Emergent Gameplay (In General, and as Compared to Linearity/Scripting)
 

After reading (and participating in) IchI's Violence/Sex/Drugs thread, I thought I'd add another discussion to the mix.

One of the more interesting and promising areas that games are beginning to explore is that of emergent gameplay. For those unfamiliar with the term, emergent gameplay is what happens when "games should be less predetermined and merely feature a set of rules governing its elements. When you throw of a bunch of agents in the mix, whether they're AI or humans, different things occur based on the rules and what sort of interaction takes place." (Steve Bauman, Computer Games Magazine)

Two of the most emergent games (as in featuring the most emergent gameplay) are Grand Theft Auto (3 and VC) and Morrowind, both of which were widely lauded as the best their genres had to offer. Is emergent gameplay the way of the future, or is there still a place for scripted/linear games?

I think the answer to this question is multifold. There are many elements of such emergent games that detract from their own advances, and many technological limitations that will hold them back for quite some time, I suspect. To create a true emergent game, the developers are there merely to create the world, something that requires things we don't yet have (I'm writing about that right now, actually, and I hope to have a piece for you in a few days). However, the potential to create immersion and feeling with scripting remain powerful, because the tight control the developer has means that the player's emotions can be tightly handled.

There are two general categories of games right now (in terms of emergent gameplay): those that offer wide-open worlds, stories and missions (e.g. Vice City or Morrowind), and those that offer tight, linear gameplay with scripted thrills (e.g. Call of Duty). Each has their benefits, but I believe that one has a future (and, to a degree, no present), while the other has no future, but a lush present.

Open-ended games such as VC offer one type of thrill (the thrills of a game being the things that draw players in). This thrill and excitement stems from being part of a living world that can be explored and tested, experienced or ignored. In Vice City, I've spent countless hours testing the limits of vehicles, flying over the city, or wheelie-ing down a highway, all for the mere purpose of "the hell of it." This is emergent gameplay in action, and I'm liking it. On the other hand . . .

Call of Duty is an unbelievably intense game: it features excellent (or at least cool-looking) AI, beautiful, detailed levels, excellent acting, excellent sound design, and beautifully, cinematically scripted levels. The thrills here come from being the center of an exciting narrative experience, one which is tightly controlled so as to render it the most intense, most immersive experience the developers can create.

Open-ended games are only just beginning to come into play, with only a few on the market right now, whereas linear/scripted games are peaking. The era of the scripted game is beginning to close, in my opinion, as the appeals and potential of emergent games like Vice City begin to take hold. There will always be a place for the Call of Duty's and Half Life's of the gaming world, but the bar to be lauded as a scripted game in a world with emergent games is (or should/will) be set very high, and justly so.

While both of these mediums have their merits, I think the next true level of games requires a blending of the two (and the full maturation of emergent games). To rise to the next level, the games industry must create an open world like VC's or Morrowind's, but replace the braindead inhabitants with entities that think and understand. In short, they must infuse Vice City's world with Call of Duty's immersiveness--the player should be one piece of a million in the world, but still have such things happen (randomly, not through scripting) that they feel like the star of a movie. Every player should be able to have their own experience, and have it feel just as intense and immersive as Call of Duty, but just as off-the-cuff as Vice City.

Anyway, those are some of my thoughts on emergent gameplay and such things. What are yours?

I finished it early: here's the piece I was working on about the ASI (read on). It's not terribly well-written, because I tend to ramble when writing about games, but it conveys the point:

Think of the last few games you played. Try to sum them up in a single verb conjunction—“to win.” How many of them involved some sort of violence? How many involved superiority or dominance? Most games, especially the mainstream ones not targeted at a specific niche, revolve around violence in one form or another. Whether it be Nazis in Call of Duty or trolls in WarCraft III, violence is the lens through which most games portray their world. Violence is the easiest thing to portray in many media. In games, this comes from the (comparative) lack of complex programming and art involved—it’s easier to depict a mangled corpse than a lively person. Because of this, and because of technological limitations that we are only now beginning to overcome, violence has always been central to games.
One of the elements that developers of violent games have been touting recently is their new flavor of Artificial Intelligence, AI. AI has been developing rapidly, leading to more and more realistic depictions of humans in combat. However, Artificial Intelligence is a misnomer for such entities, because the only thing that they are capable of doing (without scripting) is fighting. Developers may have developed the most human-like fighting reactions and responses, but they have not created an Artificial Intelligence, they have created an Adversarial Intelligence. To take games to the next level developers need to craft a true Artificial Intelligence, an artificial social intelligence.
This ASI would allow a true virtual world to be created, with characters, events, and situations that had much deeper moral context. Games as they stand today lack the depth and social awareness to rise to the equivalent level of, say, the movie industry’s A Beautiful Mind, but an ASI would allow the creation of characters and stories with resonance and meaning, and allow games to be created that provoked gamers into detailed thought, and inspired goodwill instead of grand theft auto (although I do not believe that games can provoke violence in all but the most disturbed persons).
Just as some movies and books are considered to “matter,” games can achieve that level of maturity and value with an ASI. However, without a socially intelligent entity that can remember and react to the actions of a player, violence is the only medium through which we can see gameworlds (with a few exceptions).

