Is there anything that better exemplifies the inherent weirdness of video games than Rockstar Games and Rockstar’s games? The company develops and publishes some of the medium’s most commercially-successful and critically-acclaimed titles – Grand Theft Auto V, the latest instalment in their flagship series, has a rating of 97 on Metacritic and made a billion dollars in 72 hours. But the company also develops and publishes the medium’s most defiantly puerile and morally-questionable titles. They are, in fact, the same titles. The aforementioned GTA V, to give one example, is at times a daring satirical masterpiece and at others legitimately offensive and inexplicably stupid. That a game can be both of these things is not entirely surprising. That it can be both so frequently and interchangeably very much is. Just how the open-world sandboxes within which Rockstar scatter their toys seek to both plumb the depths and scale the heights of American culture’s past, present and future, so too do the games that house them contain the best and worst of what the medium has to offer.
Many people of my acquaintance insist that this is intentional. That it must be. That no developer capable of such occasionally startling prescience can also be so short-sighted that they’re unable to recognise which aspects of their work are meaningless or insulting. And this may very well be true. Nobody can say for certain. It’s certainly in-keeping with Rockstar’s flagrantly cynical view of games and the people who play them; of life and the people who live it. But it strikes me as an odd way to craft an experience. Strange also is Rockstar’s approach to storytelling; their casual insistence on imbuing their stories with heart and depth and vitality, and on continually undermining those stories at every given opportunity. This, I think, neatly nutshells exactly what is so confusing and fascinating about L.A. Noire. It is a game of constant contradictions, wildly incompatible ideas, remarkable successes and mind-boggling failures. In the years since I first played it I have thought about it a great deal, and I still couldn’t say with any certainty whether or not I actually like it.
L.A. Noire is not, technically, a Rockstar game. It was published by them, but the actual development was handled by an Australian studio, Team Bondi, who have since dissolved – notably amid rumours that their employees were worked half to death during the development of this very title. Nevertheless, L.A. Noire’s genetic provenance is obvious. It might have been conceived with a mistress, but there’s no way it would fool a paternity test. And while some of its problems are unique abnormalities, most are hereditary. It’s a shame. L.A. Noire’s single greatest failing is how desperate it is to get on with the rest of the family.
Take, for instance, L.A. Noire’s desire to tell a coherent, linear story in an open, non-linear world. One is fundamentally detrimental to the other. This is a problem unique to video games, but not unique among them. It is exceedingly difficult to tell a story that allows for any audience participation at all; telling one which offers the audience a large amount of behavioural freedom is arguably impossible. Rockstar’s games have always been this way, but they made concessions. Their player-controlled characters were antiheroes theoretically capable of nutcase rampages. They were criminals, and the law was their enemy, and the player understood that. Of course, the player’s and the story’s definitions of these characters were often at odds. Niko Bellic, the protagonist of Grand Theft Auto IV, was a hair-trigger European immigrant who the game’s story insisted was tortured and ambivalent. And while the cut-scenes were playing, he was. Yet on the way back to his apartment Niko would plow over throngs of pedestrians and send their ragdoll corpses skidding along the asphalt. Often, he would pull over to collect the floating stacks of ghostly currency they left behind. Whenever the story paused, Niko became a container for the player’s manic excesses. When the story resumed, he became a character again.
Niko’s career criminality was enough justification for his psychopathic outbursts that nobody really thought about it too much. L.A. Noire has no such advantage. Its story concerns a post-war detective named Cole Phelps, who, in 1947 Los Angeles, is speedily promoted through the constabulary ranks and tasked with investigating crimes in multiple divisions, from Traffic to Vice to Homicide and, eventually, through a series of extraordinarily odd events, Arson. By definition Cole is opposed to how most players will want to play this game. He can’t assassinate innocent bystanders or solicit prostitutes. He can steal cars, technically, but he does take the time to show his badge first. Once he has stolen a car, he can either find another car, drive past an iconic L.A. landmark, or simply head to his next story-mandated destination.
