I'll start off by saying this is NOT a full fleged idea, its merely a concept focus. Imagine an MMO with two warring factions genre does not matter. Each faction is controlled by a separate AI program script to generate each sides overall strategy. There are numerous territories in the game world. And the AI decides in a Risk type concept of which territory to attack, and which to bump up defense in. The players have nothing to do with the strategy part of the game (yes design decision). Once the decisions are made by the computer, it begins generating random-style missions directed at various things. Gathering supplies (gather quests), Intel runs into enemy territory (dungeon style crawls), Spec ops raids (kill quests etc). Over the week, the computer also pays attention to the other side, and decides where to re-allocate resources in an attempt to bolster defenses. When the battle goes down, (depending on game design) the computer resolves it, OR there is a massive PVP/PVE battle, where the status of supplies, available NPC bots etc vary depending on the the number of completed quests, varying based on type, also the number of failures affect the out come. After the battle, the two AI's stop, take a look at the situation and how its changed and generates its plan for the next week.... Over all the concept here is that the computer dynamically generates the missions based off of the game world, then turns around and dynamically affects the game world based off of mission results. What do you think.

I think that's pretty interesting. Instead of having a computer magically narrate a story for us, which most mission generating pipe dreams rely on, you've allowed the mechanics of generating quests to stay pretty much the same, and derive their merit from a dynamic world, rather than from incredible variation of minute detail. Good on you.

If player performance was tied into faction success such that you could "grind" your side's chance of victory, I think it would be a dynamite MMO dynamic. It lets everyone contribute without getting steamrolled in PvP, and puts real weight and value on missions that are abyssmally boring in conventional games. God knows how many invading fleets I've single-handedly repelled in EvE, but they just don't learn. If borders shifted and assets were gained and lost, it would be more than a mindless chore.

You could even set up staged PvP encounters like that. Have a faction offer a mission to a handful of players that requires them to be at a certain place as a certain time. When the time arrives, a door opens or something and the first five volunteers from each side get to go through and duke it out to decide the outcome of a battle or other scenario.

Man, this system has all kinds of potential. I'm not convinced that a computer could do it, but if the devs and GMs got involved, it could be really effective.

Quote:

think that's pretty interesting. Instead of having a computer magically narrate a story for us, which most mission generating pipe dreams rely on, you've allowed the mechanics of generating quests to stay pretty much the same, and derive their merit from a dynamic world, rather than from incredible variation of minute detail. Good on you.

My actual pipe dream is a system that has enough IMPORTANT variables that it takes you more than 25-30 missions to realize you're doing the same thing over and over again.... :) I just want the story to be a little more varied than the same 10 canned approaches.

However the biggest key in an MMO in my opinion is to have the PLAYERBASE to affect the world greatly, but to only let the PLAYER affect the world very limitedly.

I do however think that computers could handle this scenario. If they can create a computer to beat the greatest chess player on the planet, then this should be alot simpler.

I see over all about 3-4 Script sets running.

1. Overall Strategy Script. This is the script running the "attack territory" or the "feint attack" over all strategy.

2. Logistics Script. This one is a little more dynamic. This keeps track of the missions needed, the number of successes, the number of failures.

3. Mission Generator. Probably the most difficult. It takes the numbers from 2 and generates missions...this is the actual guts of what the player sees and thus needs the most work.

4. Results Script. Takes the numbers from 2 and generates the results for the battle this can either be what decides the NPC's present, or just calculates the actual outcome.

Those script sets are the ones above and beyond the standard MMO scripts needed. I don't think that it would be TOO complicated compared to the rest of the game.

Quote:

Original post by Anonymous Poster I think the concept is nice, but you should be really carefull to have a great variation in quests. Even if a quest is set in a different environment, players will notice that it's essentially the same quest they did a 100 times by now.

Artificial, meaningless variety is easily perceived by players; after breaking into the enemy HQ and reaching the commander's desk there is no difference between stealing a data disk with a superweapon design, embarassing pictures, battle plans or laundry lists: the player doesn't do anything really different.
Since tactics and player experience depend on the terrain, different situations are enough to create variety, even if the missions are very similar; as long as the mission is meaningful it will be welcome.
In a military setting, every mission consists of going to a dangerous place, doing something useful and returning to safety; there is nothing to vary in this basic structure (going to unimportant boring places, having vague objectives and lacking an exit strategy aren't good mission characteristics, both in games and in real life).

Interesting idea definitely makes grinding less of a chore and quests more interesting and varied.

But I would suggest some changes, instead of an AI setting overall strategy why not have a human controlling the overall campaign (or war or whatever). Like setting which province to attack, which province to scout out and so on. AI will generate the quests based on the strategy the commander has set out.

Add in an ability for players to elect the commander and we will see politics in game.

Quote:

I do however think that computers could handle this scenario. If they can create a computer to beat the greatest chess player on the planet, then this should be alot simpler.

