It's a grim fact that a lot of games that set out to critique war, video games, or both tend to work out very, very badly---more Lions For Lambs than Apocalypse Now.

Partly it's a conflict between making a good game and critiquing game conventions---that's a tension that can, in theory, be productive, but mostly just leaves you with a game unwilling to be good out of misplaced guilt, and scolding you for whatever enjoyment you do get out of it. That would be the problem with your imagined scenario---you would end up with a game that seems to punish you for enjoying it, which is a problematically self-destructive attitude to take ("You think I shouldn't play this game? You got it!")

Part of the trouble, too, is that games that try to engage with real-world war often make grander claims for themselves than justified---making your game more like real war only exaggerates, for me, how unrealistic the game really is. All the brutality of, say, the Soldier of Fortune games doesn't move me to reflect (any more than a standard war game does, which is somewhat), it just makes me say "Hunh, they put more blood in this game." If your goal is to comment on how video games affect real soldiers' perceptions of violence, you're better off making a documentary, with real soldiers in it; a medium is best at commenting on itself, not other mediums.

This is part of why games work so comfortably in "genre" territory---a more stylized story, detached from real events, is better suited to the basic unreality of the gaming medium (where everything, from the people to the ground, is artificially generated), and gives you a much more honest place from which to speak.

...you would end up with a game that seems to punish you for enjoying it.

Well, I'd hope you'd actually end up with a game that asks you to play it within new restraints--don't torture people, don't shoot innocents, don't kill if you can persuade. It wouldn't necessarily be the same game promised by the first ten minutes, and if that's punishing then yes, I can see the issue.

All the brutality of, say, the Soldier of Fortune games doesn't move me to reflect (any more than a standard war game does, which is somewhat), it just makes me say "Hunh, they put more blood in this game."

Yeah, that's a good point. I suspect you have to engage intellectually more than viscerally when it comes to games. My thought was that the shift itself, from bloodless to bloody, would provide the engagement that Soldier of Fortune (which is, if I remember correctly, fantastically violent from the get-go) does not.

But I disagree that a game can't explore these issues. It won't explore them the same way that your hypothetical documentary does, and I don't think the topics are directly comparable, yet there had better be more room for introspection in gaming than currently exists. If there isn't, the medium's in trouble.

By punishing, I mean: If you've built a game where the most enjoyable way to play is by using violence, and then smack the player around for enjoying the violence that you've crafted, you're punishing them for doing what they're supposed to be doing, and have broken your contract with the audience. It's like the many horror movies that show some lovingly-crafted piece of violence, then show some villian enjoying the sight of the violence as a way of condemning the audience---a bait-and-switch that leaves you with nothing to enjoy.

You could make a game where negotiating and persuasion are more useful tactics than shooting, but if you do that, you must make a negotiation/persuasion engine which is as viscerally enjoyable as a shooting engine---audiences won't (and probably shouldn't) accept a less fun game on the basis of it being better for them. That doesn't mean I don't think games can't be more introspective, or do more thematically, but I'm not sure action shooters are the best place to take this particular anti-violence stand.

A more productive route, if your goal is to make a game with a strong anti-violence position, would be to step out of the action-shooter approach entirely. Like, make a game where you play as a civilian in occupied Iraq trying to get home. You're completely outgunned by the occupiers---any attempt to get a weapon will result in a suicide attack (which many players will doubtless do anyway, which is interesting in and of itself). Instead, you have to play it as a stealth game---deciding which checkpoints you can get around, or when to follow the rules, and so on. You can level up your negotiation skills by taking English lessons (some word jumble minigame, like in Bully), but negotiations are always a series of reflex-based timed events, so they're always a tense challenge---more challenging, and therefore more fun, than the hopelessly quick death of shooting.

Or you could dive into Nintendoworld, by making a game about a bunch of magical grass-fairies whose world is being trampled by big heavy army guys, and you must use your platforming mojo to somehow convince them to leave, or use building skills to protect your village, building by building. Either way, if your point is to make a game about how bad military violence is, you probably shouldn't do it in a military-shooter engine, unless you have a very, very clever trick to make the essential sarcasm of your approach work.

A) I'm not sure the point is anti-violence. I'm actually just more interested in having players acknowledge the violence and think about it. I like recreational and harmless violence, personally, but I don't think it should go unexamined. Warren Spector's distinction between "aggression" and "violence" is useful here.

B) I do take your point about making the game fun, but I don't think an FPS engine is intrinsically meant for military shooters. Portal is a weak example, but I think it does have a point--a game where the gun doesn't have any direct effects on the enemies at all.

So yeah, you've got to make a decent mechanic for persuasion, and I have no idea what that is. And you probably would have to add elements of stealth to it.

As you mentioned in your post on the GTA games, this is kind of where the rubber hits the road. How do you combine game mechanics, which are slightly surreal and mechanistic, with real-world simulation? I agree with you, I think, that GTA shows it can be done in a thought-provoking manner. Unfortunately, it also shows that most players will ignore that subtext, and just shoot people.

Yeah, the question of mechanics is a biggie. One thing I really appreciated about Mass Effect was that it at least attempted to shake up the mechanics of conversation---something like that seems central to making a persuasion and negotiation game fun.

Often, I think too much interpretive weight is put on a game's story, and not enough on its mechanics, even though you could argue that the mechanics are the real semiotics of the game; a similar problem afflicts film reviewing, where many people write as though the script is the film, and ignore the effects of things that're harder to talk about, like cinematography, editing, and sound design. Hence my GTA post---horray, it was read!---where I tried to more or less ignore plot in favor of talking about the thematic weight of mechanics.

Portal is a great example of how long-established engines can be upended---I still love Bonnie Rueberg's "Portal Is For Lesbians" piece for its wide-bore approach to how Portal subverts the last letter of the FPS acronym, and what that could mean. Then there's the larger question of how games, which are basically all about agency (according to Will Wright), can deal with questions of loss of agency---Bioshock did wonderful things with the killing of Andrew Ryan, but fell down in the last couple meters of track by ending with an utterly ordinary boss fight.

In the case of GTA, people ignoring the mechnics to shoot people isn't entirely a failure of design---besides it being a very violent game even if you do play by the rules, the whole point of GTA is that it's a mayhem simulator (unlike Oblivion, you can't just keep your head down and get a respectable job). Now I'm all for some good mayhem, and I think GTA actually does some very smart things by building its stories around the consequences of anti-social (in a literal sense) behavior. But it's not quite auto-critique, at least not yet.