[MUD-Dev] Level abstractions - Realism vs Game Issues
root at mpc.dyn.ml.org
Thu Jul 3 09:39:56 New Zealand Standard Time 1997
On Wed, 2 Jul 1997, Nathan Yospe wrote:
> On Tue, 1 Jul 1997, Matt Chatterley wrote:
> :On Tue, 1 Jul 1997, Nathan Yospe wrote:
> :> On Sun, 29 Jun 1997, Jon A. Lambert wrote:
> :> :I understand. You subdivide the global XP total into XP counters for each
> :> :skill and assign points when the skill is exercised.
> :> Hmmm. Crude, still. A start, though.
> :I can see where you get "crude" from - if the entire concept in mind here
> :is "skill based" with some form of realistic advancement, advancement
> :should be out of the players hands - an "XP" context is really not
> :required in the physical, useable sense, only as internal counters.
> Right. I don't like giving the mechanics to the players on a silver
> platter. If they want to figure it out, they have to play physicist or
Heh. I feel (vaguely) similar, and dislike giving players a look at the
works of the game - if nothing else, it really hacks into the atmosphere
you are (presumably) trying to create. I use something along the lines of
the standard LP approach to training (in a sense - many LPs let you train
by doing advance <amount of exp> <skill|stat> and such in class halls) -
players can choose what they are conciously training when they have the
opportunity, provided its the sort of thing they can train alone. They
cant dictate how much they train, or how much it'll ultimately cost (since
the cost is more in terms of gametime & effort).
> :> :I assign XP for failure also.
> :> I use a bell curve assignment. Failure teaches a highly skilled person,
> :> where sucess teaches a skilless person (actually, it is a LOT more
> :> complicated than that, as sucess factors into maintainance of a high
> :> skill, and failure increases short term chances on a poor skill
> :> [desparation points])
> :I can see where you get the overall idea from though - I like it! I learnt
> :far more about capacitors once I knew how they worked, and was
> :experimenting with alchoholic di-electrics, than when I did when I learnt
> :that E=1/2(QV^2), etc, and I learnt it all by making horrendous mistakes.
> This aspect, at least, I have covered. Knowledge is seperate from skill
> experience, and is specific to a skill (I have thought about
> interdependancies of theoretical knowledge, but have yet to implement it)
> and gives a skill specific bonus to learning and performance. Some skills
> (as it is a percentage bonus) are impossible without a decent theoretical
Integrating knowledge and skill together is probably an unenviable task..
but the above sounds like a pretty good initial approach to take. A good
example of something impossible without the theoretical background would
be an experiment of some form, without a basic understanding of the
quantities being tested - the opposite? Something like playing football.
Knowing the rules intimately helps, but is not absolutely essential for
you to have fun!
> :> :Right. I prefer to keep the abstraction and put it all of it in a global
> :> :pool for some of the game reasons I mentioned earlier. Your method
> :> :comes closer to "realism".
> :> And a bit closer to mine, but still a bit off...
> :I think the "skill web" ideas which were bouncing around some time ago
> :illustrate this best. If you have interlinked skills:
> :Combat, attack, knife, stealth.. etc, might all be linked together, so
> :getting better at combat will improve those below it in the web, and
> :getting better at knife will make it easier to improve in those above it.
> :Or something of the order.
> Roit, thanks. I proposed one of the skill webs, and started the
> discussions on usenet. Perhaps its time I seeded those grounds again? Or
> do we have all the candidates for intelligence (sans Reese and McClelan)
<g> A return to skill webs could be interesting, now that more people have
heard of (and hopefully understood) the concept (notably folks on here I
should imagine). As for usenets deadly-duo, who knows. :)
> :> :Nathan has some interesting ideas about how increases in certain skills
> :> :will pull along other skills. For instance, activity improving a
> :> :"writing" skill will affect the "reading" skill. Its sort of like
> :> :related skill categories/groups.
> :> :Hint: Why don't you repost some of your skill-net goodies, Nathan?
> :> Yeah, yeah (what is this, the fourth hint I've gotten?)
> :<g> I expect more than fourth.
> I've been busy.
I think we can establish we're all always busy. ;)
> :> OK, this is on the fly, as I don't have access to my earlier notes right
> :> now.
