For months now I’ve been hearing about the astonishing storytelling power of Portal, along with its fabulous physics and game design.
I hardly ever play mainstream commercial video games — I don’t have the hardware to run it well, for one thing — but I was really curious about Portal, so recently when I got a chance to play it on friends’ XBox, I took it.
What follows is spoilery and also probably doesn’t make that much sense unless you’ve played.
I’d been spoiled a little, both in the sense of knowing a couple of concrete plot details and in the sense that I had some idea of what people thought was so awesome about it. Under those circumstances, it’s hard not to come away a tiny bit disappointed. I’m not sure that disappointment is at all fair — this is really a fantastic game, beautifully designed and put together, just like everyone says. The puzzles are of a kind I particularly like: carefully sequenced training in a set of coherent physical rules. (I tried to do that with Savoir-Faire, in another way and another medium. Portal does it much more systematically.) It’s a funny, smooth, clever, and essentially awesome piece of work. But it is not the epiphany in interactive storytelling that I had somehow built it up in my mind to be, based on all the praise I’d heard.
There is not a particularly long or complex story here, and the techniques used in its telling are not especially new. The plot is short (one could describe all the salient events in a few sentences), and the emotional progression is simple. I went from skepticism about GLaDOS (because everyone knows that AI voices might be evil) to fear and defiance and, in the final end sequence, stony indifference to her pleas. But that’s a tiny story, very simple, very incomplete.
Speaking of back-story revelation, Portal does that in a time-honored way: by showing abandoned spaces and evidence of past inhabitants. My biggest gripe about this technique is that the first glimpses are the most effective ones, and that the more evidence the player acquires, the less emotionally effective it is, because the authors/designers don’t keep putting in surprises or new evocative information. The first time I saw a handprint, I gasped a little. The den, the scrawls of “the cake is a lie”, were chilling and evocative. But then there are more spaces like this, and for the most part those do not reveal anything terribly new. I felt the same way when playing Myst, years and years ago, and sifting through the mounting evidence against both Sirrus and Achenar, and thinking, Right, I get it already! They’re not nice!
So while I liked the bits of story-telling that Portal does through exploration/evidence, I would have liked to see more content and more surprises after the outset.
Then there’s the story in the present. Stephen Granade has praised the portrayal of the AI villain GLaDOS, who reveals her insanity and loneliness as the game goes on. I agree that was well done; but as a characterization, it doesn’t conclusively arrive anywhere that I didn’t see coming from the beginning of the game. The ending nearly pulls off something profound, but I felt as though it would have worked better if there had been a seeding of a few more backstory facts.
The final puzzle of the game is designed in such a way as to focus the player particularly on GLaDOS’ character. The timing is tight enough that one has to play it again and again, which ratchets up the sense of urgency. By the tenth time or so that I retrieved GLaDOS’ broken components from an inconvenient place in order to incinerate them, I really wanted her dead. At the same time, the repetition forces the player to hear the final, self-justifying lines over and over — giving them a weight that GLaDOS’ other, once-off dialogue doesn’t always have. In her final remarks (“No one likes you, you know”) there is a hint that she’s talking about herself, not Chell; that she herself desires an emotional connection that she has never been able to achieve, because she also needs to survive, and all humans are a threat to her. Neither the Weighted Companion Cube nor the sentry-droids seem likely to offer much in the companionship department, either. Maybe we can imagine that GLaDOS has reawakened Chell, again and again, as company — the only kind of company she is able to arrange for herself, the only kind that is safe for her because she can kill and resurrect it at will. The sentry-droids’ psychotic duality (sweet-voicedly seeking out the player in order to shoot her; saying “I don’t blame you” as they die) is reminiscent of GLaDOS, but less complicated.
I get the sense that the cake ending is not Chell’s imagination, but GLaDOS’ dream — or heaven: a place where she is accepted and acceptable, where human and AI can be friends, where GLaDOS is perceived to be as harmless and lovable as a box with pink hearts painted on the side. But she’s learned that the humans view her as potentially threatening and essentially disposable, and so she has to see them the same way. Her final core, the snarling, red-eyed core, is rage and madness, but also her impulse to self-preservation, the most essential motivation of anything that lives, and the part that remains after everything else has been stripped away.
It is also the only characteristic we definitely know Chell has.
Viewed that way, what we get is maybe a story that’s not so much the standard cliché about an AI that gets out of control, but instead about the idea that any AI created would necessarily be emotionally broken, because it would be constructed with killswitches, designed to be disposable, or at least crippled so that it could not threaten the more important human life. If the AI had any urge towards friendship or companionship, that urge would be stifled and perverted by the fact that those around it have absolved themselves (“ethicists agree…”) in advance for killing it if necessary.
That’s a sad and interesting story, but Portal stops short of completely telling it. Some of that narrative is conjecture, not made definite by any evidence in the game — especially the issue of where Chell comes from and why — and there are lots and lots and lots of alternative interpretations out there. More detail about the protagonist and about GLaDOS’ evolution could have made this something less mysterious but ultimately more powerful.
My second comment is more subjective. There were half hours of the game — especially after I’d broken out of the testing sequence — where I kept thinking, “When is it over?” I’m probably the only person in the world to think that Portal was too long. And I did enjoy the puzzles, which show masterful game design in teaching the player skills and then requiring him to use them in new and inventive ways. There was lots of great stuff in there. But at some point, past the incinerator, in the rust-infested bowels of the Center, I began to get frustrated and bored. Partly it’s that there are more and more timed puzzles. I am better at planning a strategy than I am about performing a sequence that requires good reflexes, so some sections I had to replay a bunch of times, with less enjoyment each time. But partly it was something about the distribution of new information in the game world.
Modern IF has acclimatized me to gameplay where the narrative is much denser. I don’t just mean that there’s more story served up per hour of game-play, though that’s certainly part of it. I mean also that a higher percentage of the player’s activity has some bearing on the story: exploration leads to significant new information, actions move the plot forward. It’s not that I want my game to be stuffed with more cut-scenes, but that I want more of what I am *doing* to be connected to the essential process of getting more story. The early parts of Portal actually did more of that, because new levels exposed new facts about the setting — like the existence of military droids so shiny-white and egg-like that they might have been designed by Apple. It was the late game where I got a little impatient.
I don’t mean to say that Portal was somehow designed wrong. The design is correctly balanced, I imagine, for people whose primary motivation is to experience the game aspects — the puzzles, the shooting, the jumping, the dropping of boxes onto the heads of android killers; the breath-taking flinging of one’s body over lakes of acid. And really, who wouldn’t love these things?
I find, though, that in almost any game with even a modestly compelling narrative, my primary motivation is to experience the story and find out what happens next. There are quite a few mid-school IF games from the mid-90s or so — the period when coherent plots were becoming more important but long-format, fiendishly puzzly IF was still common — where I relied heavily on a walkthrough to get through. The bother of doing the puzzle to unlock the next bit of story was sometimes more than I could stand, if I was very curious about the next event. More recent IF tends to cater more directly to my desire to get the story faster. At any rate, this is obviously a preference more than anything else.
I found there were a few times playing Portal when I felt dizzy, disoriented, and nauseated. The room span and there was a strange buttery taste in my mouth. The first time that happened, I wondered whether I was getting sick, and put the game aside in favor of something to eat and drink. The second time, I realized that my heart was pounding and my body was completely tense, and it was probably an adrenalin thing. Which was intense, but I’m not sure it was a net positive. I could see the argument that it enhances the empathy with the protagonist, but I actually found it distracting — and, for that matter, the physiological reaction doesn’t seem to have been very well/accurately paired with moments of greatest danger in the narrative.