We''ve heard this again and again: the player makes their own story. What does that mean? How do you as a player interpret & remember stories you''ve made in a game? It''s said that drama is life compressed until all the boring bits have been removed. In drama, nothing happens that''s not supposed to happen, and few things repeat. People are exactly and only where they''re supposed to be, and nothing happens without a good, narratively appropriate reason. Only the good stuff is captured and shown. "Thou shalt not bore" is a literary commandment. (And bore means "to make weary by being dull, tedious, and repetitive") If you''re going for non-linear stories in games where the player makes the story by their choices and actions, this is a real problem. Most games are very repetitive. Worse, I think that much of the actual experience of playing is nothing more than the sum of the individual moments: When you''re carefully creeping down a hallway in Voyager; when you move a unit to overtake an enemy capital in Civilization; when you skin a Gecko in Fallout 2... all of these moment to moment experiences make up the majority of the game-- and the majority of the fun!!!! . But while that''s the meat of playing, the story seems to be something else . Story seems to only be the big things; the important things; the things where something changes. So maybe then the best of these two worlds is your traditional, consistent (repetitive?) gameplay and choices where what you do causes big, important events? This isn''t quite right, so maybe somebody out there can encapsulate this better... -------------------- Just waiting for the mothership...

"the best of the two worlds" is relative though, I think... "Roleplaying" stuff has developped into quite a variety now. If someone manages to code a full blow 100% free world where YOU get to truely make the story (and hence the game), then that''d be cool. But on current specs, I doubt it. And the sheer size of code... But otherwise I actually enjoy linear games, with a lot of epic moments and dramas and twists and such... I''d prefer that to any pseudo free-style game, which tries to be free style but is never truely so. "Free style" games of today has the tendency to end up "empty", where you don''t really know what you are doing. If you don''t know what you are doing, you lose sight of the goal. If you lose the sight of the goal, you lose the entire game sensation with it. This happens because a game has yet to achieve the level of freedom real life provides. For me anyway.

The Present vs. Past
---------------------

In relation to other media.
One problem / opportunity that games possess is that they really take place in the present! Novels and film may give the illusion that they are happening in the present, but they are in fact predetermined. Everytime that you watch Driving miss daisy (or Aliens), the same things happen. Everytime you read a book the same events, etc occur. Whereas computer games offer the opportunity to play and have different things occur each time. One of the most important things in many films and novels is a chain of cause and effect.. with human factors inbetween.

Character
---------

Whilst games are well placed to create cause / effect ie. I shoot the alien, SO the alien gets upset. These are not USUALLY very SUBTLE, ie. they aren''t based on the level of UNDERSTANDABLE human relations, desires , or psychology.

Ie. A (greedy, dishonest) farmer who mixes his wheat with chaff. But denies it when confronted (cowardly).

In games we have not (often) got to the stage where the player tries to allocate psychological traits to the characters that they meet, and that these traits can be seen through GAMEPLAY - not scripting.

Question.
---------
How often in a game do you ask yourself WHY a character did what they did? Instead of saying that is just preprogrammed behaviour.
Is Artificial Intelligence necessary to make things a better story?

Obviously, there is the designer''s story. This is the predetermined plot found in most RPGs.But there are different levels of player story as well.

There is a low level story, which does not necessarily feature any climactic events. For example, how exactly did the player outmaneuver the enemy forces in the last round of an RTS? This is not part of an epic storyline. It''s just the player''s own, personal story.

On the other hand, there are high level player stories. These do involve major events which profoundly affect the plot. These high level stories include such things as choices the player makes which significantly alter the course of the game''s storyline.

These two types of player story are very different. Even a game like Tetris has a low level story. (The blocks were almost at the top, but I managed to keep cool. Just when I thought the game was lost, I got that straight piece I needed, and was able to battle my was back for another 12,000 points!)

So when we talk about incorporating player stories into games, I think it is important to consider which type we mean. Non-linear games have a long way to go before high level player stories are able to really work well.

