My Photo
Blog powered by Typepad

« Far Cry 2 Map Editing Guide | Main | Live and Let Die »

June 13, 2009


Feed You can follow this conversation by subscribing to the comment feed for this post.

Clint, it's my last paragraph that really portrays my thoughts on the matter: "To really pull this off, a game would need pretty sophisticated relationship modelling, but the player payoff would certainly make the effort worthwhile."

I have no particular interest in just hiring a whole cadre of writers who can map out the entire probability space. Not only is it uneconomical, it doesn't really give the true expressiveness that games are capable of. The larger your tree becomes, the more difficult it is for subtleties to emerge.

"I strongly suspect that if we want to develop methods for enabling players to truly engage in ethical decisions (or other meaningful decisions) in games, we need to embrace the malleable, multiform, player authored nature of what a game *is*, not reject it in favor of something else simply because we understand it better."

This is what I am really about. The direction I am most keen to explore is modelling human relationships within the game engine. It's a much more difficult challenge, but one that distinctly appeals to me (I'm a software architect/designer). This is to me, the only way to have player actions directly influence the game world, rather than just reflect authorial decisions that have been made up front.

There's a lot of work that's being done on the subject. The aforementioned Façade, Chris Crawford's Storytron, the unnanounced project from Warren Spector's Junction Point studios (before they were bought by Disney), and the Sims are some examples. There's also a lot of academic work that gets published under the heading "Interactive Storytelling" that relates to the same topic.

All that said, effectively modelling characters and relationships - in a believable way - that evokes sympathy and drama for the player (and ultimately regret) is a really difficult thing to pull off. Tying that into an engaging storyline (or as Chris Crawford calls it, a Storyworld) with authored, rich characters is even more complex. Generating appropriate dialog and allowing the player to engage by making meaningful decisions is another issue entirely (which is why The Sims sticks to gobbledygook).

The question then becomes: Can we adopt some of these techniques into the next generation of games, without going all the way off the deep end? That is the challenge for game designers today.

A very interesting discussion.

I feel we need to move away from the good/evil dichotomy, or at least throw shades of grey. Most ethical dilemmas in games really aren't, they're simply ways of asking the player 'how would you like to portray yourself for the purposes of the outcome of this story?'.

I think the main reason for this is the way games end. Usually, after the player completes the game, they are shown a cutscene that wraps up the story - and, more and more commonly, presents the cliffhanger that leads to the sequel. Thus, the dilemmas we present players are simplistic, because they must necessarily lead to one or more predetermined conclusions.

Now, in videogames, unlike in movies, we have the capability to have a variety of endings, usually based on how the player navigated the content of the game.

So, as one way of creating a narrative more unique to videogames, I'd propose a solution used in film that games can probably make more of. I'm sure it has a name in film, but it's the epilogue at the end that shows what happened to the characters in the movie after it ends (see Animal House, for example).

An ending like that - and indeed events in the sequels - can probably be tailored to the player's actions during the game. The ending could be composed of several short 'epilogues' for each important decision in the game. This would allow the player's actions in the game to have the obvious immediate short-term consequences seen in the game, both narrative and mechanical, as well as long term consequences, which a lot of dilemmas in games nowadays seem to lack. From there, it probably wouldn't be too big of a leap to have these choices recorded in the game's post-ending scene save file. This save file could then be used to check what choices the player made during the game, and be used as the basis for certain events, quests, or interactions in the sequel.

Finding a way to deliver such an ending and sequel to the player in time and on budget is another matter entirely, but that's not the subject of this discussion. So I'll move on.

In my mind, the lack of long-term consequences is crippling to ethical dilemmas in videogames. In meatspace, two strong emotions associated with dilemmas are doubt and regret. Doubt of whether the choice made was the right one, and regret over the consequences of our actions. The two are probably intrinsically tied together, but I am no psychologist, so I wouldn't know for sure. Evoking these two emotions in videogames is hard, for a number of reasons:

1. The consequences of choice are usually shown to the player straight away. This is what enables players to go back to a previous save point and check out what the other outcome is. My suggestion would be to have several different sets of consequences that are revealed gradually throughout the game. For example, a game might have a quest in a village to go kill a monster that is attacking said village, stealing a sheep every other day. The locals explain that the monster used to only attack once a month, but it has become increasingly aggressive and the situation has reached the point where the mounting losses are costing them enough to pool some money together to pay someone such as yourself to rid them of the beast.