Many video games, perhaps most, implement systems that are designed to limit how easily the player can euthanize the surrounding fiction. In open-world games the potential for player disruption is far greater, because opportunities for mayhem are more frequent and accessible. Grand Theft Auto’s lawlessness is a convenient catch-all solution to this problem. Because Cole Phelps is ostensibly a peacekeeper, L.A. Noire instead elects to hamstring its world. There are no fun or interesting or silly diversions. There can’t be. A lawman can’t draw his gun and indiscriminately mow down the populace. He can’t launch a stolen car from a ramp and barrel-roll it over a bridge. He can’t turn up to a crime scene on a BMX or leave a murder suspect in an interrogation room while he goes to play pool and lift weights. In Grand Theft Auto: San Andreas, the player could hijack a fire truck and use it to extinguish an arbitrary amount of fires, thus making the player-character impervious to explosions. That is fundamentally ridiculous, but it’s expected of an open-world game. It’s the point of an open-world game. Nobody gets in a toybox just to sit there.
Don’t get me wrong: L.A. Noire’s City of Angels is often astounding, both in its vastidity and its superlative realization. To this day, very few other video games have been as painstakingly rendered or as devoted to digital landscaping. I was often content to stare at various storefronts, museums and monuments for minutes at a time. L.A. Noire accommodates the idle observer remarkably well. But the game wants you to stop there. Quiet admiration is enough. Should you poke and prod at the game’s systems, you begin to see how limited your available interactions are. The player has no scope for influence, nor can they deviate from the tracks on which the game sets them. A minor act of rebellion such as, say, walking to your next destination rather than driving, accomplishes nothing. You will reappear in your vehicle, alongside your current partner, right where the game wants you to be. Not only does L.A. Noire bind you in place, it tacitly refuses to acknowledge any attempts you make to break free.
For years now, I’ve written about video games. I’ve often written about their stories. Sometimes I feel as if the medium is incompatible with authored storytelling – usually after playing a game with a bad story. Sometimes I feel like good fiction will invariably compel a player to behave in a way that the fiction’s author intended – usually after playing a game with a good story. I view video games primarily as a storytelling medium; a relatively young and experimental one, but a viable one nonetheless. As that kind of person, I have no desire to make Cole Phelps lose his mind and slaughter half of L.A. I don’t think he should be able to. A close friend of mine, who is annoyingly smart and surprisingly thoughtful, was furious that he couldn’t. This is the dilemma that game-makers face. The rules of the medium are so nebulous and often contradictory that two smart, thoughtful players can approach the same game with radically different expectations of what it should be, and what it should allow the player to do. In the case of L.A. Noire, a game that focuses primarily on narrative, problem-solving and dialogue, I see no justification for an open-world. But the open-world is there. Freedom is dangled tantalisingly in the face of the player, but when they reach out for it, the game snatches it away. That L.A. Noire limits the player’s capacity for interacting with its world in order to preserve an authored narrative strikes me as counter-intuitive. I don’t want Cole Phelps to be able to do these things; I want the game to not care that he can’t.
Stopping to admire newly-discovered landmarks is a fine thing, as is finding a rare, hidden car to drive around (although you can’t keep any). Yet these are activities very few players would eagerly pursue, and they’re present out of obligation. But an obligation to whom? Nobody can accurately predict the wants of an audience as diverse and insufferably picky as those who play video games. The desire of game-makers is to make enough money to pay their bills and put food in their children’s mouths, and that frequently manifests as games which are as broadly accessible and anodyne as possible. Much of L.A. Noire feels like this. The needless open-world is a symptom of it, as are these asinine collectible diversions. Even worse are what the game calls “Street Crimes”, which are essentially capsules within which Team Bondi stuffed the game’s worst mechanics: Wonky cover-based third-person shooting, dreary on-foot and vehicular chases, appalling fisticuffs. None of this stuff needed to be here. In the context of the story these mechanics at least make sense. They’re still lacklustre, but they have a purpose beyond merely justifying the open-world. There was a rumour that circulated around the time of the game’s release that these more “traditional” elements were forced upon the game at the last minute by Rockstar, who were worried that without them L.A. Noire would be seen as too boring, or, perhaps more charitably, too niche. How ironic, then, that the elements inserted haphazardly to stave off boredom are easily the game’s most boring elements. Some enterprising designer at Team Bondi obviously recognised this, and inserted the option to skip them.