Hmmmm... very interesting... You was talking about Violance and Ai and that mainly games designers have only managed to construct 1 main Ai that works properly and thats fighting... I dunno if u noticed (obviasly u have) that the human mind has different emotions. When in battle u can feel a lot of emtoion but they are only about 4 major ones. Anger, Fear, Love and proberly something like Survival (if that can be classed as an emotion, i supose it could but its more of an instinct) Because the human is so complex but in my opinion has the most basic of emotions its easy to prodict how somone will react when shot at. Humans are very complex and when u talk about AI and artifial inteligence and been truly imersed in a game where nothing is scripted. If you think about it. Wouldn't u feel for the people ua re with? Games like gta3 could get thrown out of the window in terms of what ua re actually doing. If you are playing a game and inside this world the AI has generated characters, these people talk to u and have emotions. Even though theya re information how could u kill them? surly it would be like comiting murder. No matter how much u try and say ye but its only a computer. How do u know what reality is? In my opinion there is no reality. How do u know every day is not a dream? When u go to sleep u forget about what has happened so far and start a new dream? U wouldn't know at all if u was given information to go on day by day. Nothing is impossible in my mind and if there was AI who is to say we would need reality? Now we are getting into the matrix kinda... But if u think about it. As long as it seems feel, what does it matter? I mean you will always have that thing in your head saying no this is not realy. But then u have no boundrys. I could see quiet a lot of people wanting to be in the real world after there fun in cyber space. But the posiblitys are endless. Imagen going sky diving and not having to worry about dieing? Imagen been able to go into space and not have to pay millions of dollors. I do not see there been games in the future. They will be more of experiances. Thats my opinion anyway.

I kinda got a bit lost in my head with some wording and explinations sorry if its a bit confusing. It is 2:30am :D

This is exactly the kind of game that would be enabled by the creation of an ASI, and exactly the kind of game that could take the industry to the next level. I believe that this is the next true frontier for games--a social simulation (perhaps stretching the definition of simulation) that created games that really did matter--games that make the gamer think, feel, and live the world, feeling guilty for the wrongs they commit just as they are proud of their achievements and embarassed by embarrasments.

If developers can create that, they can legitimize games and create a whole new world.

Interesting. I mostly agree, but the main problem with open ended games is the lack of story. Now theoretically the ASI system would generate a story in the end, just by having large amounts of 'agents' interact a 'story' would develop. I'm putting that between ''s because it won't really be a predetermined story with a beginning and an end, but more something of a series of events which could, looking back, be seen as a 'story'.
Problem with this is that the chance the 'story' would not reach the same level of quality one would reach with a human writer. Sometimes a writer needs to make something which only has a very small chance of happening in real life/the simulation top happen in the story in order to get it to be interesting. Now if you would just implement an ASI system like that, 90% of the time spent playing would be storyless. And people like stories.

So basically what I'm saying is, you'd need more than just a social intelligence system. You'd need to find a way of 'story-ifying' the events, getting illogical things to happen at the right time. I have no idea how this could be done, but I do think it will be required. Just a game world with people in it won't be enough if nothing truly interesting is happening.

Just some thoughts :)

I don't thinka s I said they would be just a normal world with people in because thats not really a game. There would be very very basic scenarios. A good example of this would be. Basiclly, u start of as a lonly hobbit (ring a bell :P) and u are given a ring. Your objective is to get rid of the ring. First of all u would have to find out about it and see what people know in this world and also then destory it. As long as u have 1 objective that is enough to have the world go round in a sense. You play a game for the experiance, u wouldn't buy the game if u didn;t like the idea of been a hobbit and wanting to destroy a ring. My point is that all u need is something very simple setting up e.g.

hobbit
big guy wants ring back
wizards are evil and good

simple... :P

The more complex the AI is the more simplistic things get when generating scenarios. You do not need to really script everything do this and that. Let the person live there own story and is the big who wants the ring managed to leave the front door open to the big lava pit where u destroy the ring then lucky u :D this was of playing a game each person would be able to tell there story of how they did it with infinate posibilitys. You never know in your story your hobbit could be a miggit pimp with attitude that drives up to the place where u destroy the ring in his pimpmobeal. Get my point yet? :P

theGreenBunny: You have some good points there, but I think some of your skepticism comes from misunderstanding. The ASI in itself won't make a game, but as a componenet, it could make games infinitely more real. Developers would still be needed to create stories, settings, and background--it's just that the focus would change from "who can make the best graphics engine?" (as this will probably be infinitely real by this point anyway) to "who can make the alive-est world? best story?" Even if used in shooters, this would create a much more realistic environment for the game to take place in, as enemies would react much more realistically to the player (if a guard is friends with some guy you just shot, he's going to get pissed).

Thus, developers are still essential to a game, but they serve the role of storytellers, not picture-painters.

Another facet of this is the possibility to tweak the ASI and create characters. If a developer were to create such an entity, it wouldn't difficult to change some minor things about it (I would think). With tweaked characters with their own motivation and drive, a storyless world might well work, in my opinion.