Killing the monster that is terrorising the village means the villagers like you more and call you a hero, as well as giving you much-needed money for gear upgrades, training, and other adventurerly expenses. If you explore the area where you find the monster, you find it's lair, where it's helpless young mewl at you pathetically and try in a weak attempt to imitate their mother's behaviour towards intruders. If you leave them be, they might die, or they might survive, or a band of evil humanoids mght find them and enslave them later, rearing them and using them to terrorise - or even annihilate - the locals. But the humanoids only show up later in the game, and only if you kill the mother monster. If you find a different way of getting rid of her, the humanoids still appear, but they don't have access to warbeasts of doom, so they harrass the villagers, who call you back to solve their new monster problem. If you don't get rid of the monster, the villagers are still subjected to it's depredations, but when the humanoids show up she also starts preying on them, meaning that her attacks become less frequent - until the end of the game, at which point the cubs are grown enough to cause lots of trouble to villagers and humanoids alike.

In the example, chances are the player would kill the beast, and then find the lair. They then get to choose whether or not to go back and load their previous save file, and not kill the beast, or not. Each choice has an immediate consequence, but that is not the end of the dilemma. After killing the beast, the player needs to choose whether or not to kill the cubs (if the player happens to have met a dealer in rare beasts, they could always try to find a way to sell them the cubs. That could, in and of itself, be another moral dilemma. Killing the cubs might seem to be evil, but selling them to a stranger whose intentions towards them are unknown?).

No matter what the player's choice, they will have set in motion a series of events. Eventually, the long-term consequences of those events occur, and the player is shown the repercussions of their earlier actions. With enough separation between the action and the reaction, chances are the player will have made other choices in the game, saved a few times, and progressed, to the point where going all the way back to the beginning of the game is impractical. This might, of course, raise the question of whether the player will feel regret at what they have done, how long they will feel it, and whether they will feel their experience cheapened by the whole event, or whether they will appreciate the nuance woven into a seemingly simple choice, and face future choices with some doubt as to the long-term consequences of those choices.

Of course, 'long-term' is a relative term when talking about videogames, since a game usually plays out in a relatively short period of the player's life, and gameplay itself lasts anywhere between 15 and 80 hours these days. By causing some of the player's actions in a game to have consequences, even if minor ones, in a sequel - if it is appropriate to the nature of both games - we can make the choices we present in the player even more meaningful. If the player spares a minor villain in the first game and makes them see the error of their ways, and the villain then appears as a secret recruitable character in the sequel, having reformed themselves and taken the character's words to heart, the choice of whether to spare the villain in the first game becomes more meaningful. If the alternative, killing the villain, and receiving a powerful weapon as a reward, makes that weapon available in the second game as an heirloom, we make the choice even more meaningful, in that it has very long-term consequences (and this system can be touted as a feature to encourage people who didn't buy the first game to buy it, or it can be a feature that can only be unlocked once per disk, serving as both a pseudo-DRM and a means to encourage people to buy the game first-hand, instead of second hand). It also gives us another element unique to the narrative of videogames, in that the user's actions and choices in the game can be woven into the on-going narrative of the gameworld.

This comment is already overly long, and has taken me a fair while to write, so I'll stop here and I look forward to seeing how the discussion evolves.

On Nels Anderson's comment : Regret-free decision
As humans we desire what we don't have or can't have. A decision as choosing between a range of possibility that are either presented to you (as in dialog options) or that you make up as you go (in the systemic range allowed by the game in that case). If the decision is hard to make it makes it, as apparently most people agree here, an interesting decision because it creates the art loop.

[parenthesis on what I call the art loop]
What do I mean by that ; entertainment would be a straight line (you projecting yourself into a game/movie/book) and art happens when the media itself comes back to you as a person. For instance, MGS's fight against Sniper Wolf is entertaining and the final cutscene would suddenly make art happen.
[/parenthesis on what I call the art loop]

So if a decision makes you involved in the process, if you as a the "conscience" of your character has to choose, then game art happens. Not just pulled from other medias, not just used from things that already exist, a form of feeling and meaning only available in games. Now this choice must involve you as a human being that is, if you don't stand by your choice because you have the opportunity to correct it, if you detach yourself from the game, then you are not making a choice as described above, you are trying systemic exploration which does not involve feelings, it's only us playing with the toy we have in our hands.

