Probability Distributions: Characteristics, Skills, Difficulty & Success

By [unnamed sender] ·

TML message #4371From bundle 362

Here's a bunch of general gaming stuff stirred up from the idea that GDW is
revamping Traveller:

Here's a general gaming simulation question:

I've homebrewed a couple of rules sets, but I've never been able to
definitively decide what is the best (=realistic, for now) way to handle the
different probabilities of: Characteristics, Skills, Difficulty of a Task, and
Success.

Different dice give different ranges of numbers, and different probability
curves:

   linear, same chance for any number, 1 die, or percentile dice

   bell curve, middle numbers are most likely because there are more
     combinations which result in them, multiple dice added together.

   asymtopic, a *very* high chance on one end, a very *low* chance on the other
   end, usually done by needing a certain number or higher to succeed, say, on
   a d100, 30 or higher is a good chance, 80 or higher on d100 is a low chance,
   99 or is a *tiny* chance.

Common Dice Combinations:

1d4  = 1 -  4, lin
1d6  = 1 -  6, lin
1d8  = 1 -  8, lin  2d4  = 2 -  8, bs
1d10 = 1 - 10, lin
1d12 = 1 - 12, lin  2d6  = 2 - 12, bs  3d4  = 3 - 12, vbs
                    2d8  = 2 - 16, bs                      4d4  = 4 - 16, vvbs
                                       3d6  = 3 - 18, vbs
1d20 = 1 - 20, lin  2d10 = 2 - 20, bs
                    2d12 = 2 - 24, bs  3d8  = 3 - 24, vbs  4d6  = 4 - 24, vvbs
                                       3d10 = 3 - 30, vbs
                                                           4d8  = 4 - 32, vvbs
                                       3d12 = 3 - 36, vbs
                    2d20 = 2 - 40, bs                      4d10 = 4 - 40, vvbs
                                                           4d12 = 4 - 48, vvbs
                                       3d20 = 2 - 60, vbs
                                                           4d20 = 4 - 80, vvbs
1d100= 1 -100, lin

Chance to roll a target number of higher:

       1         PERCENT CHANCE
       0         9         8         7         6         5         4         3         2         1
Dice   0987654321098765432109876543210987654321098765432109876543210987654321098765432109876543210987654321

1d4  : 1                        2                        3                        4
1d6  : 1                2                3               4                5                6
1d8  : 1           2            3           4            5            6           7            8
1d10 : 1         2         3         4         5         6         7         8         9          0
1d12 : 1        2       3       4        5       6       7        8       9       0        1       2
1d20 : 1    2    3    4    5    6    7    8    9    0    1    2    3    4    5    6    7    8    9    0

Is there a better way to display this?

2d4  :
2d6  :
2d8  :
2d10 :
2d12 :
2d20 :

3d4  :
3d6  :
3d8  :
3d10 :
3d12 :
3d20 :

4d4  :
4d6  :
4d8  :
4d10 :
4d12 :
4d20 :

The proof is left as an exercise for someone else. :-)

Notice that rolling 4 or higher on a 1d4 is 25%; there is no 0% chance number
unless the target is shifted off the range, and then it is abruptly 0% chance.


Having been pedantic, my question is:  what combinations of number ranges and
distribution of probabilities for characteristics, skills, difficulties, and
success work well together?

  Characteristic:  Strength, Dexterity, Intelligence, etc.  Probability should
  be bell shaped.  Characteristics are generally fixed attributes that everyone
  has some score in, with most people having median scores, regardless of
  training, etc.  General aptitude...  2d6 and 3d6 are common.  I typically use
  3d6.

  Skill:  Amount of study or experience.  high scores should become exceedingly
  rare as the score increases.  ie 1/2 people might have score 1 in a common
  skill, 1/4 might have 2, 1/8 might have 3, 1/16 might have 4.  I have
  considered increases in skill to be linear, being asymtopically harder to
  achieve, and rarer.  I typically have skill levels 1-6.

     Should each increase of skill be linear equal advantage, or asymtopic?
     Should Each increase in skill be harder to achieve?  How much harder?

  Difficulty:  How difficult a task is:  Typically, I have considered difficulty
  as being either linear, or perhaps mildly asymtopic.  I usually use a range of
  1d20 to rate difficulties.

     Should each increase of difficulty be linear, or asymtopic?

  Success:  chance of success at a given task:  Generally asymtopic.  I usually
  use a 1d20, and require a target number of higher for success.

     How should the chance of success be affected by a higher or lower skill or
     difficulty?  Should it be asymtopic?

What different ways can characteristic, skill, difficulty, and success ratings
be combined?

  I sometimes ask characters to roll a characteristic or higher then the
  difficulty to succeed at something, if it's a task which has no applicable
  skill.  The Characteristic 3d6 = 1 to 18 compared to a 1 - 20 difficulty
  rating seems to work well.

  For tasks which require a skill, the 1 - 20 success roll must be under the 1-6
  skill rating + a 1d6 random chance roll.


  For Tasks more applicable to a characteristic:   Roll a 1d20 success die; roll
  under the 3d6 characteristic + maybe a 1d6 skill rating to succeed. ie,
  long distance swimming, endurance + swimming (required).

  For Tasks more applicable to a skill:  Roll a 1d20 success die, roll over the
  1d20 difficulty rating - the 1d6 skill rating to succeed.  is to hit at long
  range with a gauss rifle, difficulty 16 + gauss rifle skill.

In my opinion:  having 3d6 characteristics, 1d6 skill levels, 1d20 difficulty
rating, and 1d20 success die, work well together.

Comments? Alternatives? *please*?

Jim Baranski

James Baranski            | It's Not Easy Keeping Vaxen Working ...  Underwater!
Vax Systems Analyst       | [email removed] [phone removed], 889-4968
Lanman Hill, Norwich CT   | Be: Thoughfull, Honest, Yourself, Happy