There is a good reason why it would be a lot more complicated. It is because with a computer, and for chess, all you have to do is go through any number of previously played games of chess, and play the move with the most chances of winning the game in the end. In the case you are imagining, the reality will be greatly different. It would be more like having two StarCraft AI warring against each others. Not unfeasible, but not exactly highly imaginative or DESIRING the victory, and entirely dependant on PREVIOUSLY HARD CODED strategies. Since both AIs will have the same panel of strategies, at no moment should they be able to unbalance the whole game, even with players all on one side.

But if the strategist was the higher ranking player, for a limited amount of time, then the strategies could vary, and so could the gameplay.

Let's say this military game has permadeath. there is an almost infinite supply of men for that war (in terms of gameplay, it means that, although your soldier does die, you, as a player, never lack a moment of play, since you are reborn as a new soldier some seconds later). For each particular feat, you gain a rank in the army of your faction, and are awarded new bonuses for your squad or something, with any new rank. It means that each time you succeed in accomplishing a particularly difficult mission, or in getting all your men back safe, or any other secondary objectives, you are awarded with promotion points, which are in turn traded for more responsabilities and new objectives.

When you get to the rank of general, you get to play the map on a whole new level, the strategy level, where you create missions by designing your needs. If you need more wolf pelts to create more armor, then you create a mission to gather THAT MANY wolf pelts, and then transport said pelts to THAT point where they will be crafted, and then as many missions as needed to move said created armor to the points where they will be most needed, according to your own judgment.

AT any particular moment, only the higher ranking general can assign lower ranking generals to limited areas of responsability. Only the higher ranking general has all the responsabilities. Each time one of the missions you have issued succeeds, you get promotion points, and each time it fails, you LOOSE promotion points. WHen one of the generals is deranked, then the ArchGeneral also looses promotion points, meaning that even the top spot isn't safe.

How could anyone create new missions for the rest of his men remains a complete unknown.

Maybe it could be done through static structures being needed to gather some resources (like mines or wells, or fields, mills, rivers, anything), static points needed to limit the possibilities of opposing faction generals (forts or bridges to sever the supply lines), maybe it could be done as a result of spying missions, like a spy could find the missions of opposing generals, and report them, so that YOUR general could issue an order to destroy a convoy, or something.

But mostly, I think it should be done in a way that enforced PvP, instead of PvE. When in a war, you fight as much the environment as the ennemy, but not with the same intensity.

Now, on a side note, what if the "resurrections" and other fast healings needed supplies to happen? What if you could REALLY besiege a keep, and slowly kill one after the other the defenders, without any possibility for them to resurrect? Let's say for this to happen, what you mainly need is a counter, limiting the number of spawning in a spawn point, until something triggers the refill of spawn possibilities. So unless a smuggling mission succeeded, the defenders would slowly be killed, while the attacker could maintain his infinite supply of men. But still in the shape of supply lines.

Maybe, to materialize the supply lines, you could have the generals see colored and blurred lines, making them see when or how to cut someone off his bases, and how to ask for resources? Maybe beyond your own lines, there is a "fog of war" hiding the moves of the enemy, making your deicions more difficult, as a general?

I think this thread
http://www.gamedev.net/community/forums/topic.asp?topic_id=314511
is relevant to our discussion.

Maybe reading it could trigger some response...

Fournicolas

My reasoning for not having player generals is the following.

Quote:

However the biggest key in an MMO in my opinion is to have the PLAYERBASE to affect the world greatly, but to only let the PLAYER affect the world very limitedly.

Besides, the point is to have the missions reflect from the dynamic world.

I personally belive that players in MMO SHOULD NOT personally affect the world. They should be bit players.

But that is my Opinion.

Have you guys ever played Falcon 4.0?

The campaign system is pure genius. It generates missions in real time in coordination with ground and naval attack, while the enemy AI do the same thing. As a player, you choose the mission you want to do and execute it. Everything you do as an effect on the world. If you bust a tank, it could change the outcome of a battle for a bridge, which means another brigade might need to take another route, etc. Sure, there are only about 12 different kind of missions, but the fact that you directly affect the world you play in changes everything. Comming back after a mission and looking at how you changed the situation gives an incredible rewarding feeling. That system alone is the reason while I still play that game 8 years later.

A similar system in a persistent pvp world would be enough to make my nose bleed. (In a good way).

In WWII Online, there are missions to what you describe, generated by players who have advanced in rank and (by the server???). The missions are designed to focus the player resources at attacking and claiming points (towns, etc.)

I believe a system such as yours would work only on a large scale map such as used in WWII online where the ability to travel between points is limited due to time involved, otherwise you could jus have players spamming certain points and then spamming the next point such as in some online FPS games.

You could set up the hierarchy within the program to select targets using a risk and adjency profiles.

I also would think this would work in a online game better that would not have advancement levels.

I am not that familiar with RTS programming, but I would have to imagine this is how their NPCs determine their objectives.