So the evil circle here is : you want the player to be implicated as a person in the decision making process, so you make him choose in a meaningful way so that he will have a hard decision to make and must involve himself, therefore, as a human regret his act BUT you know that regret leads out of the game... So how do we solve this ?

First lead for me is the Oblivion route : You say "ok, bad or good you choose" and then you hope that player will hold on to their choices and eventually, as in Jade Empire, help them keep the initial orientation by promising them huge rewards and making their choice visible at all time. So you just design two long term consequences. The problem is that, with this model, your player will never choose anything, he will just pick what his character would do, and could you please make it obvious ? Like Kill/Help the innocent ?

Eventually as Warrior King did very well, you can add a third "no commitment" zone (the technological "faction" of the game if you don't build churches nor satanic temples). This doesn't solve the problem : because with each new "end choice" you design, you create adjacent grey zone between this one and the others, building outcomes for these grey zones could work but, as was said earlier in the discussion, it'd take a lot of time, and will still be taken from other medias: it's not because you can choose between three lines that they are not lines.

Second lead is the technical one, if you can program AI efficiently enough, you can develop as was also said, a non-linear story. I don't really know if that's a good solution, games being predictable is part of their appeal so there might be a choice there to make into who you can present with a choice that will make a game or a moment of gaming become art.

Now is there a regret-free choice ? I think as a designer you can easily reduce the regret but I don't think you can remove it.

What do you people reading this comment think about locking choices ? Like saving automatically each time something happens ? I believe that is where Heavy Rain is going.
At least, if you don't have a choice, having to live with what you choose is one way to make people think more carefully about their choices... We'll just have to be good enough so that people will choose from a narrative point of view rather than from a systemic one.

I think it's important to understand that emotions CAN come from irreversible decisions. It's also important to know that it's not the ONLY way emotions can come. I brought up irreversible decisions because it's important to me to try to evoke new emotions any way possible, because we don't do this well at all now in games. At the exact same time, I'm fine with saying the types of games we are making now are decendant from film and aren't the best way to take advantage of our medium.

However, I'm unwilling to say that we shouldn't even try to make these designer-authored games more emotional. In fact, I think it's important we start pushing the limits of the this model for emotional impact, since we are so far behind the curve when it comes to developing games that are NOT designer-authorial in nature but still are able to create player narrative and have emotional, meaningful impact. Hopefully, something like this plays out down the road...

Clint goes out and you figures out how we can make games in a whole new way that lets us finally fully realize our potential as a medium. Then, he takes the lessons of emotion and ethical decision making that someone like me figured out in a "traditional" game that and apply some of those lessons to his new game design paradigm. And then we break through with a new type of game that also has deep, meaningful impact and emotions right away. We don't sit around and wait to figure that stuff out with the new model.

In other words, I think that by "solving" the problem in the designer-authorial context we will learn a lot about how to eventually solve the problem in the player-authorial context. Not every lesson will be applicable, but I think we'll learn enough that we'll have a leg ahead. So again, we need to do this concurrently. That way, later when both problems are "solved" we can use each other to make something truly epic.

It would be like if I started a new game project for an FPS and built an entire FPS level from scratch. Once I finish the level, my boss comes and tells me "we're changing the game to a racing game - make a new level". Yes, I am going to throw away the FPS level. But I've still learned a lot of lessons. I've learned how my pipelines for getting assets into the game work. I've learned how the engine I'm using works. I've learned something about the staff I'm working with. I can still take all those lessons and make the racing level and do that better and faster than had I started the racing level first. Yes, work is wasted when we go deep down the wrong path but it doesn't mean it's not valuable or useful.

So by doing this, we CAN keep the industry and the craft healthy when we hand it off to the next-generation of designers as you say. We can invent new models to use to build our games. None of this is laziness. None of this is a waste of time. None of this is disingenious to the full potential of the medium. It's a way for all of us, as a community of developers, to solve the problems while still making great games (and marketable games since, let's face it, many of us work for large publically-owned companies). We must try and do both as an industry.

We cannot discount the value of lessons learned from the going deep down the wrong path.


I completely understand your argument and your points. But...

