[MUD-Dev] ghost mode (was Re: SW:G)
J C Lawrence
claw at kanga.nu
Sat Dec 20 00:23:23 New Zealand Daylight Time 2003
On Wed, 24 Sep 2003 17:18:45 -0600
acius <acius at hedwig.simud.org> wrote:
> You can avoid this sharp threshold by having a smoothly sliding scale
> -- so there is no gateway of financial doom, but rather you lose a few
> more cents a month every time you gain another experience
> point. Either way, though, it's still a loyalty tax.
I've been musing on the definitions and implications of "levels", their
somewhat immutable nature (or do they really need to be immutable?), the
implications of level immutability on social formation and peer groups,
and the more interesting metrics which levels simplify and attempt to
I'm not interested in defining levels -- that's been done and well.
I'm also not arguing that levels are a Bad Thing, as some form of
tangible progress indication appears necessary to GoP games. What I
am interested in is playing with the concept and seeing if something
refreshing can be done.
Early conclusions for almost all level cases:
A) Level are usually implemented as a single "maxxed" scalar, where
"maxxed" indicates that it has the maximal value of a defined range,
can have no other value, and that it inherits the properties of all
lesser values within that range.
B) Levels are uneditable except for (typically) linear progression
along a well defined and constant treadmill/ladder (which in turn is
largely the purpose of levels).
C) Level values are tightly bound to characters, and are neither
separable or mobile.
D) Changes in level value very rarely require any cost mechanics,
in-game or out.
E) At a social formation and cultural definition level, "level" is a
rather complex term that has many implications, both useful and not,
in regard to group formation, peerage, status, and identity
F) In GoP games level is a fundamental metric in producing and
maintaining social hierarchies.
Note that I am explicitly not distinguishing between singular level
values and complex skill mesh/distributions. They're both levels, just
in different clothing and variant numbers of bells and whistles. In the
complex cases humans tend to reduce the complex values down to a
singular expression in a single linear scale.
None of these 6 conclusions above are surprising, but itemised, they are
interesting to consider and subvert. Quick musing extrapolated 5
principles I found interesting, not in themselves, but in how they could
be made to collide with each other:
1) levels need not be a simple maxxed scalar
2) levels need not be fixed
3) levels can be fungible, within a player or character, among
characters, players or other game objects, or even with RL and RL
4) there can be multiple forms of level transformations and
transactions, both within a character, a game system, a player,
multiple players, the intersection of the game world and RL, etc.
5) other surrounding metrics can partially supplant, support, and
amplify any new definition of "level". The assumption of singularity
is not necessary.
Simple points, and all previously discussed on the list. However all
the intersections haven't been so well covered. Each of the 6 in
principle is an assumption. They are an assumption of a commonly
understood model that is appealing as it readily communicates and won't
negatively surprise players. However, like most assumptions, they're
also deceptive as the real assumptions are more basic and less examined.
Unedited thoughts (these are rough):
1) Allow a character to arbitrarily reset his level and all related
statistics to ant value (eg slider controls).
a) For the literal, events and accomplishments of a character are
logged against the range of levels used during the accomplishment.
There are exploit vectors in here for GoP games related to variance
of level within an accomplishment, as well as detecting the real
beginning and end of an accomplishment action.
b) A secondary statistic, "renown", could be scaled or increased
proportionally to how much lower the character's level was than the
"par" for that accomplishment. Of course this then makes renown a
component statistic of level, and arguably not subject to such free
2) Each character has a set of "skill points". These are allocated in
various ways to form the skill mesh of that particular character. The
total number of skill points on a character is fixed and never changes
over the lifetime of the character. In the early stages of a game or
character some may be allocated to the category "nothing", in which
case they represent potential and unassigned talent.
a) A second character attribute, the equivalent of "level" marks the
character's ability to redistribute his skill points among
categories. A high level character can edit his own allocations and
readily make wide sweeping changes to his skill matrix.
i) Such changes could bear a cost, reducing the ability or
rapidity with which such future changes may be made.
ii) Such edits could have a latency which is proportional to the
magnitude of the change. Thus rapid re-definition of a character
as a functional method of problem accomplishment would be
iii) Having selected a new distribution, the rate at which the
character mutates into the new form can be a function of the level
of the character. Thus higher level characters could mutate more
or less quickly (which ever fits your model better). with the
mutation itself bearing either a constant or proportional cost
based on arbitrary factors.
3) A character may pick a level which is "higher" than his current
setting at any time. Upon picking a new level the character must then
pass one or more "tests" (gating factors). Should the tests be passed
the new level assignment becomes permanent.
a) The other related characteristics of the new level could be
instantiated before or after the tests are passed.
b) The character could:
i ) be restricted from world participation until the test(s) are
ii) have limited functionality in the general world until the
tests are passed.
iii) have the full functionality of the requested level, but have
to pay a fee (real $$$ or in-game value)
-- the fee could be waived if the new level is validated within
a defined period
-- the fee could be proportional to the savoire faire detected
in character behaviour while the condition was pending,
adjudicated mechanically or thru a function of the other game
-- have the cost of the new level (real world or game world) be
a function of how the tests were passed (speed, time, detected
effort, other character participation, etc).
4) Each character has a level range. As the character advances the
level range grows. The character may set his "level" to any value
within the range.
a) Changing the level value could bear costs.
b) "Range" could be given to other characters at a cost to the
donating player. Possible payment forms include real world (eg
$$$), game value (eg reduced level, reduced level range, lost items
etc), action requirements (eg must perform quest), or assumed
limitation (eg character assumes a liability or (temporarily or
permanently) loses an ability).
5) An edited character (level, skill allocation, whatever) could loose
relevance to the game world. This could be as simple as loss of
identity definition (name, species, guild affiliation, etc), location
(new character is located in arbitrary new/distant spot), etc.
6) Having requested a level change the adjustment could happen in
a) instantly (pending other criteria as above)
b) gradually (possibly with associated costs)
c) gradually, but with a starting position far lower than the
i) In the case of the re-allocation form, the total number of
"skill points" allocated to the player could be reduced (eg
minimum of each range value), and then gradually approach the
d) as a stair-step or laddered function of other in-game behaviour
(test, quest, whatever).
Skills can be zero sum games in and of themselves. A skill is a
specialisation. Each skill can be accompanied by a proportional
liability. You get better at XXX, but are more subject/worse-at YYY.
ie trade-offs. Jack-of-all-trades is nearly ideal, but sucks to
play as they're good at nothing. Experts are wonderful at their
specialty, but are risky to play due to their incapacities.
Levels can be made a negative sum game: each gain made is at the cost
of larger/more expensive losses/liabilities.
This would seem particularly interesting the more competent a given
character becomes in a particular skill.
J C Lawrence
---------(*) Satan, oscillate my metallic sonatas.
claw at kanga.nu He lived as a devil, eh?
http://www.kanga.nu/~claw/ Evil is a name of a foeman, as I live.
MUD-Dev mailing list
MUD-Dev at kanga.nu
More information about the MUD-Dev