[MUD-Dev] Looking for a Magic System (inspired by Bartle's book)

[Akiba Liebskind]

On 8/15/05, Eli Stevens (WG.c) <listsub at wickedgrey.com> wrote:
> erik wrote:

> I've thought about this some as well*, and came up with a few
> rules that define what I see as an "interesting" magic system:

>  - More outcomes/possible spells than rules.

>  - Wide range of effects.

>  - Allows for both creators and consumers to enjoy the system.

> As I see it, DnD magic is one rule = one spell.  All of the spells
> fit into a general pattern, but each spell/rule is pretty much
> one-off.  In general, each spell exists in a vacuum.  I want to
> use the combinatorial explosion.

I agree. As someone pointed out earlier, if it's possible to churn
out the best possible spells and post them on a website, what
incentive is there left to keep players using the spell crafting
system.

Ideally the system allows to generate very large numbers of spells
many of which are of comparable usefullness. On the other hand you
don't want all these spells to be interchangeable or feel "the
same".  I think a way to achieve this is by making a larger part of
the usefullness of your spells not in the spells themselves but in
the spell combinations.  Give players the ability to craft
incredible variety of spells but make them go into battle with only
a subset (the classic "you need to study the spells to have them in
memory ready for use but you can only have so many at a time in your
mind" or some such). An analogy would be Magic the Gathering deck
building.

In such a system there is a place for a wide diversity of spells.
People will want to experiment "concepts" that require spells
different from other concepts. There is room for several highly
effective concepts, several highly specialized concepts (e.g.: pure
defence) and tons of fun concepts. All these should help keep the
demand up for spells crafting above and beyond just trying to find a
very slightly more effective fireball. Make sure to include both
positive and negative effects for spells so that people will play
with degrees of tradeoff for minimaxing and so that many spells that
really aren't that usefull can be made (this can appeal to the
explorer and social people with contests for the most
useless/unusable/self-harmfull/funny/interesting spells including
spells that people end up using as emotes).

> What I've concluded (correctly or not) is that what I want for
> such a system is a programming language, once you strip off all
> the flavor. Look back at the three rules - they fit to a T.

>  - Define an "API" that provides/limits the spell's ability to
>  impact and react to the world.

>  - Give primitives that allow hooking the world/character input
>  APIs to the output APIs.

>  - Assign cost to the above.

>  - Coat with eye of newt/gestures/runes/words of power/pixie dust.

> Now, when it comes down to actually doing it...  Yeesh.  Ideas are
> cheap; the money's in the implementation.

I like the idea of clearly identified ways that a spell can exchange
information with the environnment but how does one control this?  If
the crafting is programming-like then it can be hard to
balance/tweak.  Either you give very little expressive power to the
system (how then to make it fun and varied?). Or you give much power
but then you can't easilly take it back from them if they create
something that breaks balance or harms the game somehow.

How about a more simulation/generation engine instead?  The user is
given a fair bit of freedom in the design of what he is crafting but
then the system computes what exactly the result will be. You can
tweak the computation in general and even tweak it to detect
specific kinds of cases in order to balance it properly. I'm
thinking something like "Spore" where you can make all sorts of
things with an idea of what the result might be but you actually
have to try it to find out how good it will be because the system
will compute/generate some behavior for it that you cannot 100%
control.

-Akiba
_______________________________________________
MUD-Dev mailing list
MUD-Dev at kanga.nu
https://kanga.nu/lists/listinfo/mud-dev