As it happens I am reading Reinventing Comics right now, and at the end of Part 1, McCloud is talking about a kind of genre collapse that buried the American comics industry. Basically, there was one kind of comic that was the most popular - bullet-proof guys wearing their underwear over their pants - and that created a market pressure to make the romance comics and western comics and war comics and detective comics be more like superhero comics. Whether this involved the creation of a tights-wearing detective such as Batman or whether it involved tonal and stylistic shifts that forced the action and dramatic elements of war or pirate comics to be more super-heroic in their structure and form, the end result was the same:

They gave the audience what they wanted, and now >95% of comics are superhero comics and the medium is completely lost to general public. (This seems to be shifting slightly within the last 2 years, but since we're inside it there is no point is trying to speculate as to whether that medium will be 'healthy' again in North America at this juncture).

My concern is that the points you make might well be true in theory where you and I are free artists able to make the games we want. But we're not - or not completely. We're still answerable to the guys who make the money, and we still need to return profit (and explosive growth) on our work.

Given that:
a) the general public who are increasingly playing games are highly fluent in appreciating emotionally engaging content doled out by authors
b) our ability to deliver authored emotional experiences is improving and the games that are leading the way are more successful than those that try more game-native approaches to driving emotional engagement

I am gravely concerned that we are only a small step away from being actually (or even just practically) forbidden from exploring these new approaches because they are too risky or not profitable enough.

If we can continue to see the levels of explosive growth we see in the industry today by focusing on improving our ability to deliver authored narrative-like experiences duct-taped to the sides of our games which themselvesd remain mechanically and dynamically meaningless, then no one will ever want to finance a riskier approach that will yield much slower growth.

Worse than that, those who strive to develop the new vocabulary of expression may end up left behind.

Worse still, the audience may never realize what they are missing.

And worst of all, we may construct a new, wonderful, even beautiful new medium that is amazingly successful and leads to the birth of a new industry, but whose emotionally impactful artfulness is relegated entirely into the non-interactive authored space between the repetitive and insipid mashing of 'X' to chop down legions of enemies until you get to the next bit of Academy Award winning artful story.

That is a grand and beautiful vision - but I *can't* subscribe to it, because it literally means giving up on what I am compelled to do and letting the directors and actors and editors of the 'real meaningful parts' of the games I work on do their thing while I continue to do meaningless crap to bridge the gap between segments of their art.

Now - I appreciate your 'why can't we work together on this' approach (and I'm not being facetious, I do appreciate it) - I am concerned that this approach itself *is* what kills the future of our industry.

And I don't want to sound like a Crawfordly old curmudgeon, but I see the building pressures, and they're very real.

"Conversely - I find it intersting the sheer number of films recently that try (often too hard) to provide multiple perspectives (from Reservoir Dogs to Crash to the films of Inarritu and many others) when this idea of mutliple overlapping perspectives is non-filmic in nature. Of course you can still make a film about it, but it is working against the grain to say the least. This kind of thing is, in fact, and among other things, what games do best."

Clint: Were you at E3 and did you see Brink? The game will allow the player to see the game from both sides of a war in which neither is necessarily "right." Most games, by comparison, parse the "two sides" idea into good and evil, so I'm excited to see whether Splash Damage succeeds in creating something meaningful.

Someone may have mentioned this above -- forgive me, I haven't read every comment -- but there's a way to have save games and still provide moral/ethical ramifications of your decisions: any game that tracks the moral dimension of your behaviour does so using some sort of stored variable, so the game just has to make that variable persistent despite saves. In other words, if you save at a certain point and then do a bunch of evil things, and then later decide it would have been better to be good, when you revert to your save your "moral meter" will fail to revert, instead reflecting the actions you took after the last save.

...Yes, but that would be kind of like 'not saving', right? I'm not saying it isn't an interesting proposition, but what your telling me is that I can go back in time, but I have to deal with issues I did in the (last) future... Like, I could be punching someone in the face, then revert to the last save, but still have a sore hand and an angry guy standing opposite of me :-D But in all seriousness that would become really hard to track as a player after several retries, not to mention the development. But then again, we're not talking about an easy solution here.

Hmm, Brinks sounds interesting, but it would become really hard to see the adversaries as the good guys once you have experience it from one side I think, not impossible probably as some movies revolve around this and succeed.

@ Clint: Stick in there man, keep fighting that battle. I truly believe there are ways that mechanics will be able to tell stories and that the player can experience the story through them. Your reading a great book by the way, and like comics we have our own language, our own way of presenting and experiencing a story. We just didn't find it yet, but I'm guessing with lots of people trying out different things it is bound to happen somewhere this coming year. Keep it up!

