Monday, July 21, 2008

Gamers are Maximizers of Utility


-- Coletta Factor AKA Spoiler Warning: GTAIV--

Most modern games tell a story by superimposing a narrative logic on the logic of gameplay. The latter consists of the rules in the game-world that define “progress” in the game: how you control your character, how you get to the next level, how you defeat enemies and avoid dying. The former consists of the conventions of storytelling: making characters with cogent motivations and giving them a plot with includes difficulties to be overcome. Because these two governing logics work side-by-side in most games, the player can give two sets of explanations for why she is doing what she does in a game. On one hand you are flipping out and killing hordes of ninjas because it will allow you to clear the next door and move further in the level, and on the other hand you are flipping out and killing hordes of ninjas because those ninjas stole your girlfriend.

Games are successful to the extent that they are effective at conveying (with some degree of subtlety) what you are supposed to do (what counts as success in terms of gameplay) and offers you a set of interesting scenarios in which you pursue that end. For this reason, getting through a game-- figuring out its gameplay-logic-- involves some complicated pieces of means-ends reasoning. You know that you are supposed to reach some room or platform, it's all just a matter of how, and reasoning out the means to that end can be tricky. The story logic, on the other hand, is usually not left in the player's hands. The designers and writers give you bits of story in cutscenes between stretches of play that indicate why the player's character is so bent on killing that horde of ninjas or getting to that platform. Because story isn't left in the player's hands, games are very good at inculcating good instrumental reasoning (how to get ahead in the game-world) but are bad at making the player reflect on the exigencies of character and fate that make for a story.

I bring all this up because I read an interesting article on MTV's Multiplayer blog about a central decision in Grand Theft Auto IV. In the course of the game, the player becomes embroled in a conflict between two of the major supporting characters in the game-- an ex-con named Dwayne and an up-and-coming street hustler named Playboy X. The protagonist, Nico, spends some time assisting each of the two characters in some criminal activity during the game but tries to avoid taking sides as enmity grows between them (Playboy had taken the reins of Dwayne's criminal enterprise while the latter was serving a bid, and Dwayne becomes jealous and depressed when he is not given his leadership role back upon his release) Each of the men asks Nico to kill the other, and then the game tells you that you are to make the choice.

What I admired about this scenario is that the choice carries a surprising degree of moral subtlety. Unlike the clear-cut binary decisions in games, which I decried in a previous post, room is left for ambiguity in the game's presentation of this decision. Both characters are portrayed as men with recognizable virtues and flaws as human beings-- this speaks to the general strength of the scripting-- and the game provides you a well-fleshed-out picture of each character's perspective on their conflict. When I was faced with the choice I thought about how I would make sense of that decision in terms of my understanding of Nico's character and motivations. Maybe there is something inherently corny about buying into the game's fiction this way, but Duncan Moore of Insomniac took the same tack: “I felt a certain empathy towards Dwayne and felt that ‘my’ character was a man who upheld a certain moral code. As cold, decisive, and murderous, as he was, Nico was a man who related to the suffering of his fellow man (at least this was the side of his character I chose to relate to). I just couldn’t put Dwayne down.” I thought that the very fact that the game could prompt this sort of reasoning was an achievement.

However.

The game rewards the player for killing Playboy X: after you kill him, you get to use Playboy's apartment as a safe house. Killing Dwayne gets you nothing. I thought this decision to attach a reward to the choice represented a failure of nerve on the part of the designers; by wedding a gameplay-reward to a decision that ought be governed by one's sense of character and motivation, it turned a moral decision into a straightforward piece of instrumental reasoning. As Tim Schafer reported, “My moral choice was corrupted by [gamefaqs, a site that offers how-to guides for games]. I heard that if you kill Playboy X you get his apartment. So I did, and you do. I have no regrets.” Most of the interviewees mentioned that they thought this gold star affixed to the act of killing Playboy X marked an endorsement of their moral choice, and I thought this gesture ruined the game's otherwise laudable attempts at narrative ambiguity. And it also revealed how much gamers, as a class of people, seem to be guided by a purely instrumental imperative: when presented with a narrative decision, gamers find out the consequences of that act by going to the game's walkthrough and then make the choice that helps you win the game. It's what we've been trained to do.

8 comments:

Bobby said...

The use of walkthrough to make moral/gameplay choices is a very poignant one. Toward the beginning of the game I looked at a guide to see what the result of my choice of to-kill or not-to-kill was going to be, though by the time I got to Playbox X / Dwayne I had stopped referencing outside material to make my own choices.

This observation raises a handful of questions. The first relates to the use of walkthrough. While walkthroughs are not cheating, they do lay outside of the game's gameplay/narrative structure that you laid out. It's extra-narrative and extra-mechanic in terms of the game's world, but has become an ever-increasingly important part of the game industry. After all, Rockstar gave Brady Games the material needed to have their strategy guide published the day the game came out. And game publishers are also aware of the tenacity of the game playing public and their quick GameFAQs submitting fingers. This begs the question of how many people actually look at these walkthroughs when it comes to making moral decisions. If it is indeed a lot, then something needs to be done to make these decisions more surprising and impactful.

