With the recent release of games like Morrowind and Freelancer, I have noticed that random missions in games are severely lacking. Specifically, the random missions are pretty much all the same. This brings down the "epic" and "huge" labels on such games, as it seems like no effort was put into the game world beyond the main storyline. But how to generate content that is interesting? A friend of mine suggested the use of an rts AI as the AI for a faction or group, that has conflicts with other groups, controlled by AI. The AI fights other AI''s, and you are the bystander until it wants you for a mission, be that - kill A - steal resource depot A - Build house, whatever. Then the conflict adapts around that. Then you have a teired AI that oversees the entire conflict, and makes sure that either the player wins it, or the conflict rebalances, by "cheating" for the BG, i.e. New weapon is discovered on side of BG, massive arm is unveiled, etc. Anyone else have ideas? Gamedev for learning. libGDN for putting it all together. An opensource, cross platform, cross API game development library.

*bump*


Gamedev for learning.
libGDN for putting it all together.
An opensource, cross platform, cross API game development library.

I wholeheartedly agree.

I''ve been playing Freelancer (and am also a big Daggerfall/Morrowind fan) and I''m always unimpressed by the lack of diversity and variance in the mission structure.


I believe there are a lot of different that need to happen to make more ''interesting'' random missions.

For starters, and most importantly, you need a good diverse world/universe. In Freelancer there is the potential for so much, but they really limit it. You can''t fly anything but fighters and small freighters. You can''t blow up anything that you can land on. The economy is static and shallow.

You''re world (let''s use a fantasy-based RPG as an example.. just to be different ) needs to be able to support different things. Dragon-slaying and dungeon crawl missions are fine, but what about an assasination attempt? Maybe someone wants you to sink a boat full of supplies headed for a bandit camp. A wizard wants you to break into a castle and free his daughter who''s being forced to marry against her will. You can''t have these missions unless you''ve got important characters to assasinate, boats to sink, or castles to storm.


Secondly, you need to identify all these things and ''generalize'' them. Figure out everything that ''can'' happen in your world (interesting side note: We''re talking emergent behavior long before the game reaches the player) and build a good random generator that can intelligently create these missions.

Other points? Diversify your rewards. In Freelancer it''s always money. Never a new weapon or ship, never a cargo load full of anything. It''s all about money. Hopefully in your game the player will be motivated by other things. If you slay that dragon, chances are the people in the nearby town are going to be real nice to you. Free nights in the Inn, discounted prices at the store, etc... Even the chance to run more interesting and difficult missions is a valid reward. I start doing boring runs for the thief''s guild, but after I''ve run some of those they''re willing to trust me with more interesting and dangerous missions.

Just a few of my own ideas, from an armchair designer . Don''t worry, I''m all talk and no do.

--Ben

I agree. I think an rts AI behind the missions would be highly beneficial to a diverse variety of quests. However, it doesn''t address the more obscure types of missions, like some of the ones you mentioned. The RTS AI could effectively drive the action behind the scenes, but we still need to generalize the other mission types somehow. Any ideas?


Gamedev for learning.
libGDN for putting it all together.
An opensource, cross platform, cross API game development library.

I think these kinds of levels are a great asset to a game. A lot of people don''t quite understand that just poking around and discovery is an awful lot of fun because so many perceptual inputs are attached to that activity: the unkown, delight in discovery, the pure joy of wandering for it''s own sake (it''s a great way to get away from the pressures of life and the pressure of the game challenges), there''s mystery, tinges of fear, anticipation, excitement -- all these things go on and probably more just through this one activity of ''ok, I''m gonna poke around a little bit and see what I can discover''.

I believe it adds a great deal of individuated and personalized enjoyment in a game that another type of player might not experience or enjoy because they are more objective and beat the game oriented.