To push the same argument from a slightly different perspective (though now we're not talking about ethics, we're talking about how we make meaningful decisions in games, with ethical decisions being only a part of that)...

I just came across this post over at Brainy Gamer that criticizes the industry's resignation as viewed through the lens of E3.

One thing this touches on - and something I've been thinking a lot about recently - is the psychology of how we behave in games as opposed to how we behave in life. Its funny how these things happen at the same time - there's a relevant "experiment" going on just now as well, in which people are playing Far Cry 2 and accepting any death they have as permanent.

On the subject of save/load and player death, I'm split. I don't have a clear opinion of my own at the moment, but I'm going to play Devil's Advocate for a moment and call Clint out on something: are you sure your insistence on the save/load system being part of what makes a game a game isn't simply your attempt to avoid finding a better way of dealing with player death in your compose/execute/recompose cycle of failure, as shown in your GDC slides?

I would thoroughly agree with you that dealing with player failure is part of what makes the gaming experience unique and apart from the cinematic experience. I disagree, however, that we can't make use of regret and that players should always be able to have a "do-over".

Couldn't we just deal with player death in a better way? Perhaps we're still into branching vs emergent narratives again, but if we're in control of the game world, why don't we just cover player death in a way that does both?

Let's take Far Cry 2 as an example. Now, I'm riffing this from the top of my head, so its not going to be perfect, but what if you couldn't really die in the game? What if, every time you died, the game would automatically save your progress at that point, and you'd have to start over with a randomly generated character who has been sent to pick up the slack of your previous player character? You'd get your stuff back, your diamond account would be transferred over, everything remains unlocked, but no-one recognizes you any more.

The aim of this is to make it technically little more than a brief set-back, but to discourage it with grave emotional impact. Nobody LIKES to fail, even within a game. In the example you gave, Clint, of two monkeys play-fighting - yes, it is important that the monkeys have agreed not to kill each other. That would be the social contract that creates the "magic circle", surrounding the game and the player ( But it is also important - at least to that particular game - that neither monkey wants to lose.

(Its still a game if one does accept (a "training fight") but its a different game. I'm not sure what that means for this discussion though)

Getting back to my original point on the way we behave in gameplay experiences, I also wonder what our goal in this area should be. To create an experience so realistic, with so much immersion, that a player plays as though it is real life is to make something other than a game. Clint is correct in this assumption. If I want to run around "real" Africa and be shot at all the time, I'd fly there and ask people in bars about diamond mines, and last about five minutes. A good example of this is in Far Cry 2 itself - the "magic nut" that fixes every problem with your car in seconds. On the surface, a ridiculous concession to gaming mentality, but its acceptable within the environment, and streamlines our passage through unnecessary drawbacks of reality to the experience the designer(s) intended us to have.

Is death one of these unnecessary drawbacks, or is it part of the experience? Should we be streamlining past it, or integrating it into the environment?

I think there's a line we need to draw in the sand between creating the "real" experience and the "gaming" experience, and I think what it comes down to is knowing what is the essence of the experience we are trying to convey. I'm planning an at-length blog post when I can find the time, but the best example I can give is the new control system in the upcoming Splinter Cell shown at E3 - we don't want to be some awkwardly-moving mannequin with the potential to be Sam Fisher if we can work out how, we want to BE Sam Fisher, a man who doesn't miss his head shots, doesn't stumble awkwardly into his targets, and doesn't mistime his steps.

And finally, on the subject of emergent vs branching stories: I appreciate both sides. Branching stories are a concession to technology - I think we can all agree that emergence is what we want. The problem is people saying "it can't be done". Well, I'd like to throw in my two cents: NOTHING is impossible, especially in the world of game development. Some things are just really, really, REALLY difficult. So, to those working on emergent narrative systems: stick at it, we'll get there, and we'll all be better off for it.

Verify your Comment

Previewing your Comment

This is only a preview. Your comment has not yet been posted.

Your comment could not be posted. Error type:
Your comment has been posted. Post another comment

The letters and numbers you entered did not match the image. Please try again.

As a final step before posting your comment, enter the letters and numbers you see in the image below. This prevents automated programs from posting comments.

Having trouble reading this image? View an alternate.


Post a comment

Your Information

(Name is required. Email address will not be displayed with the comment.)