But a designer can make good use of low level player stories without much trouble at all. Simply having the game remember certain player actions, and then showing the player that his actions are remembered, is an easy step in the right direction.

For instance, the player might overhear NPCs discussing the player''s latest exploits. Sure, this is simple for a designer to accomplish, but it could certainly be effective.

(In fact, I think Wav has been working on just such a system.)



Jonathon

quote: "Mathematics are one of the fundamentaries of educationalizing our youths." -George W. Bush

Never believe that realistic AI cannot be done in an open ended RPG with current spec,it''s a vicious lie. I am amazed at the ammount oof people that believe thaqt to make realistic AI, te NPC''s need to be constantly in action even when the player is not there, and that the AI has to be human-like to the very detail. You do not want to create realism - you want to simulate it. Make the player believe that everything is happening realistically, and then it matters not how you acheived it.

I am trying to implement an RPG with realistic AI. It is based on a scne/event generator. Nothing really goes on when the player is not there, but when a player gets to a location, the world-AI figures out what events would be likely judging on the positions of certain story NPC''s, and what events have already happened in the world. NPC''s start dumb as a brick, but if you speak to them, they will be promoted, and it is calculated as to what they would realistically know. NPC''s can then be demoted to thick gain if the player has no further use, but the world-AI might decide that an NPC warants to be promoted to story-NPC. Fromn then on, that NPC can be used for the basis of an event, and hopefully a bond will be created between them and the player (good or bad). Also, all through the game, it records the player''s play style, and tries to tailor NPC''s and future events to the player. All hopefully resulting in a story that the player has a lot of controll over, and a story that they will enjoy.

Yes, that is interesting, about that RPG you are doing. This idea of "simulation" of real world comes in different varieties for sure. And some like it one way while others like it another way, all depending on your expectations and level of desires. Jonathon''s Low/High level story concept is fascinating too... I can see what you mean about that Tetris "story" that individuals develop. Also Ketchaval mentioned the human psychological behaviour, which would be such an addition to "realistic" simulation.

When I jumble this all together, I feel that it''s the "infinite possibility" of reality that fascinates me. It''s actually the combo of those low and high level story concepts, where your character is living and experiencing his own stuff in a polymorphic (always-changing) world. "The world" here means the environment, which includes NPCs. And this certainly brings in that element of Ketchaval''s thing on human desire, psychology and so on. Sure Artificial Intelligence isn''t necessary to make a better story, but better AI certainly does help generate a more complex, realistic behaviour within the world. And you will interact with it to gain the Low Level Story, all the while being in High Level Story, as Jonathon might have put it.

I think my expectation is too high. But at least, for me, a "realistic emulation" is rather close to emulation of human mind. This is the problem for me, I think. When you say Hello to an NPC, the NPC may reply Hello. But that''s not because NPC "feels" my existance and the essence of my being. It''s just "scripted" to return H and E and L and L and O. And I think that is "Impossible to emulate with current home technology.

If one could, that would be SUCH a Story for me. Story of Life. Story of Virtual Being.

So the bottom line, I think, is What you mean by story, and What you expect the story to be. Personally, my idea is that Story is either My Life Story, or it''s something I read in a book. Alright, "non-linear" games are nice too, but if you think about it, "non-linear" games as we know it today are actually linear... it''s a combination of linear stories. The more combinations, the more "non-linear". There are other types, but they all end up feeling... incomplete. Something is always wrong. That is inevitable and understandable, given our current technology.

I guess it''s that high expectations that I seem to have unconciously, which makes me sceptical of modern "emulation" of the world... such as the game Sims.
I bought that game 2nd hand, and already sold it. It wasn''t free-style enough (Now I''m being greedy.) If I put a bath and toilet outside the house by front garden (without walls), I want to see the resident complain and not use them... I mean, would you? If God placed a toilet in the middle of road in front of your house, you wouldn''t use it. I''d pee down the kitchen sink. If god took away kitchen, I''ll pee in mugs and cups. If mugs and cups are taken, I''ll go to a neighbour for help That''s human psychology.
In the game Sims, they use it. And they start screaming and get out to get dressed ONLY when some other NPC comes close by. That is because each NPC is designed to react to each other... and they don''t "think" themselves, like we do.