> :Well, it's probably more suitable to this exact case than something done
> :from notes. ;)
> Nah, those notes are pretty extensive. Gotta get back to that machine. Or
> remember to upload the note files or copy them to floppies
Hmm, the only thing this triggers in my mind is a memo never to eat donuts
near floppies - jam really wrecks them. Guess I'm getting hungry.
> :> Each player has an array of values coresponding to a set of existing
> :> skills and attributes. A given skill has its own list of dependancies,
> :> including physical and equiptmental dependancies...
> :I think I follow..
> Some things to think about: "fly" can trigger a half dozen skills, ranging
> from the raptorian soaring skill of the tre'laeci to the use of a rocket
> pack by a human or glah*drahk to the behavior of a swarm, not to mention
> the flight of a small bird (simplified, as only applied to NPCs) and the
> responsive versions tied to, for example, a jet or stingship (damn, I
> think the rocket pack is a response tied to equiptment. No matter.) As
> such, the parser attempts to choose the method most likely to suceed,
> unless specified. This requires a check to the current status of a skill,
> so if a character, say, had cybernetic wing grafts, as well as a jet,
> there would be a check to piloting, as well as one to use of those
So attempting to do something will select the most currently appropriate
method, and use the appropriate skills to that method to determine
success? I think thats the basic concept, and I like it.
> :> An example linkage in the global net:
> :Definitely the best way of tackling something of this order (and
> :magnitude) that I've seen yet.
> Thanks. Of course, this brings me back to the motive for hardcoding most
> of the game (allowing extension only within the context of the initial
> game world rules) ... this sort of thing takes up so much of my resources
> that I don't have much choice but to make optimizations where I can. Its
> still easier on the system than an out of the box Rom2.4 mud.
Heh. Well, it really is a quite unique (I believe) way to look at skills
(and the most realistic I've run accross). It's probably not hard to beat
the efficiency of something like ROM, made for mass consumption - you
understand your own code a lot better, for a start.
> :> :I do have 2 classes of skills when it comes time for a character to
> :> :do development. Skill categories and individual skills. Spending
> :> :points to improve "one-handed edged weapons", a category, will provide
> :> :a small improvement in all of these weapons. Spending points on a single
> :> :skill "short sword" will provide greater improvement, but only in the
> :> :specific case.
> :> As I don't have points to spend, that particular point is moot. As for the
> :> distribution... why not both? Practice a shortsword, and you improve the
> :> kinds of dexterity and strength attributes (I have them much further
> :> subdivided than in my above example... fast chopping is a special
> :> attribute in itself) that you need for any one handed edged weapon... but
> :> you won't be instantly AS practiced with a one handed axe.
> :Yeah. It's certain that in skill based systems of this order, some sort of
> :linkage between skills is really required, even if only in a basic sense
> :(ie the better you are at combat overall, the better you are at all
> :sub-skills of combat).
> Right. The difference between single and multiple inheritance, almost (and
> the third option, of a truncated, bi-layered neural net approach, more or
> less, that I use.)
I've seen single-inheritance skills on several muds, and
multiple-inheritance as well (but only in a limited sense, and usually
only two-layered). Seeing things go through many layers of inheritance
could be very interesting - you might get some quite obscure (but
> :> :They may be relevant if the assumption is they automatically increase
> :> :in a levels implementation. Hit pts. and mana pts. figure into the power
> :> :equation because they are often tied to it.
> :> I don't have character mana, and wouldn't even if I had mana. What I do
> :> have is the ability to carry bateries, power generators... this would
> :> carry well into magic.
> :Yeah, it would. I view magic as being a form of energy, in leakage
> :(something like electricity propagating by radiation), and those who use
> :it as a form of capacitor for magical energy. This is not an uncommon
> :view, I believe.
> Either that, or it is portrayed as something "borrowed" from elementals,
> deities, other realms, etc, or as a force tapped directly from the user's
> body (psionics).
Yeah. I think most of the time its just taken as "magic" - what I'd term
"abuse" of a tool, were it in a novel (I keep harping on about this).
Asking people to blindly accept something as "magic" won't work if you're
aiming at an intelligent, inquisitive (and downright nosey!) audience.
They want to know how and why something works, why it makes sense that it
does, and why they should accept it.
