[MUD-Dev] I Want to Forge Swords. [Another letter to game

rayzam rayzam at home.com
Wed May 9 22:32:25 New Zealand Standard Time 2001

----- Original Message -----
From: "Adam Martin" <amsm2 at cam.ac.uk>


> I bring this up because of the number of posts recently that have
> alluded to wanting to be able to be a crafter as well as an
> adventurer, not in spite of it! Long discussions with fellow RP'ers
> some time ago led to a handful of conclusions:

>   1) Managing the skill increase and associated decrease behaviour
>   is a very hard balancing act - not least because by definition it
>   takes a long time to experiment with a new set of weightings! -
>   and you want to try to do this procedurally.

I've tried this, with hunting/assassination/stalking skills. Basically
you can build up a combat bonus against a species [hunting], race
[assassination], or in a terrain [stalking, in: jungle, versus cavern
versus beach]. On the one hand, the more powerful an opponent is
defeated, the more it should count towards the skill. Plus, the bonus
was impeded, so it became harder and harder to get up to Deadly.

The problem comes in the decay. When the decay appeared balanced for
the effects given and resulted in someone being an expert at a
limited, but not all, number of the species or races or in some of the
terrains, the user was limited in socializing. That is, the player
didn't have time to just chitchat or hang out because skills were
decaying. When the decay was minimized, the player would spend some
time just building them up, and then like inertia, go about their
everyday business and socialize, and as soon as it dropped a level in
effect, go build it up again. The real problem here is that it became
very easy to max out a large number of the subtypes, which negates the
idea of it being a specialization.

Instead of experimenting in-game from the start, I plotted out the
various functions, and varying the parameters, including various
'player-types'. The player-types were information on average time
on/adventuring/socializing/catching up on events & news
in-game/etc. So some spot checks along the power-gamer ->
community-minded continuum. [I hesitate to use achiever -> socializer,
mostly because that leaves out the other 2 :)].

It never reduced to a good solution, wherein it was still useful
enough for a casual killer to specialize in a few, while keeping a
power killer from keeping many, many choices at maximum effectiveness,
with little effort.

I opted for reducing the maximum effect slightly, but setting the
parameters towards the community/social/casual end of the scale. Thus
it wouldn't be unbalancing for the power-gamer.  In retrospect, the
one limitation to the system, I used was that decay only occurred when
a character was in-game. I didn't want players to feel like they had
to log that character off [though they could and play another], and
thus reduced the decay effect. If the decay occurred regardless of
being in-game, then the parameters could be shifted, due to the
general belief that everyone needs an average amount of sleep a week

Another alternative is to have the decay be tied to being in combat,
and thus any other activities in-game won't cause decay. This is
unrealistic, but the larger problem would seem to be that it dictates
what a player kills. That is, only kill what you're good at. If you
kill anything else, your hard-earned skills will decay. Seems like it
would end up with 1-trick ponies, not because they're masters at that
one subtype, but because killing anything else comes at a cost.

Now that I've stated the system, and how I analyzed the parameters,
does anyone have a better suggestion? I used exponentials/logs. Anyone
have experience with systems that decay when players aren't in game? 
Does that change the manner in which people play?


>   5) Any significant effort to improve a skill should result in SOME
>   permanent reward - you have "regression limits" where once you
>   achieve a certain level in a skill, it can't slip back below a
>   certain other level. Alternatively (or in parallel?) the skill
>   would regress as normal, but it is very easy to get close to the
>   highest level you've previously attained in any particular skill.

> Thoughts?

The problem I see with this is that a player would work hard on a
single one, to get it up to a high level, and then rest on the
'regression limit'.  Then choose another skill, and repeat. Thus, by
serial skill concentration, the player ends up with all of them at the
regression limit.

The alternative method, where it's very easy to get back to the
highest level previously attained, would again work the same
way. Concentrate on a single skill, get it up to maximum. Then
concentrate on a different one, until you've concentrated on all of
them, then it's easier to get multiple/many of them up to those
previous [maximum!] levels.



MUD-Dev mailing list
MUD-Dev at kanga.nu

More information about the MUD-Dev mailing list