No one said that classes didn''t exist. But when you are role-playing you assume to be one role, which - (typically) has a class. From here, we''ve never actually tried to do something other than control their actions by use of stats. I''m not saying that there are no classes - I don''t see that anywhere.

Players should be able to define their own class, not choose their own class. And in this I don''t mean you write in a character sheet (which I''m trying to abolish just about) that you are "A Master Fruit Picker" - I''m saying that by the method that he solves puzzles in game allow him to basically play out a class, rather than have some statistics that say he''s such and such a class.

And before anyone says that this means players can''t play particular roles because they won''t be able to act like a thief because they aren''t one in real life - I say that there would be no challenge in games then. The fate of all the players actions wouldn''t be based on anything but pure (somewhat weighted) luck. There needs to be a challenge in each game and the challenge of playing a role is much better than the challenge of finding out how you can possibly have the greatest chance of killing something/solving problem.

The character in game isn''t defining the role in the case I''m trying to present, the character is more like an aid to the player that tries to play a role that he/she can''t do in real life. We can''t limit the character''s ability to that which is in game, the character is just a medium, not the actual object (ie - the role - the player should be much more directly involved with the control over that role).

I''d love to keep discussing this, and hammer out the details of everything we''ve said using a "story award" system instead of the usual stats. My ICQ and email are in my profile I think, and the GameDev boards also remain a very good place to talk.

Now, in reference to some of the posts here:

quote: original post by Paul Cunningham
What you''re talking about is a reward system never the less, correct? But a reward system not based on player capital but story advancement.

A reward system, yes. Story advancement, hmm, sort of. I think "story adherence" would be more appropriate. I think it would all be much clearer if you guys had ever done an improvised roleplaying session with me and my friends here.
The "story" there, is usually a pretty vague thing. A few lines to know what the "main problem" is, and who or what causes it, perhaps with a few pointers on how to defeat it.
The rest of the game is directed by the DM/GM''s judgement, on what fits the advancement of the story at that time. It depends on the game mood that has developed, previous actions the characters have attempted, previous interactions with non-player characters, and the role they are playing. These all play a part in the story, perhaps that''s important for me to stress.
The "role" a player plays should not just influence the story, it should become a part of it. At the end of playing a full game, the player shouldn''t think of it as "I solved the problems using thievery skills", but rather "I just played a thief story".

How exactly should be do this? If I knew, I''d probably be making a ton of money already .
I think a lot of it depends on how you implement your "obstacle resolution" in your games. You could have adventure-style obstacle resolution: one, or possibly more, predefined exact ways of overcoming an obstacle. "Use key on door", or "kick door", triggers that specific event for that specific door. "Kick door" on another door might not do anything.

Then there''s old-CRPG style resolution: a door has a key, you can try to pick the lock, or try to force the door, all based on your abilities. Here, the system allows for a bit of flexibility, but completely disregards if it''s appropriate to the story for the lock to be picked. Now, using modifiers you could add it again, but then you''d need to explicitly add modifiers to everything, sometimes even time-based, and that seems like a hopeless overcomplication of things.

Unless ( and sorry if I''m thinking on my feet here, it''s early in the morning and I have a headache ) the possible actions to open/unlock the door are predefined ( for instance, use key, pick lock, force door ), and the chance of succeeding is based on things not that predefined: How many times has a character tried to pick a lock before? How many times did it succeed? Does what''s behind the door have an important effect on the story? Is it the right time to discover what''s behind the door?

I don''t know if that makes sense to you, but it does to me at this much-too-early-sunday hour .

The first two of those things are easily trackable, and have to do with the role the character has been playing so far. The last two are more difficult to evaluate. I think that''s where some really inventive design has to take place. Perhaps you could assign importances to certain locations at certain times, for events that you can only witness at those times. Importance to certain items, increasing towards the time that you will need them to possibly solve a problem. Importance to some NPCs, increasing when they can tell you something you need to know, or assist you. ( Get a locksmith to pick the door for you! ).


If it all sounds like mindless babble, keep in mind that I''m really just brainstorming, and I do not have a solution (yet). I am enjoying the way you are picking apart these ideas, because hammering out the weak points will only improve the idea over time.


Give me one more medicated peaceful moment.
~ (V)^|) |<é!t|-| ~
ERROR: Your beta-version of Life1.0 has expired. Please upgrade to the full version. All important social functions will be disabled from now on.

Having an open-ended story (i.e. non-linear) is very difficult to do, and is one of the advantages of having a human GM. The disadvantage of having these 3D graphics and AI is that it''s very hard to create a new place and environment on the go and also update the AI to respond to different plot development.

