Analysis: Three Roles I Played In L.A. Noire

[Writer and designer Emily Short outlines narrative inconsistencies in Team Bondi's L.A. Noire in this Gamasutra column, carving her own path in the game to bring it closer to home.] Tom Bissell takes L.A. Noire as final proof that "interactivity sabotages storytelling," because the game is so good in so many respects and yet the story is often undercut by the play. Much as I like his read overall, he's wrong about that. The problem with L.A. Noire's story is not the presence of interaction. The problem is that the interaction is so inconsistent about how it positions the player relative to the story. I found myself taking three different approaches to the story -- one I think the designers intended, one I think they should have intended, and one they couldn't possibly have envisioned. [I am going to discuss elements from throughout the story. It's best to play before reading. Spoilers follow.] Cole Cole Phelps is not a player avatar. He's his own character, with his own history and personality. We know things that he doesn't, and he knows things that we don't. The game breaks this fact to you slowly. We get used to controlling Cole, making his decisions and advancing his career, but gradually as we go we realize we don't actually know the man. He mentions a wife and kids. We don't see them. He goes into rooms and we don't get to follow. He makes choices on his own, ones we might not agree with. Finally our viewpoint comes completely unglued from his and we get another protagonist to control some of the time -- and this happens right about the time that we're realizing Cole isn't what he wants to appear. It's not a flawless arc. It takes far too long to reach interesting character material, and we lose a lot of precious time at the beginning seeing Cole as a somewhat stiff person with bizarrely angry episodes during interrogation. Still, it's a perfectly workable design decision to say that the protagonist has a distinct personality and that the player doesn't totally control him. But it's not clear what the player is supposed to be accomplishing via interaction instead. Rising to a challenge? Demonstrating understanding of the mystery plot? Testing how far she wants to be complicit in Cole's story? Developing empathy for the characters and learning to treat them as people? The design is not consistent. For one thing, the systems for driving, shooting, and brawling are designed with opposing approaches to success and challenge. Reviewers have said that it's problematic for Phelps to be allowed to drive psychotically; I'd say the issue is not that the simulation fails to impose narratively appropriate limits, but that the inappropriate behavior is what the engine is designed for. The GTA-style driving is tuned to encourage mayhem. The controls make it easy to drive recklessly and a lot harder to drive well. The display makes it easy to miss red lights or not see an oncoming car. (A certain amount of bitterness speaks here. I would commandeer a sporty little convertible because I liked how it looked, and then my blundering driving would smack it up, and I'd need to commandeer another one, and that wouldn't last long either. It gave me a complex about how I couldn't have nice things.) I'd take all this just as a sign that I personally am bad at the driving system, but friends who are better at such games report a similar experience. Besides, this system is at its juiciest when you're screwing up -- plowing into taxis, demolishing fire hydrants, taking turns with a squeal of brakes. Contrast the systems for shooting and brawling, which focus the player's attention on doing the right thing. There's room to fail, but by default Cole tends to aim in the right direction and punch the right characters. Hitting things is rewarded more than failing to hit them. Admittedly the brawling gets fairly same-y after a little while, but the systemic goal there is success. Is there any reason the driving has to be failure-optimizing, other than that Team Bondi had access to the GTA engine? What if, instead, it were fairly easy to avoid crashes, and the juiciness came from being able to (safely) point the camera in other directions, to look around and sight-see while you drove? Wouldn't that be both more in keeping with the game fiction and a better use of the superlative scenery? Meanwhile, the interrogation and investigation systems are partly designed as though their purpose is to test the player's comprehension, a common trick of mystery games, but they don't really work that way. Point-and-click adventures, when they're going well, are about parsing your environment and figuring out what's interesting there, not about the pixel hunt. (Andrew Plotkin's essay "Characterizing, If Not Defining, Interactive Fiction" (available in the IF Theory Reader) is a good read on this topic.) In L.A. Noire, I almost never found that I could pick out the objects from the environment that were supposed to be meaningful, with the exception of the actual corpses. The rest of the time, who knew? Sometimes a photo was critical and sometimes it was just a piece of paper. So it was entirely a game of moving carefully around the perimeter of each room, waiting for the controller to buzz, dowsing for