On the other hand, mega epic story, which has twists and dramatic moments, are enjoyable as it is. Just so long as it''s well written. And that is easily achievable by today''s home computer standard; it''s like picking up a good book. And if a game is going to end up semi-realistic and leaves me feeling that something is wrong, I''d rather go for a linear RPG with a good story.

This post of mine doesn''t seem to have any point in it. I guess it doesn''t. It''s hard to explain how I feel about it, but I think you see what I mean. This "Making a Story" discussion could just as well be "Can PC emulate a Human Mind". Either way, it''s a very interesting topic to talk about.

quote: Original post by Wavinator

So maybe then the best of these two worlds is your traditional, consistent (repetitive?) gameplay and choices where what you do causes big, important events? This isn''t quite right, so maybe somebody out there can encapsulate this better...

I think that providing important goals (killing the general), then setting up systems which give second-by-second gameplay (infiltrating the prison camp, then the castle: whilst evading / killing guards, creating distractions etc.). But finally putting a recognisable human context to the minute-by-minute gameplay.. (the novice soldier that doesn''t want to kill - or die, misfires and has to use his bayonet).

Human Context
-------------

Combining gameplay with a good IMPRESSION of "psychological processes / personality" of the characters.


Being able to reason out, or hearing what is going on in these character''s minds will make the gameplay become more of a story.As it will put it into a human frame of reference:
Ie. You know why they did it: fear, greed, anger, boastfulness.

This would help to raise the "story" above the enemy shoots- but misses; I throw my grenade at him.. it was all over.

"Whereas if the soldier misses because his hands are shaking with fear.. and then you shoot him as he runs to get reinforcements" - Is this more of a story?

*because it happens due for a reason that is understandable rather than the manipulation of gameplay elements*


Ie. The "cowardly" henchman that won''t follow you into the shadows when you run there.. instead choosing to find a torch first or to try and set a trap for you. vs. the foolhardy caveman that charges headfirst after you with his club raised.

Hey, folks. Hey, Pug Penguin, welcome to Gamedev!

Anonymous Poster''s suggestion regarding simulating only when the player can see the results sounds like a great work saver.

Pug Penguin brings up an important point when he says that even the non-linear games currently available often feature very linear storylines. This is all too true.

One of the reasons for this, is that in order to create a more non-linear high level story, the developer has to do much more work. A story with only four possible diverging storylines requires four separate games. Sure, they may all begin in the same way. But if they truly diverge, they will each end differently.

Another problem is that it can be difficult for a designer to justify that extra work. The company or individual funding the project is likely to look positively stupefied when he learns that the designer is asking to fund large portions of the game which the player may never even see. "What do you mean, the player won''t see this part until the third time he plays the game? Are you insane? If it''s not necessary, let''s just scrap it!"

And one of the other major problems with a diverging high level storyline is that its possible storylines are not infinite. The designer may be able to provide a number of main branches in the story. But in the end, the only high level stories possible are the ones the designer has created.

That is why low level stories are so much easier to work with. By allowing the player enough possibilities, the game''s stories can travel in directions the designer never anticipated.

I remember a quote from the designer of a strategy game. (I''m sorry, but I don''t remember it exactly.) In essence, what he said was that when the game was completed, he didn''t know how to win his own game. He designed each of the strategic systems himself. But when they were all put together, the possibilities were so limitless that even he wasn''t sure what the best strategy would be.


Jonathon

quote: "Mathematics are one of the fundamentaries of educationalizing our youths." -George W. Bush

quote: Original post by Jonathon
One of the reasons for this, is that in order to create a more non-linear high level story, the developer has to do much more work.

