Who the player is plus the game environment determines what should be. Awesome GM''s are masters of non-linearity. They know their players. They know what they like. If the game''s dragging, they know what to axe and what to toss in. On the fly, they can tailor the gaming experience to the player pefectly. We know computers can''t do this (yet, anyway). But one method that might get close-- and still be non-linear-- are dynamically changing encounter tables . I''m thinking about using something like this: Who The Player Is The player''s strengths and weaknesses (given by skills / equipment / allies / etc.)... Determine conflict types that will be difficult or easy. This gives a weighted encounter table. The Game Environment Humans seem to hate discontinuity and non sequitur. So environment re-weights possible encounters so that you only encounter what should be in the environment. This covers not just the area you''re in, but the situation / mission / events / story so far. This does open up a philosophical question, btw: How much should the player have to do to find action? Is it up to the player to find the conflict (greater weight on the environment) or up to the conflict to find the player (less weight on environment, but more out of place encounters)? What Should Be The notion of *logical spawning* would be the best way of making this work. You spawn in such a way as to make it look like the encounter occurred naturally. If the player can explore freely, then you will want to spawn in areas that he hasn''t been in, and areas where encounters could realistically happen. Example: You''re fighting your way through a closed space station, sealing bulkhead doors behind you all the way. If you reverse direction, you should not encounter anything unless it could have been hiding, stuck past you or have teleported in. But the airlocks-- access points to other parts of the game world, in this case-- would be perfect places for a spawn to happen. Spawn Rate Help Makes The Game Great Controlling the rate and type of spawn will allow you to control the challenge you give the player. Just like real GMs, you can lull them in to a sense of calm with easy encounters backed by a difficult one. You can drive them away from areas you''re not ready for them to enter with extremely difficult encounters. You can lead them as if by breadcrumbs with areas that have encounters happening vs. ones that don''t, or ones that offer greater over lesser rewards. The CPU GMing Since you know what you''ve thrown at the player, you can vary it to stave off boredom. And by monitoring success, failure, and the amount of time the player takes to solve an encounter, you might also determine their frustration level and ease back on the level of challenge (or spawn something that could help) Maybe applying this and a similar approach to AI could make the foundation of an awesome and very replayble game? -------------------- Just waiting for the mothership...

quote:Original post by Wavinator

The Game Environment
How much should the player have to do to find action? Is it up to the player to find the conflict (greater weight on the environment) or up to the conflict to find the player (less weight on environment, but more out of place encounters)?

If I had my way, it would be up to the player to find conflict most of the time. I am an advocate of player choice, and we all know w/ choice comes responsibility. In this case, it's responsibility to find the adventure.

Very great ideas on how to use AI for our own interests. We'd be working with the computer's short-comings instead of against them.


Need help? Well, go FAQ yourself.
What a plight we who try to make a story-based game have...writers of conventional media have words, we have but binary numbers


Edited by - Nazrix on February 17, 2001 11:43:05 AM

I love that idea. I think that random encounter tables are still with us primarily because of the limitations with computers in the past. It always takes someone to come along and make use of new capabilities to accomplish something truly innovative. Static encounter tables are just the easiest way to model reality in some situations. But a truly good game designer is not content to simply use the same old solution, simply because it is good enough.

While your idea is a great step in the right direction, I would like to make some suggestions. First, changing the encounters based on character strengths/weaknesses is a great idea. But it is easy to interrupt the "suspension of disbelief" if taken too far. It is important to ensure that even the dynamic tables conform to a strict sense of reality, whatever that reality is.

Also, I would suggest modeling the behavior of entire groups of creatures. In the original Dungeon Masters'' Guide, Gary Gygax gives a good example of how an orc tribe''s behavior should change in relation to attacks on their village.

For example, the first time the village is attacked, after a long period of peace, the residents will be wholly unprepared. Few, if any, of the inhabitants are likely to be standing guard. Reaction to the attack will be slow and disorganized. Panic is likely at first.

If the village is attacked the following week, response will be more swift, and resistance is likely to be better organized.

If the village is attacked repeatedly, the village will be increasingly better prepared. Guards will be posted and vigilant. A clear warning system will be devised. Traps are likely to be prepared, and a clear strategy will be followed.

If the attacks stop for a long period of time, vigilance will die down. Some traps will be removed, or fall into disrepair. Yet others, like pit traps, may remain for a long time.

Clearly, modeling such a complex set of reactions is more work, but your posting indicates that you are not the type of game designer to settle for what is easiest, yet inferior. And this system is not necessarily all that difficult. A simple "battle readiness" variable is a simple way to keep track of this. Another small array containing which particular traps or strategies are used is another detail.

Using your space adventure scenario, this could be applied to the behavior of individual ships and crews. Damage sustained, crew type lost (engineering, helm, etc.,) and similar details would have a tremendous impact on the strategy they employ. A heavily damaged ship will flee, if possible, for example.

I don''t know whether this has been helpful, or whether this is something that you have already considered. Certainly you know the value of those late-night gaming sessions with nothing but a few friends, some dice, and a whole lot of imagination flying around!

Good luck with your project. It already sounds like you''ve got the soul of a great designer. I hope you will let us know how it goes.

-Jonathon

Hmmm... This all sounds kind of familiar. A good idea, I might stress, and something, if implemented well, could go a long way.

Now, where did I hear about this idea...

quote:Original post by bishop_pass
Now, where did I hear about this idea...

Did you come up w/ a similar idea before? Sorry, I honestly don''t remember. Perhaps, you should use the "Wavinator Format" and accent things w/ bold and italics...It seems to get more attention




Need help? Well, go FAQ yourself.
What a plight we who try to make a story-based game have...writers of conventional media have words, we have but binary numbers

quote:Original post by Nazrix

Did you come up w/ a similar idea before? Sorry, I honestly don''t remember. Perhaps, you should use the "Wavinator Format" and accent things w/ bold and italics...It seems to get more attention

Yes, Nazrix, I will admit my posts seem to lack the ''style'' and ''presentation like'' feel of Wavinator''s. But we all know he is a hot rodding angel when it comes to stylish and effective discussion here.

Regaarding Dynamic plotting, yes, I did post a similar topic a while back. I even like to fancy that my contributions have had some small influence on Wavinator''s prolific design ideas.