Unless you can perfectly represent the way an environment responds to any given action it gets a bit hard - and this basically means that the computer would have to be able to generate a compelling story - and then that means that we need to define what makes a good story and allow the computer to play around randomly with combinations of events etc - which still is a bum to do.

quote: Original post by MadKeithV
A reward system, yes. Story advancement, hmm, sort of. I think "story adherence" would be more appropriate. I think it would all be much clearer if you guys had ever done an improvised roleplaying session with me and my friends here.

So the game is just the blue print and the player/s fill in the gaps on the conditions of role playing. Kind of like having "roleplay hotspots" through out the game?

Or, kind of like: the game is more fun when they player role plays. The game solves this by matching players actions to the game blue print and rewards them appropriately?



I love Game Design and it loves me back.

Our Goal is "Fun"!

On classes: How do you define a classless game? It seems to me, by allowing a player to ''define'' a class, then the game is classless until that class is defined in some way. how do you plan to allow it to be defined? And how do you give the player bonuses to his class without knowing his class? This falls under the "in the first 30 minutes you use X skill Y times and get Z bonus for it and it forever remains a skill you get a bonus for because you used it in the first 30 minutes." I don''t follow how this would benefit the player, at all.

As for what Keith is talking about.. a DM outlines a mission. "You must rescue the princess from the orcs". naturally, this brings up questions: who''s the princess, where''s the orcs, why must we save her?
All of these questions need answers: She''s the daughter of the king of the realm, they''re somewhere in the wilderness to the west, and she''s rich and there''s a huge reward for her.. and she''s cute, too.
Well, now that you''ve got a basic plot.. now you build the story. This is how D&D is designed. You set up a plot, and you tell a story.. the characters interact in the story.. like the books in which you choose which way to go, and based on that you get other choices. The players are confronted with several modular choices. Then can go into town and ask questions, then can go to the forest and become hermits..
What they choose makes the story, defines it.. shapes it. Is it wrong to let them do an action they shouldn''t be doing? Sometimes

The key is to think "if i expose this to them this early in the game, they''ll die" or they''ll win, and the story is over. The game need not make things on the fly.. just have a LOT of modular pre-made things which come about only once the characters have reached a certain point in the plot. For instance.. they need to get inside this door, they can''t find a key. Well, perhaps the guy with the key is inside the door? Well, after say.. they complete X item, the guy with the key comes outside. now they can kill him and get the key. Or maybe they kill his wife, and he comes running out and forgets to lock the door.. then you don''t need the key!
It''s all about modularity and forced linearity. If they''re not ready for something, then most likely they won''t get to do it. It''s possible to do this in a game, but takes a lot of design, programming, and work. Keith, i''d love ta help ya out with this if we can get something going. As a DM, i''ve been waiting to see a game that can do almost as good as we can.. hehe!

J

quote: Original post by Maitrek

Having an open-ended story (i.e. non-linear) is very difficult to do, and is one of the advantages of having a human GM. The disadvantage of having these 3D graphics and AI is that it''s very hard to create a new place and environment on the go and also update the AI to respond to different plot development.

You''re right, it''s very difficult to do. That''s one of the things holding us back from trying. We all think it is impossible to get it right. I''m not pointing the finger at anyone but myself here, I''m not sure myself if it can be pulled off, but for some reason I just refuse to give up on the idea.

Some things I want to make clear:
- I''m not aiming for the open-endedness a real human GM can provide. I think that''s many, MANY years away in the future, before we can achieve anything like that.
- I am however aiming for supporting at least simulated open-endedness, or a lot of variety of supported playing styles, by tailoring the entire system to it.
"Some" actions will be possible, like the ones I defined earlier. Some things will not be possible, because it''s too hard to support them yet ("push piece of parchment under door, push key out of lock, pull parchment back, retrieve key from parchment" ). However, we CAN make the most of the interactivity that we do manage to put in.

quote: Original post by Maitrek
...and then that means that we need to define what makes a good story and allow the computer to play around randomly with combinations of events etc - which still is a bum to do.

Hmm, what if we can pull it off? What then? Even if you have a bunch of set clichés that you can run a story from - lets say, about 50 templates. Lets say those can be filled in quite randomly ( crappy example: evil *something* has overrun *somewhere*, causing *something bad* to *happen to someone you know*... etc. ).
These don''t even need to be completely done by the time the game starts. What if the "location" of the *evil something* is not determined until you have reached certain obstacles? That way, you can''t come across him/her/it accidentally in your earlier travels...



I know these are all very loose thoughts, still, but it seems acceptable to me. Something that stuck, is something I believe LandFish said once... even in a non-linear game, in retrospect after finishing, it should still seem linear, and not be obvious that you chose a "different" path.