Yes, I agree completely.

quote:
Another problem is that it can be difficult for a designer to justify that extra work. The company or individual funding the project is likely to look positively stupefied when he learns that the designer is asking to fund large portions of the game which the player may never even see. "What do you mean, the player won''t see this part until the third time he plays the game? Are you insane? If it''s not necessary, let''s just scrap it!"

Good point, but that''s why we don''t ask for extra funding and just do things independantly (and suffer the consequences of less than the flashiest graphics usually)

quote:
That is why low level stories are so much easier to work with. By allowing the player enough possibilities, the game''s stories can travel in directions the designer never anticipated.

I remember a quote from the designer of a strategy game. (I''m sorry, but I don''t remember it exactly.) In essence, what he said was that when the game was completed, he didn''t know how to win his own game. He designed each of the strategic systems himself. But when they were all put together, the possibilities were so limitless that even he wasn''t sure what the best strategy would be.


Jonathon

"Mathematics are one of the fundamentaries of educationalizing our youths." -George W. Bush

Yes that''s exactly the philosophy that Wav & I have been discussing about the strategy plot concept. If the effects of a typical strategy game in all of it''s limitless possibilities can be used with plot then we could have very interesting situations with little or no intervention from the developers after the system is developed. Very good job in explaining the usefullness of low-level components.




A CRPG in development…

Need help? Well, go FAQ yourself.

(This is long... but this isn''t ranting, it''s serious )

Yes, that "fuzzy" nature of the strategy plot concept certainly is a lot more "free form". At the moment, though, such idea only seems to extend in certain genres. While all those concepts are nice, we also have to look at it practically. And for that, we have to figure out what makes these games so... unpredictable even to the programmer himself.

Strategy game is a good example. They come in varieties, but perhaps we should use a simple example for the sake of argument.

In strategy games, I believe the key is its iteration of units and attack modes, as well as possible movement of each unit, which makes strategy games more of a free-style, don''t you think? If there are 2 units and 2 fire modes, you can have 2x2=4 different combats. Put them on a chess board, now you have so many ways to get from A to B. Even with 2 pieces on it. Now expand the map to something like 1000x1000 cells... and put on 20 different units, each posessing 2 firing modes. The possible course of the battle expands rapidly.

You may implement the unit''s AI to cope with small 5x5 map. You may also implement AI to handle its only two different firing modes (say, close range fire and long range fire). But that AI will surely still work well enough on 10x10. And with 100x100. And 1000x1000. Then add another unit on the board. Then 2 more. Then 10 more, then 100 more. Alright, I am simplifying this too much. But you get the idea. The possible course of Historyline on the battle field grows rapidly by modifying few variables, though the AI remains more or less the same.

How do we get a similar effect in other game genres, say, in first-person shooter like Quake and rpgs like Final Fantasy?

Quake type game might achieve this simply by programming a very clever AI... perhaps. Whether it can be optimized enough for real time Fragging is another story, but if one bot AI is programmed vigorously enough so that it learns the surroundings and take advantage of hight, visibility, clear line of sight, weapon ammo location etc, you might be able to put multiple bots (like 100 bots...) in a huge random map to have a totally unpredictable battle, much like the uncertainty of real life army men. This is again simplifying things too much, but the concept is there.

They''re a few forms of an emulation of realism through finite amount of programming (as opposed to infinite amount of programming).

Now, how do we make RPGs and Adventure games so they inherit similar principles?

Relying on conversation with NPCs is probably a bad idea until AI can very closely emulate human mind (and our use of language through emotions). Otherwise programmers would have an infinite amount of programming to do; that is, write many many many many branches in the story and put many many linear stories together to form a non-linear story. And as Jonathon says, High level story would ultimately always rely on what the programmer hard coded during development. Given that the emulation of human mind in NPCs seem to be impossible right now (else Robocop and Terminators should exist by now), we seem to have run into a brick wall.