L.A. Noire excels when it shrugs free of the shackles of a traditional sandbox, which to the game’s credit is quite often. You can compare the experience to other open-world games if you wish, particularly those developed by Rockstar, and many people did. (I’ve heard the game flippantly described as Red Dead Detective an awful lot.) But L.A. Noire has much less in common with these progenitors than it does John Alton and Raymond Chandler; Chinatown and L.A. Confidential; even old adventure games like The Secret of Monkey Island. It brazenly pilfers from its influences – a chunk of plot from Who Framed Roger Rabbit?, an entire case based on Robert Wise’s The Set-Up. As a result, the game’s story can occasionally feel second-hand. That isn’t to say the game’s story isn’t good; it’s frequently very good, and sometimes extremely good. L.A. Noire is also utterly committed to that story, even when it falters. It aspires, plainly, to a classier version of video game narrative, one that can be called good without the need for any qualifiers. And it achieves that.
How, I think, L.A. Noire achieves that, is by making its principle mechanics delivery mechanisms for its story. In many games, story is what happens between stretches of gameplay. In this one, the gameplay is the story. At the beginning of each case, something happens. That something can be a murder, a car crash, arson, or whatever. It doesn’t matter. The player is usually given a glimpse of the crime being committed and a summary from Phelps’s commanding officer, but very little information otherwise. And then the player takes over. They drive to the crime scene (either themselves or by telling their partner to do it, further defeating the purpose of L.A. Noire’s open-world) and start looking around. Each clue they unearth, each lead they scare up, is not a MacGuffin they carry to the next cut-scene – it itself is the story, fragmented and scattered, but there to be assembled by the player.
How the player navigates this story is by alternately investigating crime scenes and interrogating suspects or witnesses. In practice both mechanics are flawed (the latter quite significantly), but both are praiseworthy in that they’re non-traditional approaches to gameplay that emphasise storytelling and problem solving. Investigation is, in essence, a contemporised flavour of old-school point-and-click adventure game, wherein the objective is to find something useful amid piles of rubbish. Written down, this sounds awful. But there’s a peculiar satisfaction in discovering a piece of incriminating evidence, or even a seemingly mundane item that somehow knits together a theory. As an experience, L.A. Noire values the journey, not the destination. Acing one of its cases is like taking a trip to Slough by way of Atlantis.
L.A. Noire wisely eschews the infuriating moon logic that characterised classic adventure titles of the 90s. Instead, every examinable object prompts a definitive yay or nay from Cole; there’s no need to rub an item against every other item in Cole’s pockets in order to determine its usefulness. The process is, mercifully, free of trial-and-error. That isn’t to say that all objects are clues. Most of the junk Cole picks up is exactly that – junk. What L.A. Noire classifies as junk isn’t always discernible, but that’s a forgivable consequence of the rules that govern inspection. Forcing the player to examine every beer bottle and cigarette butt in a given environment would be absurd, but having them inspect only important clues wouldn’t be much better. I can accept a few lapses in a game’s internal logic if it makes for a smoother, less frustrating experience.