With the things I have described, I believe it''s possible to generate a "non-linear" story in a linear fashion, simply because the details of an encounter/obstacle have not been worked out until you''ve actually met it, and attempted to conquer it.




Give me one more medicated peaceful moment.
~ (V)^|) |<é!t|-| ~
ERROR: Your beta-version of Life1.0 has expired. Please upgrade to the full version. All important social functions will be disabled from now on.

This is the kind of thing I''d really like to see done. In kind of early implementations it''d be an idea to have something of a database of plot points/features and then you could get bored enough as to write self-modifying code and get a quadrillion play-testers to keep playing and feeding back to the computer what it thought. It''d take a while to do but you could use the same AI to do that again and again?

And on classes a little : Well if the game starts out classless and it still works then that''s okay with me. I just don''t like the idea of the computer basically telling the player what he/she can and cannot do. It seems unjustified to take a player from one world where there are a billion restrictions as it is and then shove him/her into another world of imagination and do the same thing.

Well what about making the program analyse a graphic act. Then expanding the list of what the program can recognise. At the moment it would be easy for it to recognise acts like "fight" or "talk". But it becomes difficult when you create strings. Maybe there is another way you could do it without using strings?

I love Game Design and it loves me back.

Our Goal is "Fun"!

I think I''ll do a little text-version of what I''m thinking of.
It might take a while, but as a proof-of-concept it might work.

Some of the directions I''ve been thinking in:
- Plot-template database ( since that seems the easiest way to store information about what humans find interesting )
- Neural Network for tailoring the Computer-GM to the player ( play style, preferred "character class" if any, perhaps you could even have the player rate their games, so the NN knows whether it did a good job? )

Paul, you are talking about NPC conversation aren''t you? I am not sure how to go about that. Natural Language parsing perhaps, but that''s very computationally expensive, and probably very very VERY hard to implement in a way that doesn''t annoy players.

.. no, now I see what you mean. The entire interaction system. The "click your sentence" doesn''t really work ( i.e. LucasArts SCUMM system ) because it leads to "try every possible combination and quell all imagination".

I''ll have to think about it more.


Give me one more medicated peaceful moment.
~ (V)^|) |<é!t|-| ~
ERROR: Your beta-version of Life1.0 has expired. Please upgrade to the full version. All important social functions will be disabled from now on.

Just been thinking about this whole "stats" problem recently, and how to maintain classes as so many people seem to like, and I''ve got a few things to clear up.

The character - in game is - what I view as a medium. It provides an "interface" for the player to enter the imaginary world that we have designed for him. The player sees what the character sees and hears what the character hears. In this way the player gets feedback from the world through the character, but the character doesn''t control what the player wants to see. The player moves the character around like a camera, rather than sitting back and watching a film. This is interactivity as we know it - instead of watching a role we play a role and respond to what happens to the character by commanding this character that is more or less an extension of the player into another world. The same way a nuclear scientist puts on the radioactive gloves to enter the irradiated containment capsule thingoes to interact with the world that he cannot normally interact with (ie play with unstable nuclei).

I think that this kind of interface type relationship should exist further, so that the player puts his/her self into the role and anything that the player cannot do in the world should be substituted by the character, so that the character pads out the player so he/she can play the role of the character. I think that padding out strength/agility/mental ability is unnecessary for this. Sure, this can be done, but it''s not as if the player needs to see it for a start.

The player may want to be able to push this rock onto someone to solve a situation, and some people might say that a weakling may not be able to push the rock off if it''s too big, but this isn''t important. I think that the focus of what a player can and cannot do should be in the skills that the player has, and the skills that the character pads out for the player.

This is because we aren''t trying to replicate a physical world, there''s no brunt or brawn (of course, this may not be true in a medieval RPG and I''d be inclined to have hidden stats there). But in almost all other cases it shouldn''t be so focussed on the statistical physical(magical in some cases) ability of the character. It''s a game, and you can play a role without having a physical side taking over and restricting it.

The character isn''t the great focus, we aren''t trying to move the character and make him have a mental dilemma, we are trying to make the player have these kind of emotional changes in response to what happens to the imaginary world and the relations he has in the imaginary world. Which means focus on the story. The way the character plays should be dictated by the skill rather than the stats and make the computer react differently to different types of players. Don''t change the plot in any major way unless you can easily do so, start by changing the way the characters world and it''s people react to him, and thru the reactions the character gets, these are carried on to the player, which is the real part of the world, not the character.

Role playing does not mean you are restricted to a class or physical body, a role isn''t a "body" it''s something that changes the way a story flows in it''s own right.