But, perhaps, this could be remedied by relying on other features of RPGs. Even with today''s technology. Perhaps. Let me elaborate more.

The implementation of the "free-style" concept in the earlier two examples (strategy and first-person shooter) relied on the iteration of a few simple patterns, which allows the possible course of History to grow rapidly, to an extent that even the Lead Programmer has no idea of the best way to play it. So... somehow, we have to bring that concept into RPG/Adventure genres.

Say... Monster iteration, where Monsters mate to produce offsprings over a period of time. Of course, monsters must get stronger in some ways as they evolve - perhaps by taking the mean average of its mother and father plus some addition. Monsters might also fight against each other so always (or almost always) the stronger survives to achieve Darwin''s theory of Evolution. Either way, the AI has to work systematically, with some randomness thrown in much like strategy units. And you, as the main hero, must fight to save the world from these mobs. Objective must be pre-set, much like strategy and 1st person shooter counterparts - that is, "Destroy the Enemy". You can play in whatever way you want, so long as you "Destroy the Enemy" in the end. This gives players a sense of goal, though very simple it may be. That''s a good thing.

So how would the game play... I don''t know. I have no idea. Obviously, we will have to implement one single monster AI vigorously. It''s no good if it''s randomly picking its mate to reproduce - there has to be some logic behind it as well as randomness.

This is much like in strategy AI; a unit may show some random nature, but it also considers the surrounding areas for signs of water, forest, mountains etc. Then moves appropriately, depending on its needs - forest for protection, low land for fast movement, mountain for visibility, and so on. A solid strategists'' idea. Again, this is a systematic AI with randomness thrown in.

So, say, there''s always one Boss monster. A chief monster, villain, whatever. And if you kill him, all others die and game ends. Or you can try to kill all the lesser monsters to reduce the reproduction. Perhaps you can discourage monster evolution that way and use it to your advantage. But within the monster community, fight goes on to reproduce and expand. Perhaps that "boss" is simply chosen by the field engine, or perhaps the monster with highest weighted-average of his status becomes the boss. At the beginning of a game, all monsters are obviously weaker and more accessible, so some players may try to rush and take him out early to end the game... (how boring). But if players bug around too much, surrounding monsters will become stronger through mating evolution and with it the chosen "boss" monster will be stronger. Obviously players would be higher level too, having spend so much time killing stuff. A good balance is needed there, just like any other game.

Anyhow, now the iteration principle exists in there. Throw in few more variables - say, partially randomly generated map. Or throw in a volcano or two... and some mobs might extinct if they erupt closeby. Or perhaps the opposite, they might strengthen their resistance. The RPG world may be big. Or small. Initial number of monsters maybe just 10, or 1000. Either way, each variable will contribute to the iteration to produce a vary "free-style" adventure.

Something is wrong... don''t you think? I think the problem there is that it totally lacks the classic High Level story, from which RPGs earned its name over the last decade. It may be full of Low Level Story, but not High Level. RPG is "supposed" to have conversations with NPCs. You converse with NPCs to construct the story, at the same time as the story being the backbone of NPC conversation. It''s perhaps a two way exchange - Converse, and have High Level Story. Or No Conversation, No High Level Story. No Pain, No Gain ... sort of.

But I guess it really is No Pain No Gain. Right now, because of how people expect RPGs to be, we can''t really have it "free style", to an extent that even the programmer doesn''t know what the best way to play it is. If we were to introduce heavy free style nature, we have to lose something else instead.

Then, to truely implement this "infiniteness" in RPGs ("we could have very interesting situations with little or no intervention from the developers after the system is developed." as Nazrix put it) we still need to first figure a way to emulate human mind using AI.... . . . .

Perhaps, in a sense, RPG/Adventure genre is truely the most complex, superior genre of all games. Perhaps. It was born so simple. And now its simplicity fires back at it. It''s forced to become complex.

Scary, no?