And, really, there’s very little about the process of investigating crime scenes that could warrant frustration, especially not how L.A. Noire presents them, which is the same way it presents everything – with astoundingly authentic period detail. Homes and police stations and movie studios in this game feel exactly how homes and police stations and movie studios feel in the real world. The clock has been wound back to a time before most of us were alive, but the essential nature of these spaces is familiar. Evidence is found in logical places: A glasses case in a bedside nightstand; muddy boots kicked off beside a back door; a flyer dropped on a kitchen table; a bloodstained length of piping crudely hidden in an exterior drain. This devotion to forensic authenticity is vital to the L.A. Noire experience, because it distracts from the magical allowances games are forced to make for the sake of player convenience. Cole’s health regenerates, his car repairs itself, and dispatch – which he calls whenever he needs an address or a suspect’s criminal records – always immediately has the answers to his queries. This is silly, of course, but it doesn’t violate the unspoken contract between game and player that a certain amount of silliness can be tolerated. No other medium asks its audience to sign on this dotted line, which is why so many people can only see video games as interactive systems, and not as vehicles for storytelling. Interactivity, by its very nature, sabotages traditional storytelling. This is inarguably true.
Traditional storytelling, then, is non-interactive storytelling. Fine. As such, many video games co-opt cinematic language for their narrative purposes; L.A. Noire is fluent in such language, but it is one of the few titles to gamify its grammar. What would in film be passive scenes of dialogue are, in this game, a form of non-traditional, interactive storytelling that I haven’t seen before or since. Reading and interpreting the human face is an important gameplay mechanic. L.A. Noire’s characters are not crude digital marionettes, but actors transplanted into three dimensions with algorithmic accuracy. The technology that allows for this, MotionScan, invented by Depth Analysis, a subsidiary of Team Bondi, captured the performers using 32 cameras positioned at every conceivable angle. Unlike previous facial animation that tracked only quadrants of the human face and extrapolated everything between, MotionScan records every dimpled chin, narrowed eye and other actorly flourish that had previously been impossible for video games to capture.
The result is scarily lifelike performances that, at their best, could potentially change the way video games employ actors, and the way players consider video game storytelling. This is the aspect of L.A. Noire that most fascinates me. It applies tools and templates that the current orthodoxies of game design tell us do not work, and, for the most part, makes them work. These are the game’s interrogations, in which Cole sits down with a suspect or a witness and must use his knowledge of the case, collected evidence, and, most importantly, their behaviour, to determine whether they are being truthful, evasive, or outright dishonest. And when this works, it works startlingly well. There is tremendous pleasure to be found in identifying an obvious discrepancy in a suspect’s alibi, recognising their dry swallows and evasive eyes, and then producing the perfect piece of evidence to contradict their statement. In moments like this it becomes clear what L.A. Noire is trying to accomplish, and also that whichever game manages to do perfectly what this one merely does well might represent the most significant narrowing of the space between story and storyteller that video games have seen in their history.
L.A. Noire is not this game, for various reasons, most of them relating to the underlying mechanics of these interrogations, which often seek to undermine or distract from everything they do well. The first problem is ambiguity. Here is how these sequences work: Cole sits down with someone who presumably possesses information he needs, he asks them a series of questions, and they provide a series of answers. Cole’s response to these answers can be one of three things: he can acknowledge the statement as truth, he can call it into doubt, or he can accuse his interlocutor of lying. Here is where problems begin to emerge. Often in these situations Phelps will interpret a mild world like “doubt” as an excuse to launch a scathing verbal tirade. Other times he’ll raise a single suspicious eyebrow. There’s no real way of predicting Phelps’s behaviour, and his wildest loose-cannon tendencies seem to be reserved for the least-deserving suspects – a hospitalised teenage girl, say, or a recently-widowed husband. This can be immensely frustrating, in part because it’s a ludicrous way for someone to behave, but more so because it encourages the player to approach these situations based on how Phelps will interpret them, not on their own gut instinct.
