[Design] Why random rolls?

[Lev Lafayette]

Hey people,

Good set of comments.

Dave, Karl and Julian all really picked up on my
leading question on the range of randomness. I find it
odd (from a simulation or narrative point of view)
that random die rolling is invariably the same
regardless of the skill or attribute being tested.
This seems to be seriously strange, imo. Julian and
Dave were both also quite correct in suggesting that
the purpose (at least at this stage) is simulationist,
although the gamist element is obviously important as
well - a game that is *too* predicitable is dull as
dishwater, and a game that is *too* random simply
requires a suspension of disbelief that is not
enjoyable.

Very interested in Peter and Carl's comments on how
they don't fudge die rolls and use the results as an
opportunity for the GM and players to come up with
plot explanations. Also interested in Peter's comments
that unimportant NPCs are dealt with minimal die
rolling ; this was actually recommended by Aaron
Allston a long time ago in a Justice Inc supplement.

Carl's point on the range of skill rolls compared to
expert systems who act beyond the range of their
programmers is interesting because it shows (once
again) the necessity of clearly distinguishing between
skill and effect. Ditto for the comments concerning
'vague clauses about not needing roll for routine
tasks'.

I may also make mention of Karl's d4-d4 system where
the random effects to the base roll can be positive or
negative. That's really a nice touch.


OK; thoughts at this stage resulting from the
discussion.

Randomness is important for simulationist (the fiddly
bits not worth calculating), gamist (odds and risk
taking) and narrativist (dramatic moments in a plot)
reasons. However this also indicates that the *degree*
of randomness in determining a particular event should
likeness be *also* variable. Some skills (and
situations) are more random than others and should be
treated as such by the simulationist aspect of the
game system. Some times it doesn't serve a gamist
orientation to engage in odds calculation; you want to
know that your abilities provide stability and
exactness. Sometimes plot incidents are too
insignificant to deal with a detailed exposition of
random influences. 

As far as I know, most games simply *don't* provide a
system for this purpose. They may (as the
aforementioned Aaron Allston example, or GURPS "Quick
Contest of Skills") suggest it as an alternative
resolution mode but not in a systematic manner. 

Something that I've discovered in the Outbreak of
Heresy playtest (and contrary to my expectations) was
that players actually seem to enjoy (on average) two
die rolls for skill and effect resolution. They like
the "second chance" aspect of it, I suspect (OK, I
stuffed my Skill Roll, but I may get lucky with the
Effect Roll, despite the modifiers).

Anyway here's an initial proposition within the game
design.

As previously suggested, the capacity to perform a
skill depends on one's skill. The relative success of
a skill roll (critical success, special success,
normal success, normal failure, special failure,
critical failure) modifies an effect roll which is
derived from the trait (aka attribute, characteristic,
stat etc) being used.

However the random influence on this effect should
varies according to how random the activity is or how
important it is to the narrative.

For simple attribute tests where there is no pressure
or skill being used above what one would learn as a
child, simply apply the trait average result as the
effect roll. Walking to the corner store is an obvious
example. Likewise apply this to utterly insignificant
plot elements ("we travelled between London and Devon
without incident").

For attribute tests which have an element of skill,
but still tend to be "trait driven" use the Low Random
category. Example of this could be use of a running,
jumping or climbing skill. Likewise use it for skills
which are heavily skill driven, such as reading a
book. In a narrative sense, one can also use these as
"quick contest of skills" with fairly trivial plot
incidents.

It seems almost too convinient that conflicts between
individuals have both narrative importance (in
general) and are affected by extraneous events for
simulation purposes.

Anyway, will be working on putting this down in a more
systematic sense during the rest of the day.

All the best,



Lev


Lev Lafayette
lev_lafayette at yahoo.com.au
http://au.geocities.com/lev_lafayette


	
		
__________________________________ 
Yahoo! Mail - PC Magazine Editors' Choice 2005 
http://mail.yahoo.com