[MUD-Dev] Curing skill spam (was: Moving away from the level based system)

[Chris Jones]

<EdNote: Table slightly reformatted for readability (I ended up
graphing the thing)>

On Wed, Dec 27, 2000 at 08:00:51PM -0500, Phil Hall wrote:

> How about something along the lines of having diminishing returns
> for repeated skill use as the player gets better at that skill. When
> the player is inexperienced in a skill, repeated use should give
> them rapid skill increases. But, as they approach mastery of a
> skill, repeated uses don't have any effect, only time does. IE, as
> long as the player keeps using the skill every day (or whatever time
> frame you determine is appropriate), they'll see a small increase
> every day. The amount of the increase is NOT determined on how many
> times the player used that skill.
>
> Basically, what I'm shooting for, is that you can gain skill points
> in an untrained skill easily just by practicing, but, once you start
> getting better, practicing becomes less valuable, and experience
> (ie, time you've spent knowing that skill and making use of it)
> becomes more important.  Character age will basically be what
> determines your skill level after a certain point. You can add in
> atrophying if a character doesnt utilize that skill over a long
> period of time.
>
> This allows young characters to quickly move up in skill gain early
> on through practice (which makes sense), but keeps macroing from
> being a viable option to gain better than average level in a skill
> and keeps the 2-day uber-characters at bay.

This system doesn't address every issue you raised: it would be
possible to get an uber character if you knew exactly which skill
tests to accomplish in which order. Macroing would still be possible
and probably encouraged. This system doesn't allow for artificially
slow increments at high skill levels (based on points allowed per
day), but does allow all players to quickly achieve some average level
of skill.

The insipiration behind this system was the problem of grouping in EQ.
It was frustrating to not be able to participate in activities with
groups of disparate level. My wife's 12th level necromancer couldn't
group with my 22nd level druid effectively because of levels and the
huge disparity in skills and power that EQ enforces. One of the
requirements for any solution was to allow a near-newbie avatar to
participate in activities with veteran avatars and be able to
contribute in some effective ways for the group.

To that end, I identified three different types of skills:

  1) Professional slow skills
  2) Professional fast skills
  3) Non-Professional skills

Your profession is dependant upon a quest, guild allegiance, class, or
some other system. That provides a set of skills available as
professional fast skills and professional slow skills. All other
skills are non-professional skills except for those that are exclusive
to a profession.

Professional slow skills are skills that are difficult to master and
advance relatively slowly. Imagine the use of a dangerous and exotic
weapon, piloting a dicey aircraft like a helicopter, or learning the
most esoteric and mind-bending magics. Professional slow skills
advance along a curve described by the equation:

  x = (2 - LOG(skill)) * 12.5 + (skill - 12.5) / 91

where x/100 is the percentage chance the skill will increase given a
successful skill application.

Professional fast skills are skills that are not as difficult to
master and advance relatively quickly. Stealthy movement, typical
combat manuevers, or cooking are skills that can be described by the
equation:

  x = (2 - LOG(skill)) * 25 + (skill - 25) / 83

where x/100 is the percentage change the skill will increase given a
successful skill application.

Finally, non-professional skills take longer to develop because they
are not being emphasized. You can certainly use them, but they are not
part of your professional training. In real life, a professional
software developer can learn how to perform surgery, but without the
practice and daily use of those skills, you can't expect him to reach
the same level of skill as a professional surgeon.  That's the same
way with these skills.  An artificial skill rating limit of 50 (out of
100) exists for non-professional skills in order to avoid the master
of all trades or "tank mage". The advancement curve for
non-professional skills can be described as:

  x = (1 - LOG(skill)) * 7 + skill / 50 + 5

For reference, some data points from the three different advancement
rates is provided:

  Skill    Pro Slow      Pro Fast       Non-Pro
   1      24.87362637    49.71084337    12.02
   5      16.18045736    32.28478604    7.20720997
  10      12.47252747    24.81927711    5.2
  20       8.819542637   17.41400914    3.29279003
  30       6.728292008   13.1322096     2.260151217
  40       5.276447911   10.12922311    1.585580061
  50       4.174962858    7.826954711   1.10720997
  60       3.295087392    5.967905487   -----------
  70       2.568142632    4.414717674
  80       1.953133404    3.085400928
  90       1.423616984    1.927069794
  99       1.005109518    1.0006864

So a real new avatar advances quickly, but finds a skill plateau
relatively quickly. Powergaming, "achieving," or exploration and
exploitation of skill checks will allow dedicated players to go past
the plateau, but for the casual or social player, an average skill
level is easy to achieve.

So unlike a system based on age or time in the game, this is based on
the number of successful skill tests. I would recommend that a
threshold of skill success be used to limit the total number of skill
improvement checks (say, if the skill check succeeds by 1 to 10
points, you can check for skill improvement, but if it was too easy
and the skill check succeeded by 11 or more points, the skill didn't
improve).

Caveat: this system hasn't been playtested, but I'm willing to present
this portion, warts and all, to get an idea of how effective this
would be. My implementation is waiting on me to finish designing and
coding a scripting engine.

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