It’s tempting to blame the characterisation of Phelps here, and initially, when I was being tormented by this, I did. But in retrospect that was a mistake. Phelps is an arsehole, but not as a by-product of sloppy writing. As the game progresses, it becomes clear that Phelps being an arsehole is a fundamental aspect of his character, and a large part of what makes him darkly compelling. From a storytelling standpoint, this is perfectly okay. But when the player-character’s personality directly impacts an important game mechanic, it is not. As a result, many of the game’s interrogations begin to feel like exercises in damage control, as the player attempts to pre-empt Phelps’s ballistic outbursts and mitigate any damage they might cause. This is not necessarily uninteresting, but it also not, I suspect, what the designers intended these sequences to be.
This is not to say that odd, emotionally-inconsistent writing does not also plague L.A. Noire’s interrogations. This is a slightly more pardonable offense, given that the branching possibilities of dialogue in a game like this effectively prohibit genuine coherence. But the dialogue is so purposeful that it rarely – if ever – feels naturalistic. L.A. Noire’s story is predetermined; the in-game evidence offers the potential for choice and Holmesian deductive reasoning, and the conclusions one draws from interrogations can alter the length of individual cases and the order of the leads pursued within them, but the trail cannot go cold. As such, every scrap of dialogue must contain some kind of actionable material, so if the player – or, more likely, Phelps himself – messes the case up completely, it will still contrive ways to continue. Despite this, and for reasons entirely mysterious to me, the game still elects to tell you whether or not you chose the “correct” answers during each interrogation. What should be the most organic element of the game frequently feels depressingly synthetic.
Still more bizarre is how the game’s interrogations communicate with its evidence-gathering mechanics. The relationship should be symbiotic, and in the early portions of the game, when evidence is scarcer and the characters are more inclined to telegraph their deception, it is. I initially thought that this itself would be a problem: when you ask an actor, who is acting and therefore already lying, to act as if they’re lying, what you generally get is a bald pantomime. This, it turns out, was not the problem, partly because later suspects become much more prone to subtlety, but mostly because the real problem is in how L.A. Noire frames your available responses. The game never explicitly tells you this, but the correct time to doubt an individual is not when you are unsure of their honesty, but when they tell a blatant lie which you lack the evidence to prove. This is a strange decision. The player’s natural inclination when presented with bullshit will be to call it out, but unless you possess a very specific piece of evidence that is tied to that specific slice of bullshit, the game regards this as incorrect. Intuiting which item you’re supposed to be using is only half the battle; knowing whether or not you missed that item during your investigation, or if it even exists, is impossible.
One has to wonder, while playing L.A. Noire, that if these problems are avoidable, then how, exactly, would they be avoided? You could argue that this kind of character-driven narrative has no place in an open-world game, and I would agree with you, but what better way to express a faithfully-recreated 1947 Los Angeles than an open-world that you can drive around in and explore? You could argue that freer dialogue would improve the character interactions, and I’d agree with you there, too, but what place does that kind of dialogue have in a game which must consistently give players important motivational or directional cues? How would L.A. Noire account for the player’s fallibility if it prohibited incorrect answers; if it didn’t have the systems in place to course-correct Phelps’s investigation? These are undeniably problems, but to remove them would create more problems. Games like L.A. Noire are what make me appreciate being a gamer and a writer. But they’re also what make me thankful I don’t make games.
Towards the end of L.A. Noire, its story gets better and weirder. The gumshoeing gets more involved. At the very end, the player runs around sewer tunnels fighting gangsters with a flamethrower. And this is my problem. This is why I can’t decide exactly where I stand on the game; whether I regard it as simply encouraging, or as the closest step a digital experience has yet taken in showing us what video game storytelling is capable of. Sometimes it’s both. Occasionally, it’s neither. In 4000 words, I still haven’t been able to adequately express whether I think L.A. Noire is good or not. But maybe that isn’t the point. “Good” is a relative term, anyway. It means different things to different people, and to some it means nothing at all. I think the kindest, fairest thing I can say about L.A. Noire is this: I’m very happy that it exists.