In Part 1 I discussed the success of recent graphic adventure games despite the genre being in its infancy. In Part 2, I’ll look at some of the challenges graphic adventure games have, and will continue to face when dealing with narrative interactivity.
It is an undeniable fact to me that Heavy Rain, The Walking Dead, The Wolf Among Us, Life is Strange, and Until Dawn are all excellent games that are well worth playing. It is also equally undeniable to me that the second playthrough of each of these games is inevitably disappointing since it reveals just how little of an impact the player’s choice actually has on the narrative. In other words, these games suffer from “choice illusion.”
This has been a complaint of critics pretty much since the graphic adventure genre got started in 2005. Read through a review of any of the above games and while they garner mostly praise, they will also be slammed with the criticism that nothing the player does really matters. Detailed breakdowns like this exist all over forums and video game review websites to explain how every choice is either meaningless, creates minor alterations which quickly return to a unified path, or only effect isolated plot components. The only exception to this rules seems to be endings since nearly every graphic adventure game has multiple possible conclusions, though the level of variation amongst the multiple endings tends to be low.
(It’s worth noting that the infamous Mass Effect 3 ending was also heavily criticized along the same line that the player’s input ultimately barely mattered to the game’s story. This problem is endemic to nearly all games which utilize narrative interactivity; graphic adventures just magnify the issue because narrative choice is such a core component of the genre.)
While I was never as bothered by choice illusion as most critics (especially after playing through so many graphic adventure games over the last month), I must admit that I was disappointed after replaying a few of these games for the first time. Nearly all them were advertised as presentations of coherent stories which could be controlled in a significant way by player input, and plenty of in-game choices and QTE strings would seem to suggest this is the case. But more often than not, both big and small player choices are meaningless or close to it in the long run.
For instance, in the two Walking Dead games:
- Choosing different dialogue options usually results in only a marginally different response, or no difference at all, for the vast majority of dialogue options in the games.
- Choosing different actions often somehow cause the same result – Clementine can choose to share her food with a dog, which causes him to attack her when she tries to stop him from eating it all, or she can choose not to share her food… which causes the dog to attack her because he’s hungry.
- Choosing different actions can also create short term divergences which then reconverge not long after – Clementine has to choose between helping Pete or Nick, both of whom are being attacked by zombies simultaneously. If she chooses Pete, then Nick runs away on his own, and Clementine helps Pete flee to an abandoned truck even though he has been bitten and will soon die and turn into a zombie. They stay there for a few hours until Pete accepts his fate and sacrifices his life to help Clementine escape so she can rejoin the larger group. If Clementine saves Nick instead, then Pete is just killed by zombies, and Clementine and Nick rejoin the group.
- Complex decisions which would imply long term effects have little to no tangible gameplay or narrative effect – To whom Lee distributes the remaining food rations within his group only effects how certain individuals treat Lee in the very next scene.
- The deaths of nearly every character who can die is predetermined, and cannot be altered even by choices presented immediately before their deaths.
Given how narrative choice is explicitly claimed by these graphic adventure games to be an integral component of their design, shouldn’t players expect their choices to truly matter in these games? Well, yes, of course players should expect that. But there is an open question as to what is a reasonable level of narrative choice, and even more interestingly, what is a reasonable level of “choice illusion”?
To understand why most modern graphic adventure games feature little real player choice and use a lot of choice illusion, we have to understand the costs and benefits of increasing a game’s level of narrative interactivity. The best way to conceive of this issue is to compare the process of scaling narrative interactivity to mechanical interactivity from a developer’s perspective.
One way a developer could increase mechanical complexity in an FPS is to create more weapons for the player to use. In turn, every additional incremental increase in weapon variety creates more work of varying time, intensity, and complexity for a developer which further scales in accordance to the complexity of the game as a whole. Each new weapon added in Doom meant creating new visual textures, designing new sounds, integrating the weapon into the level design (including placement, ammo distribution, enemy interaction, etc.), balancing against other weapons and enemies, and many more minute factors. Adding another weapon to a modern FPS includes all of those factors and more, and all at a higher level of complexity, given, for instance, having to balance a new weapon against many more existing weapons or integrating a gun’s effects into an enormously intricate physics system.
Now compare that mechanical interactivity scaling to narrative interactivity scaling. Each opportunity to alter the narrative within a modern graphic adventure game can have a variety of descending effects, ranging from as small as a single altered line of dialogue to as large as fundamentally altering the entire course of the ensuring narrative and game. On the simpler end of that spectrum, offering the player the opportunity for his player-character to respond to a prompt with multiple dialogue options means the game’s writers have to write more dialogue and the actors need to perform more lines (and/or motion cap acting). On the more complex end of the spectrum, a single initially small narrative alteration (like a character saying “yes” or “no”) could shift the entire plot in a way that means producing many hours of more writing and performing for branching narrative paths.
Imagine a graphic adventure game in which you choose one of five companions to accompany the player throughout the entire ten hour game. Assuming the companion plays a prominent role in the story, this choice will have enormous effects on the game’s content. If each companion has, say, an hour of dialogue, and five hours of screen time, then the developers will have to write five hours of dialogue, perform five hours of voice acting, and perform twenty five motion cap acting or animation design just to give the player one choice which in a single play through will display one fifth of the overall content produced. And that’s not even counting the secondary effects that the companion will have on every other character and various plot elements within the game, which will add even more content which needs to be created.