Another question is revealed in terms of decision-reward structures. If you're not looking at a walkthrough and choose to kill Playboy X and are rewarded for doing so does this necessarily show a bias for the "right decision" in the game. I felt that Rockstar has set up that bias in terms of narrative already. It's really hard to empathize with Playboy X. He lives outside the way of life with which Niko has aligned. The heavy-handed narrative from Rockstar seems to imply that getting Playboy X's apartment shouldn't be considered a reward so much a not getting it is punishment. This of course may just be my reading of the game, but I felt that the designers had a general trajectory for Niko's character even though it was "open" on the surface.

People get down on Grand Theft Auto in this area because they expect "open world," whereas I've found that the GTA games are anything but. Grand Theft Auto games allow freedom on one axis of your gameplay/narrative structure, but the latter is much more finely controlled. Sure CJ can fly around on a jetpack to the top of Las Venturas buildings and snipe people from the roof, but that open play will not progress the game, as you've written. It's up to the designers to strike the balance of what moves the game forward.

So what do we do about this climate of decisions affecting gameplay. Should all games look like Knights of the Old Republic with an alignment system that changes the narrative? Does a different-but-equal reward system mitigate the inherent problems with moral decisions in narrative? What about randomly or procedurally generating results so that everybody's gameplay experience is slightly different? Or is it okay to reward decisions based on the ideology of the game?

What's most important is that game designers try all these different things so that we continue to have a wide variety of game playing experiences.

Steve the Creep said...

I've read about this moral question in GTA IV before. The thing that caught me is that you have to make the decision. There are no other options. Kill this guy or kill that guy.

Seems to me that that's the majority of the moral choices you can make in games. "Yes, I will protect your farm" or "no, I'm going to steal all your sapphires."

I liked the concept of morality in Fable because it changed what you looked like. But other than an alternate ending, what does morality choices get you in any games? A different ending? I really wish there was a way to go bad and it turns into Overlord where you can control the evil creatures and wipe out the cities and heroes.

Unknown said...

Thanks for another great post!

I guess I remain unconvinced of the notion that there's a fundamental dichotomy between narrative and gameplay. Isn't the reward of Playboy X's apartment in fact a narrative reward as well as a gameplay reward? Isn't gameplay actually a mode of narrative here and elsewhere(or rather, more basically, isn't narrative actually a mode of gameplay in the pretend-game [or the PPP] between storyteller and audience)?

At any rate, I find it pretty easy to imagine a better version of the GTA4 story in which the gameplay is better integrated with the rest of the story, and the player gains advantages from either choice--ideally advantages that are commensurate with the choice s/he made.

I guess that means we agree around the back--for you I imagine that better version would simply be the gameplay and the story, as separate practices, complementing each other better.

Unknown said...

If I remember correctly, killing Dwayne actually does have a "reward" attached to it: you make alot of money. At least 100,000 if I remember correctly, but that could be a bit high.

Either way, each choice presents a unique "reward" to the player, though one is a mostly useless safehouse (there is another safehouse better located on the central island) and the other is money, which is only useful at that point in the game if you haven't been able to afford all the weapons already.

Iroquois Pliskin said...

@bobby: thanks for the response, there's too many good points you bring up here to do justice, but I had a few thoughts: first off I think that cheating by looking at gamefaqs is a permanent part of gaming. We've all done it when we get stumped and want to make it further in the game. But the problem is when the aim of "beating" a game defeats the effects of the designers to instill ambiguity into the story. Decisions have more weight when we have to make them without knowing in advance what all the consequences will be. Second, I think you point to the very reason why there are so few satisfying attempts at moral decision in games: it's difficult to design for. How do you represent all the consequences of moral decision in the game without making the game unmanageable to design? I think you're right that we're gonna have to see what designers manage to cook up: I'm excited to see what Bethesda does with the Fallout series and their attempt to incorporate choice.

@steve: good point. In the article I pointed out Tim Schafer talks about this very point; in some of the missions (like the one where you kill a gay man for Brucie or the crusading lawyer for Francis) you wish you had a option to walk away from killing anyone at all.

@Scott: thanks for pointing this out, I didn't know you get so much money for killing Dwayne. Maybe I've gotten all this wrong and the two rewards are meant to represent the choice between money (Playboy X) and loyalty (Dwayne). In that case, the designers deserve more credit than I give them, since the conflict between old world (loyalty) and the new world (money) is a thematic anchor of the game's narrative.

LJS said...

A lovely post, and a subject guaranteed to get attention, but you knew I would call you to task for said subject not having much bearing on the topic at hand, correct?

Anonymous said...

I chose to spare Dwayne because the game itself told me to. By the time the decision becomes relevant the game had forced me to micro-manage Nico's friends to the extent that I was seeking a thumbs up from them to reach the 80+% mark that results in an in game benefit for all the work getting them there. Thus I had premeditated a decision because Dwayne had like and respect scores and offered tempting new powerz. Playboy X did not, and was marked by the game as a dead end. The loft becoming a surprising and largely irrelevant bonus. I was definitely seeking to maximise my utility. And for what it's worth, Dwayne's "power" was seriously maximal, if used.

As a side note, the dilemma of choice that exists between Playboy X and Dwayne is similar to that in the end game. Friendship / loyalty / honor vs greed. It's plausible that this was intentional, but the other main choice killing moment (Packie's brothers) has completely different themes (Corrupt lawbreaker vs idealistic druggie). Suggesting that it's just a happy coincidence.

Glen said...

Thank you for writing this post