clues. Fairly often by the time I reacted to the simulator buzz I had already moved outside the area where the investigate button worked. I spent a huge amount of time moving Cole back and forth minute distances to get him on the hot spot to pick something up, and often the something turned out to be a carrot or a cigarette butt. There is nothing about that process that communicates plot, setting, or character. Likewise, the interrogation mechanic tries to be about reading faces and reacting accordingly, but the brokenness of this system has been so widely commented on that it doesn't really need me to rehearse it again. It didn't encourage mastery (because there was so little consistent feedback about what was going on) and actively impeded comprehension (because once you'd missed a piece of information, it might be gone forever). Besides, as the game goes on, we more and more know things about the outcome that the protagonists don't; we're ahead of Cole in knowing that a serial killer is at work and we have the backstory about the arsonist long in advance. Daniel Weissenberger is right to argue that L.A. Noire tips its hand too much to be mysterious, but I disagree that the plot is the problem here. The writers are telling a story about how Cole Phelps and Jack Kelso come to terms with the non-heroic nature of humanity. Who did what is only part of the story; the rest (true to noir convention) is about how the characters deal with what they discover. Unfortunately, the investigation gameplay is designed (with some flaws) around the idea that the story is a mystery, rather than a melancholy exploration of human nature and the brokenness of a particular setting. Silent-Movie Organist The best gameplay in L.A. Noire is the gameplay of texture. Hours upon hours of playing in a sunlit Los Angeles gave way unexpectedly to rain. The atmospheric effect was intense. I didn't know the city could look this lost. I didn't know the game engine could produce this watery sheen. A rich man lived alone up a winding road on a big estate, and driving to see him made the point of how different his neighborhood is from everyone else's, how wealth insulates and isolates. Towards the beginning of the game I saw a truck I wanted to try driving, a fancy multicolored funny-shaped truck, and I chased it down the street half a block, like a sad dog, before I realized I wasn't going to get close enough to press Y and take it for a spin. And then I didn't see that truck again, not to notice it, until the very end, at one of the pest control companies. It was their truck! Parked in their parking lot! The world is full of things that look like mere scenery that turn out to be important, hours and hours later, in a very natural way. The further you go into the game, the more tightly the fabric pulls together, creating the sense that everything is part of the same conspiracy. These textural aspects are why -- despite my distress about wrecking up the town repeatedly -- I kept on driving to case destinations right to the end of the game. It felt wrong to skip these interludes. Their meditative quality reminded me of some of the long passages in Heat without dialogue, only music. The time spent simply being in this world, doing simple tasks to find out how it felt, gave me context for the plot elements. It gave me time to think. It gave me time to accept the reality of the world I was in. Driving did this best, but it wasn't the only system that did. I went to look up land registry information, and I had to use a quaint old calculator to gain access. That was a bother, but I didn't hate it -- because it experientially communicated something about the rickety nature of period infrastructure, the way that information storage and retrieval was slow and bothersome, and how this played into the kinds of things that could happen. So here's what I think the gameplay should have been to best fit and complement the story: a way of engaging with the setting. Gentle exploration mingled with performance. An opportunity for the player to improvise the thematic passages that round out the plot pacing, like the organist accompanying a silent movie. L.A. Noire has a lot of the right elements to do this, except that many of the elements are tuned wrong, and tuned wrong in different directions. The driving needed to be easier, the brawling juicier and more engaging. The interrogation needed to be more forgiving, encouraging the player to relax into the intuitive face-reading, and focusing as much on forging connections with the characters as on guessing the right level of doubt. The investigative portions needed to lose the beer-bottle red herrings and respond better to attempts to examine something when the player has just stepped out the regular hotspot: they would have been much more fluid to play without losing anything of value. As for the more ludicrous platforming bits, those could have been dropped entirely. They were just hard enough (at least for me) to make me fail a few times and lose the pace of the action, but not hard enough to be actually interesting as a tactical challenge. And they pushed the fiction to the breaking point, as the tar pits scenario more or less openly admits: if our partner has a rowboat handy all along, why go through all that