In today's Game Set Watch, movie and screenwriter Justin Marks takes the game industry to task for calling the story in Grand Theft Auto IV "Oscar-worthy" and wonders if gameplay as narrative is the answer:

The adventure of Niko Bellic, complete with its comic assortment of ethnic cliches, is pretty much on par with the rest of the franchise's self-conscious worship of movie archetypes and genre tropes. And there's nothing wrong with that. Rockstar has made clear that's all they've ever wanted to do, and they've done a damn fine job at that (although I do miss some of that charming humor from Vice City and San Andreas).

The problem here is not the quality of the story, but the manner in which it is incorporated into the gameplay. After skipping over countless cut scenes so I could get to the action, I slowly began to regard plot in GTA IV as being something akin to the Clinton marriage: why do they bother with the charade? Is there anyone in this country who honestly thinks these two people still sleep in the same bed?

After all the incredible advances in their game engine, why does Rockstar insist on making its story an accessory -- a needless, comparatively inferior element? More to the point, how did narrative become such a side bar to the real point of gaming, i.e. our ability to play out our deepest fantasies in a virtual world?

I found myself nodding in agreement at the start, but then wincing at some old, overworn ideas as his essay continues. By the time the essay starts to near the end, Marks is returning to the same old obvious claims that many game writers wind up making:

We need to stop thinking about story as a device to make us care about the gameplay (it doesn't), and start thinking about the gameplay as the narrative itself (thus, making us care). Now that the technology has finally reached a breaking point, a place where we can genuinely craft sophisticated worlds, we have to understand that plot is not forced upon those worlds artificially, but grown from our interactions within their environments.

Story design needs to be less checkpoint-focused and more focused on implementing a meta structure that makes us believe we are shaping events with our choices, even if these choices have already been made for us.

The "story on rails" has now been exposed. Game engines are strong enough that we can see the seams in the narrative fabric. It's no longer acceptable that we can take our girlfriend on a date and never once have her mention the fact that we're carrying a missile launcher by our side. We need to believe our actions have consequences within the virtual universe and that the experiences we are living are wholly unique, even if they aren't.

This is all very, very old news. His assertions and observations are fair enough, except that like all generalizations, when extended out to encompass everything it falters and fails. The truth of the matter is that in some games, having the interactive bits lead to stories on rails works very well. His timing for this assertion is especially unfortunate given the relatively recent rumors that Metal Gear Solid 4 will have 90-minute cutscenes. I'd be willing to bet that the people who have stuck by Hideo Kojima so far are more than happy to sit back and watch as his "story on rails" unfolds – which illustrates my contention that the issue isn't with stories on rails, it's with bad stories on rails.

I for one love a great story on rails, as evidenced by the number of Final Fantasy games on my shelf, but I have little to no patience with bad stories on rails, which is why after playing Lost Odyssey for a couple of hours I flatly lost interest. The game had some interesting premises, to be sure, but it squandered them way too quickly. Lately I've been anxiously awaiting MGS4 even though I haven't played through the first three, opting instead to catch up through the excellent video retrospective series being offered up by GameTrailers. It's cheaper, sure, but more to the point it takes up much less time – although I never use Cliffs Notes for books and still largely resist using hintbooks for games, when presented with the option to get caught up on the Meta Gear story through these summaries instead of playing through 100-plus hours of gameplay, the decision was an unfortunate no-brainer. I don't know about you guys, but I don't have that kind of time.

I still think that the best way to handle interactive narrative in games is to treat it like a series of rubber bands strung between nails – the key plot points are fixed (what Marks refers to as 'checkpoints') but the manner by which you arrive at those points is flexible. This is the philosophy you often find deployed in games with lots of side quests or mini games – they improve the quality and the duration of the game, but they still remain optional. I'm not a big fan of sandbox games for many of the same reasons cited by Marks, but I remain skeptical that the Crawfordesque, Holodeck-esque model that he's wishing for will ever be a realistic scenario.

What I want is the opposite of Marks' prescription: I think game writers should write better stories and work with the game designers to develop better game mechanics to mesh with the narratives. Despite the frequent claim (that Marks himself makes near the end) that "the game industry is not the interactive little brother of cinema", I still kind myself marveling at how easily these types of claims map onto criticisms of film. People that claim that narratives in games should take a backseat to gameplay strike me as characters that claim that narratives in film should take a backseat to cinematography. It's a short-sighted, tunnel-vision type of claim – because X is what media form Y does uniquely and independently, then all instances of media form Y should focus almost exclusively on X. It's a bad model and a rotten philosophy: many films do okay with an iffy story and spectacular cinematography, and many films do okay with an amazing story and mediocre cinematography. It's the ones that do both brilliantly that truly prove themselves memorable.

I think that the proper first step is to determine what kind of experience you're trying to produce when everything is said and done. This will allow you to start deciding what type of narrative experience or gameplay experience is best for what you're trying to create, and then to develop an appropriately matching narrative or gameplay right along with it in an organic, intelligent fashion. Let the ratio of gameplay to narrative – and the ratio of interactivity to 'rails' – be determined not by your media type but by the type of experience you're trying to create. Just like with narrative and cinematography in films, an ideal blend of gameplay and narrative is the holy grail – but what that ideal blend happens to be depends wholly on what your desired end experience happens to be. There's room enough in an entire media type for a wide range of experiences and ratios. Just because you don't happen to like games with stories on rails doesn't mean that they shouldn't exist. I think that to assert, as Marks does, that we should "stop writing high-minded stories. Start writing games. And let the stories grow from them", is way too one-sided and, frankly, way too simple-minded.

(Update: Kojima Productions has since issued a correction to the rumors, stating that there are no 90-minute cutscenes in Metal Gear Solid 4. Still, I think my original argument stands – MGS is a solid (no pun intended) example of a linear story that unfolds through interactions with the player, and since it has enough fans to bring Amazon.com crashing to its knees when it goes on presale, well, then, I still say Marks' insistence on Western-style nonlinear narratives is overreaching at best.)