A post was put on here some weeks back about how a developer member of the community had just finished a ''walking camera in a world'' project, nothing more than the ability to walk around in the gameworld, no foos, no bosses, no resources, just the ability to aimlessly wander about and see the sky, feel the virtual grass under your feet and have virtually ''no particular place to go.''

He gave the cam executable to his dad, dad gave it to a buddy, they were older people by at least twenty years or more, and the devver thought nothing of it for a few weeks until he asked his dad, who confessed he had not run it (I can''t remember all the specifics except the important point this post revealed about player behavior), but he had given it to a neighbor to try, and he told his son he would find out and get back to him.

The neighbot had spent two or three weeks absolutely mesmerized by this simple program, and had been glued to the exe for hour after hour. Just simply walking around, no more complex interactivity than that.

Which brings me to my point, and I think it is an important one to all game development in the future.

The criteria are: the user was an older demographic; likely in his fifties plus or minus (in case you don''t realize this; this is exactly the key demographic of the huge mass market game we all have been hunting for the holy grail of for what, a couple of decades at least?) The level of interactivity was about a low res as you can ordinate (just taking a virtual stroll about a world with grassy slopes, plains as far as the eye could see, and a postcard sky), and they guy was as engaged as a tweaker on mod for days.

This type of interactivity, which I shall term not gameplay, or replay (the things we focus on as value enhancers) but as "extended play" is a great hidden jewel in our toolkit, and will be the kind of low res activity that is still rendering to screen that will be a major, major attractor for the other than male 14-38 demographics we safely sell to.

Building in these kinds of levels, with simple exploration and discovery as part of the design, with no major accomplishment for the level with the exception of maybe some low level resources collection and perhaps some unique antiquity or object found along the way, will be something that huge numbers of gamers that are not gamers today will be attracted to.

Why? Because they are not the kind of player that needs to have a level seventy five dragon around every corner (I don''t even think core gamers want that all the time), they have all kinds of stimulus all day long from good old fashioned regular life stuff, and when they want to get away, I think they want to get away with a little relaxation involved. Who has time for walks on the beach these days with our schedules in real life? This could be entertainment value that is a sleeper.

The real interesting thing here is that generating content that is interesting for this kind of extended play is not really that challenging from a interaction standpoint, only an imagery standpoint.

Benfinkel really said it, "Diversify your rewards." I suggest that in that diversification, you do what good storytellers do; you give the audience a chance to breath after the action, and a cutscene transition cinematic piece may be taking them forward perceptually a little too fast, when they may just want to sit down after battle and rest a little, or walk down by that little virtual stream and play splishy-splashy or look and ponder at their own virtual reflection. It was good enough for Plato.

Maybe I am not widely aware enough to know if this kind of thing is not already being done, but you can bet I am doing it for my adventure design.

Addy

adventuredesign, you and I have the same idea about large scale games. I don''t look at the game world as something you simply pass through on your way through the main storyline, but as something that the storyline evolves around. A place you lead a virtual life. That''s why I think it''s important to have a lifelike world. Generating content, such as random missions, in which you are not the primary player, content which evolves and is observable, even when you are not involved, is the most important aspect.


Gamedev for learning.
libGDN for putting it all together.
An opensource, cross platform, cross API game development library.

(Edit: I noticed that I hardly addressed the original topic, but what adventuredesign struck me, and I just took it from there. I'm going to get to the main topic in a second post.)

I have experienced this "passion for the gameworld" many times, in a multitude of games, and at different ages I was transfixed by even apparently unintended exploration elements.

For example, with most any scrolling world game where there where borders to the world, e.x. Mario, Boulderdash, Zelda, etc., I would be fascinated by the idea that there might be something beyond the border(an idea encouraged both by not letting the scrolling stop so it seemed like if you went far enough it would change, and the cases where you WERE allowed beyond the border, at least temporarily for things like the Warp Zone). This was at the age of 5 or 6 or so - as I got older the "border" effect had to be rendered in a more complex fashion or I would simply accept the world as it was.