nonsense about jumping on sinking bits of wood just under the surface of the tar? Whyever would such a path exist to start with? (These segments, incidentally, were mostly in the episode about puzzly clues from a serial killer, a hopelessly outworn and implausible premise that I would be happy to encounter zero more times in all the rest of my life. I appreciated having a reason to visit some of L.A.'s quirkier landmarks, but there could have been other ways to do that. Besides, the psychopathic killer isn't a good fit for the rest of the story. Noir in general, and this game in particular, are about the banality and commonness of sin. Everyone's at least a little dirty. Unmotivated lunatic murderers belong to a different mythology of evil.) Towards the end of the game, I got into the mode of thinking of myself as the improvising organist -- and the game got better at allowing me to do that, or so it seemed. I was still regularly crashing my car when I didn't plan to, but it felt as though there were fewer faux challenges for the sake of gameplay in the abstract. Instead there were moments like Kelso struggling through the water in a flooding underground room -- not a difficult task, but one that captured interactively the sense of a lone man working against a pitiless physics. I just wish the whole game had been like that, and that the plot had engaged more deeply with the human themes earlier on. But my most powerful experiences with the game were not due to the intended interactions at all; and those colored my reaction to the rest even when I was most frustrated by broken triggers and odd tuning choices. Grandma Earlier this year I took a trip to L.A. to see my grandmother. There's not a lot she needs from anyone now except to listen to her stories. My grandfather and most of her close friends and siblings have passed on, so she's left to do all the remembering on her own. She's lived in the area a long time. She headed out to California in 1942 and got a job at a fire insurance company near Sixth and Spring, taking the bus downtown every day from Whittier. She worked her way up from a two-week typing gig to become an underwriter -- a rare job for a woman. She used to walk down the street to the bank to deposit fees collected from applicants, and there she met a teller recently back from the war. He struck up a flirtation. In time she quit her job and married him. In 1947 she and my grandfather moved into a new house in San Gabriel, part of a development with the rest of the houses still unfinished. Two weeks later my uncle was born. I grew up visiting that house, and she lives in it still. These days San Gabriel doesn't look that familiar to her. Most of the shops have signs in languages she doesn't read, and the major streets are built up beyond recognition. The orange groves are gone. She never goes downtown and she was appalled to hear that I had ridden the bus across Los Angeles to see her. But her own block still has the houses that were being constructed when she moved in. The lemon tree in her own back yard is still bearing fruit. Playing L.A. Noire had the quality of a pilgrimage to my grandmother's Los Angeles. The plot elements contributed. Cole Phelps' damaging memories of Okinawa, my grandfather's damaging memories of Germany. The Oklahoman boy's lost farm, the North Dakotan farm my grandmother left behind. But mostly it was a thousand details of setting. Those big sturdy cars, like the bucket-seated car my grandparents drive in the old home movies. The houses shaped like my grandmother's house. The insurance agency, the banks, the buses. Spring Street is meticulously reproduced -- oh, not to every detail, surely, but with many recognizable landmarks within a few blocks -- and I spent a good forty minutes just walking up and down the simulacrum of the block where my grandparents met, exploring the facades and peering into the windows. The moon set and the sun rose while I looked. Then again there were places I recognized from my own childhood: the Hollywood sign and the leaning palm trees and the way the streets look in the rare heavy rains; the tar pits that I loved as a kid; the LA Library, chiefly remembered for how upset my parents were when an arsonist set it on fire; the hobos, not so different from the Vietnam vets who lived under the freeways in the 80s. All those things from my experience projected backward to meet hers. Team Bondi obviously didn't set out to make a game that would help me connect my life to my grandmother's. But the possibility is there because the exacting observation of the setting put more truth into the story than perhaps even the creators knew. To the extent that art is a human response to the fact of time, L.A. Noire belongs to an august company. (Disclosure: I played a copy of this work that I purchased at full price. I have had no commercial affiliations with the publisher at the time of writing.) [Emily Short is an interactive fiction author and part of the team behind Inform 7, a language for IF creation. She also maintains a blog on interactive fiction and related topics. She also contracts for story and design work with game developers, and will disclose conflicts with story subjects if any exist. She can be reached at emshort AT mindspring DOT com.]