The issue at play here is that any degree of narrative interactivity above the most basic level seems to create layers of complexity which quickly make game design untenable, while mechanical interactivity typically doesn’t have the same problem to the same degree. I’ll admit that my lack of intimate knowledge of game design at a programming level is a hindrance for me on this point, but I would assume that increasing the level of mechanical complexity in a typical FPS takes less effort than making more branching narratives in a graphic adventure, if only because the game’s engine is doing most of the legwork for the FPS, while writers and performers are doing most of the work for the graphic adventure game.
The best evidence that increasing narrative complexity past a certain point becomes excessively costly for developers is the simple fact that the existing graphic adventure games really aren’t that complex. The text based adventure games of the past, and even some of their modern incarnations like Sunless Sea allow player choice to wildly alter, push, or even end entire plots, but modern games from Quantic Dream, Telltale, and other developers really don’t offer much in the way of long term player efficacy. I don’t think I can put an exact number on it, but I would guess that 85-90% of all of the games I listed at the beginning of the article play out identically regardless of what choices the player makes.
Even if a player can’t accept choice illusion, the good news is that developers seem to be adapting to the low narrative choice ceiling inherent to the genre. Graphic adventure developers are getting more innovative. While writing quality, the most important parts of the games, remains consistently high, developers are working on new ways to incorporate gameplay and narrative choice. They are expanding into new genres, tackling new subjects, and experimenting with new mechanics, many of which seem purposefully designed to accommodate or combat choice illusion.
For instance, Supermassive Games’s first outing, Until Dawn, brilliantly used a dramatic context that perfectly coincides with narrative choice. The game is broadly designed to mimic a generic horror movie by placing eight dumb, attractive teenagers in a cabin in the woods as they come under threat from mysterious entities in the darkness. Like in a horror movie, the characters are one-dimensional, not particularly likable, and make incredibly stupid decisions which inevitably land them in dangerous situations. While a set up like that in film would just lead to another mid-tier horror movie to get buried in Netflix, in the context of a graphic adventure game, it gives the developers a lot of opportunities to create meaningful player choice. Essentially, the player acts as a shepherd for his moronic flock by guiding the characters as best he can through their bad decision making.
The best part is that since the characters and plot are essentially expendable, the player has vastly more freedom to save or kill (accidentally or purposefully) the characters than in most other graphic adventure games. In games like The Walking Dead, Heavy Rain, or Life is Strange, most of the characters are too integral to the plot and strongly characterized to die at any moment without severely distorting the game’s story and themes. But Until Dawn’s cast doesn’t have that problem. Any of the eight characters can reach dawn and therefore survive the entire game, or die before the sun rises. Of course the game is not without any choice illusion; this in-depth guide can still reveals that all of the characters weren’t in quite as much danger as the game makes the player believe (ie. despite facing lots of danger two characters can only die in the final chapter). But even still, Until Dawn’s design provided the developers with an enhanced ability to create player efficacy.
It may be frustrating for players to realize that so much of the presented choices in these excellent games are based on illusions, but I think it’s just something which players and critics have to get used to. Or at least, it’s something they have to understand within in a proper context so we can adjust our expectations accordingly. Like it or not, choice illusion is a vital component of modern graphic adventure games. The cost of scaling up the efficacy of narrative interactivity is so high that choice complexity is necessarily going to have a low ceiling for the time being. Maybe with advancements in animation, costs can be brought down a bit, but with the bulk of costs presumably coming from writers, voice actors, and motion capture, it’s unlikely that we will see a significant expansion of narrative control in video games any time soon.
But that’s not as big of a deal as it may initially seem. I think the implied (or phony) importance of player choice in the good graphic adventure games I’ve mentioned has been overstated, not just in terms of how much player choice the game erroneously presents, but in how much value it provides to the games as a whole. The stories within these games and their sense of immersion created by even a low level of narrative interactivity are what really make them shine. The Walking Dead and Life is Strange are great games because of their characters. The Wolf Among Us is good because of the atmosphere and art design. Heavy Rain is good because of its excitement. Until Dawn is excellent because of its cleverness. The illusion of choice may have amplified all of these games for their players, but fundamentally I think they maintain nearly all of their quality once the player realizes just how much choice illusion is involved in each of them.
The only real casualty of choice illusion is that sense of “anything can happen” carried by the first couple of graphic adventure games. The early Quantic Dream and Telltale games were marketed heavily on this basis. Quantic Dream claimed that any of the four playable characters in Heavy Rain could die, and just as characters were constantly making life and death decisions on the Walking Dead TV show, players thought they could make the same exciting decisions in the game. But now everyone knows that only two of the four Heavy Rain characters can die (the other two are too integral to the plot to disappear) and nearly every Walking Dead death is predetermined. And by this point, everyone knows that the inevitable “anything can happen” claims of future games are greatly exaggerated.
This means that graphic adventure games went through an early honeymoon phase where they built, and, based on a lot of review scores, met high expectations. Then they went into a period of lower expectations and moderate backlash due to their use of choice illusion, which they are still in now. I’m hoping that soon graphic adventure games will come out of the perceptual slump and land somewhere between the two extremes, where they can be appreciated for their strengths and weaknesses accurately.