[freeroleplay] Reflections on Dice Pool Engines [was: New Fringe Mechanic Take 2]

[Pitt Murmann]

On Tue, 2006-07-11 at 09:55:00 [-0700], Troy Truchon wrote:

> >It is introducing the problem we had with the Storyteller system.
> >In ST, you roll a number of dice (d10) and those that equal or
> >exceed the difficulty (generall 6-10) are counted as a success,
> >whilst those that roll a '1' are counted as negative.
> 
> Wouldn't it make higher attributes/skills a bad thing if the
> difficulty is 7 or above? I've never really used the storyteller
> system.

In fact, with the possibility being at hand to roll 1s, stochastics
dictate that you reach a break even point once you have accumulated a
certain amount of dice. What's more, variable difficulties (meaning a
value a die has to come up with or being thrown away) tends to make
players and game masters insecure about their actual chances because you
cannot instantly derive a percentage value without diving into maths.
However, the latter problem not only haunts White Wolf's system, but
many other dice pool driven engines as well, with Shadowrun maybe being
one of the first to appear and most prominent.

In their previous release (1st edition, 2nd edition, revised/3rd
edition), White Wolf tried to come up with a counter measure that took
into account particularly talented characters, resulting in exploding
dice, i. e. 10s being rerolled, if you applied a Specialty of an
ability, for attributes and skills alike. It hardly offset the botch
issue because it took four dots (dice) in an ability to qualify for
Expertise.

Starting with 2.0, they introduced a fixed difficulty (8 and higher
counts as success), with boni and mali growing and shrinking the dice
pool (adding to or subtracting from the number of dice) instead of
raising or lowering the difficulty. There are still special rules for
1s serving as botch candidates, but only in special circumstances.

Beyond mechanics, I noticed some sort of feel-good factor for players
being allowed to cast a hand full of dice over the table - the more the
better.

My personal feeling is that dice pools tend to model long term actions
in an interesting way. Where single dice, be it d10s, d20s, d100s, or
d{whatever}, yield quite binary a resolution, a dead-or-alive chance so
to speak, casting several dice seems to imply that the simulated action
takes several steps to go through, with good results making up for
deficiencies, and vice versa. BTW, I am not talking qualities here, with
3 successes being better than 2, or thresholds above (or below) target
values, but rather simulating subsequent partial actions.

An other interesting consideration is capping (difficulties) versus
accumulation, the former ignoring dice that fail to reach a certain
difficulty, fixed or variable, while the latter simply adds the results
of all dice. The first method has the advantage of fast resolution: drop
the shells and count the pearls. The second method requires primary
school maths for each roll, but also takes into account the tiniest
fractions of the ultimate success. West End Games' D6 Classic used
accumulation, while its successor, D6 Legend, went for capping instead.

Regards,

  - PM