SHAPE UP!: Repetition and Difference

Ian here—

The sixth episode of the ongoing Shape Up! series is posted. This one addresses a structural innovation present in Oxenfree and NieR: Automata—two old favorites of mine I’ve mentioned before in videos and/or posts—as well as Doki Doki Literature Club! and Kitty Horrorshow’s ANATOMY, which I haven’t written about before but am excited to do so now.

Full script below the jump.

Hello everyone, and welcome to another episode of Shape Up!—an ongoing series about form, structure, and pacing in games. This video will be all about the experience of re-playing games—certain games, in particular, but I’m going to start out with some basic table-setting. What I’m about to say will be a mix of things both commonplace and contentious, so to help smooth everything over I’m going to state it all as a series of postulates. I know it’s a bit pretentious—but I’m going to play the role of philosopher here.

Proposition A: No game is ever the same the second time you play it.

This is most obviously the case in games with a lot of randomization. Each time you play poker, everyone is going to have a different hand. But it is also the case in games with no randomization. Chess has no randomization. And yet each time you play it you learn more about the game, and potentially more about your opponent. You aren’t the same player, and so the game won’t be the same. Gone Home has no randomization, and yet each time you play it your experience will be different.

This last point is contentious, I realize. Many people have charged Gone Home with not being a game at all, and the evidence most frequently cited is its “un-replayability.” Gone Home, the argument goes, lacks both challenge and choice, so your experience doesn’t change on subsequent playthroughs, and therefore it is not a game.

But this just isn’t true. If you do play Gone Home a second time—and I have—you won’t have the same experience as before. You’ll remember things from the first time you played it. Maybe you’ll remember the layout of the mansion, or the character’s relationships. You’ll probably remember that its feints at being a horror story are only feints, that it is a family melodrama at heart, and the red is only hair dye and not blood. Now that you know that there’s nothing to fear in the mansion, you might go through it more quickly, making a bee-line to the key items, speed-running the thing. Or you might go through more slowly, luxuriously soaking in detail that you missed the first time around, because you were afraid of the basement, or whatever.

And you might respond that none of what I’ve just said is unique to Gone Home. No subsequent experience of any work will ever match your first experience—that’s a banal observation. And yes: it is. But I’m still going to list it as Proposition B. There is no choice or challenge involved if I re-read a novel, or re-watch a movie. But as long as I remember the first time I did those things, the experience will necessarily be different. Maybe I know who the killer is now in a mystery, so can no longer savor the pleasure of the puzzle. But there’s a new pleasure, because now I can pay more attention to how the narration artfully hides which clues are most pertinent. And not just mysteries—with every story, now that you know where it’s going, you can focus more on aspects of craft, including how future developments are foreshadowed. [“Oh, I never thought I’d miss a hand so much!”] This observation that there’s something unique about your first experience of a work is in no way groundbreaking, but it is useful to state outright, as a sort of baseline.

And it leads us into Proposition C: There is a meaningful difference between a work that changes because you remember it, and a work that changes because it remembers you—a work that (in some admittedly anthropomorphized sense) knows that you’ve experienced it before, knows what your memories and expectations are, and then zags when you expected it to zig.

For much of human history, this sort of on-the-fly permutation was limited to live performance: in-person storytelling, live theatre, stand-up comedy. Once you record something, in a book, or as recorded audio, or a film—this aspect of it dies. A musician can upset audience expectations during a live performance. [“I’m sorry ladies and gentlemen—there’s no reason to do this song here.”] But if you record that performance, the recording can’t continually surprise you in the same way. It’s set now.

But digital media brings new possibilities.

Let me ask you this: could The Stanley Parable exist as a Choose Your Own Adventure book, printed on actual paper? The answer to this, I think, is “sort of.” You could have a book where the narrator talks about Stanley, and says that Stanley goes through the left door, and then at the bottom it says turn to page 37 to go through the left door, but you can also turn to page 64 to go to through the right door. And the book would continue from there.

I think this Choose Your Own Adventure would be basically functional. But you would lose things, because The Stanley Parable changes in little ways as you play it. There are the basic quality-of-life features. If you’re doing a lot of playthroughs in a single session without quitting to the menu, the game will auto-skip the opening cutscene. And it will also spawn you in a slightly different position, poised in the doorway ready to head out, rather than at Stanley’s desk facing his computer screen. But beyond these, the more you play, the more likely it is that you’ll see other, more unexpected differences. The computer monitors occasionally display a version of the “end is never the end” loading screen. The path to the first two doors can sometimes be longer—once when I was playing there was an extra blue hallway inserted, that I had never seen before and have never seen since. Other times, the path is much shorter: just a single hallway from your office, and there the doors are. There’s also a small percentage chance you’ll spawn in a completely different office. The narrator’s dialogue also changes. Sometimes he’ll talk about how much Stanley longs to experience weather. [“A soft wind blew outside and perhaps rain started, and if it did it stopped shortly after. Stanley hoped that he would one day see weather.”] There was one P.T.-esque moment where the narrator out of nowhere told me there was someone following me. [“Someone was following Stanley. He was sure of it. If he checked over his shoulder now, he would surely catch them. It was only a matter of time.”]