This branches off slightly from the topic but I think it's relevant enough to mention.

The biggest issue with scripted games (CoD, WarcraftIII, etc) is lack of replay value. You can go through the game a hundred times and barely ever see anything new. I think that's what appeals to me so much about Half-Life2. You have the intensity of CoD, but you retain a certain aspect of open endedness by allowing the script to "think for itself" and creating variations where once there was only a single choice.

I think the next logical step from Half-Life2 is (as Pendragon already mentioned) a hybrid. A seemingly open ended game that upon certain decisions, turns momentarily into an intelligent scripted scene. That would give you the feel of freedom, the intensity of a well developed story, and the replay value of an adaptive script.

No One Lives Forever 2 is a pretty good example of an open ended game for an FPS. While it has much scripting for story events, all the AI is mostly unscripted, and is just fantastic. The AI just isn't there, they live there. Guards will change shifts, eat lunch, take naps, go to the bathroom, and even converse about it with other guards! For example, as two guys change shifts or something, they may discuss their work.

PC Gamer had an article on NOLF2 before it came out, talking about the great AI. They had a cool example; There was a room, like an office, with a single guy in it. They pointed out that, in real time, the NPC would do one of many things with the surrounding objects. Do paperwork, read a book, look at some bulletin board postings, take a nap, eat lunch, and other such things, and it was all unscripted.

I think that's only a step in AI, though. Once you start seeing more day/night cycles in games, then the NPCs are required to do something outside their little short periods you see in FPSes now. NPCs have to go home at night, they'll have to do different things in the day than in the night or evening, that sort of thing. This is why STALKER appeals to me. From what I understand from the website, it has some pretty immersive, "living" AI like NOLF2, but also has day/night cycles. Of course, due to the game's story and setting, most NPCs wouldn't do much more than curl up in a sleeping bag or trailer at night, but that's plenty an explination for me.

I think we're also starting to see some big changes in how puzzles and levels are designed. Used to be that levels were designed around the key puzzles, but now we're starting to see the opposite. You've got realistic environments because they give you access to everything that the real place would give you. A military base that lets you explore every room, and then some realistic problems are introduced once the world's created. I haven't played through much of Deus Ex 2 yet, but I hear that's how it's supposed to work. They developers said they expect people to find solutions to puzzles that they never even knew existed, mainly because of the heavy use of physics with puzzles and gameplay. Thief III also claims to do that. You get to do things like push a crate or trunk infront of a window to block the light so you can hide from enemies, that sort of thing. I think once we start seeing physics used more directly, and more stuff destroyable (especially on the Soldner level, that game claims you can use a crane to drop a tank into a building to kill the people camping inside!), we start to see all sorts of creative ways to figure things outs. I can't count how many times I saw something cool in a game and wished I could have used it to solve a puzzle my way.

But anyway, enough ranting. I don't think scripted games are going to go away any time soon, if ever. But I think since the introduction of games like Deus Ex or System Shock 1/2 (which went rather unnoticed unfortunately), we'll be seeing a lot more of them. Already we've got Deus Ex: Invisible War (even if a lot of it does suck), next year we've got STALKER: Oblivion Lost and Far Cry (check out the videos for that, it's literally many, MANY miles of open jungle on various islands, with no boundries or limits, wow). I think we're going to be seeing some major changes in game design soon.

Anable: I agree absolutely about the lack of replay value in scripted games, but I'm more interested in the creative side of emergent vs scripted gameplay. It's pretty clear, even to scripting proponents, that scripted games have little replay value, but the creative side is much less clearcut.

Emon: As to NOLF2's (great game) AI, I agree that these cycles add a lot to the game, not only in terms of visual variance, but in terms of gameplay. I would have liked to see more of an emphasis on the little things they do there, because seeing the AI do things randomly, like smoking, reading, and napping was very cool. However, even with that element, the AI was centered around fighting--the other pieces were just to prop the fighting up and add some flavor before the characters opens up with whatever weapon they have. Even when they "talk" about their jobs, it's still a scripted (or similar enough to scripting to not count)--when I played through it listening/watching for things like that, I saw lots of repeatitions (spelling?). The ASI that I wrote about above would overcome things like that to a degree (although all henchmen have the same job, so repitition would be inevitable), by having distinct personality/memory entities.

As to puzzles, I agree with everything you said: instead of the levels being centered around puzzles, developers are beginning to make part of the challenge finding the puzzle, part of it solving the puzzle (using the environment, in newer and forthcoming titles), and part of the level dealing with the solution you choose. Even in early games like Deus Ex 1, there were instances of emergent gameplay--the developers received word of ways to beat the game without a single kill (when they had thought there were 3 kills that were absolutely necessary), some of which centered around using sticky bombs as ladders (place one, jump on it, place another). Seeing more games that have elements like this intentionally is hopefully going to be happening in the near future, as physics, lighting, and AI all advance.

However, despite the bright future for emergent puzzle-solving, the potential for ASI-based games and games that focus on social situations (not involving gun barrels) won't happen for a long time, I suspect. For one example, look no further than your own post: all of the games you speak of are shooters. When I see the emergent puzzle-solving fuse with ASI, then I'll be a happy man.