One later game that struck me with this quite powerfully was LBA/Twinsen's Adventure, when I was about 9 or 10, and it was not deliberately designed into the game either. The border, in this case, was caused through my inability to figure out a puzzle(The Temple of Bu, I believe it was), and because I KNEW that the game could be finished, that it could be completed, I sat there for quite a while wondering, replaying the earlier parts of the game, thinking that there was a mystery there that I could solve. Eventually I gave up, came back to it years later, and found out that the solution was really pretty idiotic(I had assumed the whole time that I had to fight some skeletons that I could never beat to pass, but really could just march by them pushing a statue), but that's beside the point. The idea that there was more there probably was a factor in my getting the sequel too.

Going back to the greater discussion, the games that have inspired this in me most deeply in recent years are adventure games, mostly the Lucasarts ones, though also perhaps even more so the Jeff Tunnell/Dynamix adventures. The problem with them all , though, is of course that the places and the characters will stay the same every time, and so the feeling comes in only scraps and pieces, with the puzzles and story line coming secondary to chasing after the experience.

I envision 3d to be a great help in getting this to happen more fluidly in the future, fortunately.

[edited by - RTF on March 26, 2003 1:00:18 AM]

You need to play Uncharted Waters: New Horizon (RPG), it has a similar system like you mentioned, random/unlimited quests.

The quests are divided into three categories:
1. Merchant
2. Kill pirates
3. Treasure hunting

1. Merchant
If you go to a guild, and ask for a Merchant quest, the guildmaster will assign you a random quest. "Go to city X, meet the merchant there, he will give you a more detailed info." Once you go city X, and meet the merchant, he will give you a more detailed quest "Bring me 100 pieces of gold bars in 60 days, I will give you 10000 now, and 20000 later if you did the job." Where you got the gold bars is up to you.
(note that this merchant won''t say this if you are not in his quest, and the merchant may ask you ANY goods in ANY number, not just gold bars.)

2. Kill pirates
The guildmaster will assign you a task "Kill pirate A, he has been causing so many troubles to our merchants. 50000 reward." If you successfully killed pirate A, go back to the guildmaster, your quest rating will increase and you receive the reward.

3. Treasure hunting
The guildmaster will say "I am looking for a Super Crown artifact, but I have no clue where it is. But, you can visit the nearby tavern and ask people there." When you visit the tavern, and "Talk", some guy will tell you who knows where the artifact is. Once you meet with the guy who knows where it is, you will be given a map, now all you need to do is to travel there and find the artifact.
In certain cases, that artifact might be in a possession of some pirates or other commodores, so you need to defeat them first.

Note that all the quests do not "change" the world/story, and they are completely optional. If there are two sides fighting for dominance of a region (assuming it''s the story of the game), the quests are not going to interfere with the fight. The quests are somewhat "neutral" to the story of the game.

This, however, have one major issue. I played this game more than once, and I do these random quests only at the beginning of the game. As the story evolves, you are getting more rich, the need of random quests are no longer needed because you will be assigned quests that are related to the game stories, thus making random quests have less priority.

If UW:NW had no stories, no goals, perhaps, these random quests would contribute a lot in the game replay value.

Current project: 2D in Direct3D engine.
% completed: ~20%

In games, I think it''s the little details that make the difference between a good game and a great game. When you''re talking about an immersive world, it''s things like birds in the trees, conversations between world inhabitants, conflict that you''re not involved in, that evolves around you. A changing world. That is what will keep people coming back. So, implementation wise, I think that''s where AI such as rts AI comes in, because it can evolve the situation around both the character and independently. Anyone think that might be the way to do it?


Gamedev for learning.
libGDN for putting it all together.
An opensource, cross platform, cross API game development library.

The players actions are to some extent random as far as random meaning how predictable the results are going to be. If you could think of a way of using the players actions to create new missions this would hit many birds with one stone.