling. Interest piqued.
|
06-07-2013, 06:42 PM | #9 |
Join Date: Feb 2005
Location: Berkeley, CA
|
Re: Bell Curve on 1D6 [math]
|
06-07-2013, 11:26 PM | #10 | |
Join Date: Dec 2007
Location: Brooklyn, NY
|
Re: Bell Curve on 1D6 [math]
Quote:
__________________
-JC |
|
Tags |
math, modifying dice + adds |
|
|