> :> Hit points are another story. I have always hated them. I allow direct
> :> damage of various kinds to be applied to physical objects. This does not
> :> mean that I keep a damage counter. I make fundamental changes. An arm,
> :> made of meat with a bone core, gets COOKED by heat damage, and is often
> :> unrepairable. It has mechanisms, on the other hand, to actually start
> :> repairing shear and compression damage immediately. However, a torn arm is
> :> not capable of outputting the same strength as a whole arm, which has
> :> serious consequences for use in that manner. (I have considered removing
> :> those physical attributes from the primary skill net altogether, leaving
> :> them in the age counter, and putting a set of them to each body part.)
> :Yipes, savage. Do you have some way of keeping track of the overall
> :situation? For instance, if you've almost lost an arm, blacking out, etc?
> Almost losing an arm takes care of itself. No way you can miss the damage
> if its giving off pain and refusing to respong when you try to raise that
> rifle, or worse, spasming and sending another jolt of fresh pain through
> you. Blacking out I handle as a single message, if its quick, or by
> something I borrowed from a Spider Robinson book, if its slow...
> Spoonerisms. I scramble text output... and verbal input, when it is
> reoutput as well. That, and a series of related symptoms of contusion.
> Something clobbers you in the back of the head. The floor ruches up at you
> and smacks you rudely in the face, and a parade of chirping stars rushes
> by, tweeting madly. The air swegins to bim, dna omesone estps on yrou...
> blackness claims you.
> :> : We are pretty close on this, I suspect. I have a species maximum for
> :> :base hit points. For most of the humanoid-size species its around 100.
> :> :The wild boar that was mentioned in another thread has a species max of 180
> :> :hps. Hit points are actually an abstraction of concussion damage and are
> :> :tied to creature size/fortitude. Thus while the boar is half a humanoid's
> :> :size, it's fortitude is incredible. How they relate to combat is probably
> :> :unrelated to this discussion. Let me just say that "wounding" figures more
> :> :into it than concussion points.
> :> Interesting. I do have a "three steps forward, two steps back" pain
> :> threshhold concept that may serve comparably... players, however, can
> :> fight through force of will to remain alive, to a point. After a certain
> :> point, if no one is coming to save you, you might end up better off
> :> recreating.
> :Heh. I have something along the lines of a similar concept, in that when
> :the "die" function is called in something, it may suffer a negative
> :effect, or actually die. This allows for all severe hits after a point
> :(potentially dangerous injuries which could cause fatality), and after
> :hitpoints begin to fall below a reasonable threshold.
> One thing I like about it is that you hve to maintain constant input just
> to stay alive.. if you don't, you "give up the ghost", and the lifeless
> slab of meat disassociates from its Character, which evaporates, ousting
> you (the PlayerIO) and forcing you back into shell mode.
Interesting.. No comments at the point though. :)
> :> : Power points or mana have their own skill categories and are quite
> :> :similarly done. These also have species maximums but these are tied
> :> :to other things, not size/fortitude.
> :> Hmmm. Interesting. A bit better than the old method, at least.
> :I don't think it's really hard to break the mould on this one. ;)
> Amen to that.
Furthermore, there are an increasing amount of people trying to break this
mould - one of the old gaming concepts that particularly deserves to be
smashed violently, IMHO.
> :> As I don't have magic, this really isn't relevant... but I _really_ like
> :> the idea of specific spells having recharge times, and concurent known
> :> spells being limited by some capacity value.
> :You and me both. Magic is one of the writers most powerful, and most often
> :abused tools in fantasy literature - it's also like a sword; double edged.
> I'm mostly a sci-fi writer, have only one (unpublished) fatasy to my
> name... and the only magic present in that was an ancient herbal medacine,
> and an animistic pseudospiritual lifeforce. I tend to function better in
> hard, hard sci-fi. For some reason, when I try to write fantasy, it has
> that sort of feel... and I hate reading that kind of stuff, 90% of the
Heh. I've not written anything (well, not finished writing anything) for
about six months now - putting together the legends for my mud history
will probably be the first thing, and I've started on that now.
"He can't stop us, we're on a mission from Glod!" - Soul Music (Pratchett)
More information about the MUD-Dev