There’s no gameplay function to these things that I’m aware of—with just one exception, which is that you need the blue office variant to get the Whiteboard Ending. But as for all the other changes, there’s only an aesthetic function. The game is saying to you, “Welcome back. Your user profile data indicates you’ve played this before. But things can still change. Stay on your toes.” It creates a feeling of mutuality, that the game is trolling you just as much as you’re trolling it. A paperback novel can’t do this. It might be “interactive” in the sense that you make choices while reading that affect on your route through the story, but it can’t surprise you by mutating in cheeky ways. The Stanley Parable’s “confusion ending” nests four different restart loading screens into what is in fact a single run of the game, allowing the narrator to comment on how these apparent restarts have failed to cleanly re-set the accumulated problems of previous playthroughs. [“No, I restarted—I swear! I definitely re-started the game over, completely fresh! Everything should be … or did something change? Stanley?”] A paperback could never do this. If you turn to page 1 to start from the beginning again, you’re at page 1. A paper book can’t trick you into thinking you’ve returned to page 1 when you in fact haven’t. The evidence is right there in your hands, and it’s pretty conclusive.

This potential for mutating and trolling the player really comes down to save states. We are now ready for Proposition D: the humble save file is an underrated asset in the toolbox of game storytelling. In particular, I’m thinking of possibilities of the New Game+: the utilization of user profile data that persists between playthroughs, allowing players to play the game over again, now modified in some way.

For all that “replay value” is often held up as the key criterion by which games should be judged, it’s surprising to me how few games have used the New Game+ in a narratively interesting way. A story that, in effect, knows it’s been told before, to a given audience … that’s something that allows writers to play some truly devious tricks around repetition and expectations. This is fertile ground for some of the most fascinating bits of medium-specific storytelling that games have to offer, and I’m always delighted to see it done well.

In fact, I’m so delighted that I had a hard time cutting down my case studies for this video, and as a result I’m going to be looking at a broad swath of examples. I’ll be discussing them in terms of two categories: the “repeat customer appreciation” scenario, and the “should have left well enough alone” scenario.

In the “repeat customer appreciation” scenario, your persistence with the game is rewarded in some way. Think of this as the equivalent of a customer loyalty punchcard. Except instead of “your tenth sandwich is free,” it’s that a game withholds a good ending unless you’re willing to do a second or third playthrough.

My main case study here is Doki Doki Literature Club, and here I’m going to repeat a warning from my previous video: all of the case studies I’ll be looking at in this video have surprises in store for players as they navigate their off-ramps, and I’ll be massively spoiling the experience of those surprises here. If you’re spoiler-averse, might want to sit this whole video out.

Also, I’m going to give a content warning for Doki Doki Literature Club that is arguably a bit of a spoiler in and of itself, but I think it’s warranted. I’ve covered plenty of horror games on this channel and showed a whole lot of digital gore, but Doki Doki Literature Club features emotional abuse and depictions of suicide in a way that feels different, to me.

Anyway, let’s hit New Game. At first glance, Doki Doki Literature Club looks for all the world like a bog-standard dating sim set in a Japanese high school. You have your girl-next-door, Sayori, and you decide to join her school club, the Literature Club. The club has three other members, all girls—there’s Natsuki, a tsundere-type who’s into cutesy Manga, and Yuri, an intellectual introvert who reads horror novels. Then there’s Monika, the cheery and assertive club president. The main activity in the club is writing and sharing poems with each other, and this is your main way of wooing the girls. Each poem is scored in terms of which keywords a certain girl likes. As a somewhat morbid intellectual, Yuri likes multisyllabic words, and also things that seem related to horror literature. Natsuki likes words that are bright, comfy, cheery, and stereotypically cute. Sayori’s trickier—she likes a split of words connected with light emotions, and connected with dark emotions. She in fact characterizes her own tastes as “bittersweet.” It’s hard to get that eclectic balance right, so it does seem like the game is funneling you toward dating either Yuri or Natsuki, whose tastes are easier to consistently play to. And that’s fine. Maybe pursuing Sayori is the “hard” route, best left for a second playthrough.

But the weird thing is that doesn’t seem to be any way to romance Monika at all. Her illustration isn’t even on the poem-composing screen. I guess her route was cut content, or something.

Anyway, the school festival is approaching, and to prepare for it you’re going to end up doing some activity with whichever girl you’re closest to, who is most likely going to be either Yuri or Natsuki.

Meanwhile, Sayori becomes withdrawn, and accuses us of not needing her in our life anymore because we’re pursuing another girl. As the game goes on, her dialogue indicates she’s sinking further into a depressive episode, and that sucks. It’s one thing for the game to make her harder to date, but to then bundle that with this emotionally complicated aspect of our relationship is really … intense. But, maybe we’ll try to date her during our second playthrough. This game’s not long. And maybe we can look up a walkthrough of which words she likes next time around.

The day of the festival arrives, which probably means our hard work wooing is going to pay off. Except Sayori hasn’t shown up. Since she’s our neighbor, we volunteer to go check on her at home … and we find that she’s hanged herself. This is … not just a bad ending, this is unlike anything that’s ever been in a dating sim ever before. Did we do something wrong? Was there a way to avoid this? Our player-character collapses into a fit of guilt and confusion, the game ends, and we’re back at the start screen.

Which looks a little different now—the image of Sayori is gone, and the characters and text are glitching out, as well. But aside from this change, the game doesn’t explicitly indicate that there’s a new version of this story waiting for us if we hit start again. This is pretty much a full-on false off-ramp. Theoretically there could exist a player out there, somewhere, who reaches this false ending, mutters “that’s fucked up,” and closes the game, never to boot it up again.

And … that’s pretty cool. A book can’t have a false ending in this way, because you can always see how many pages are left. Post-credits sequences in movies divide audiences between those who leave and those who stick around, but that trick is only really possible in a theatrical setting—it’s easy to spot the extra hidden scene when viewing at home. So with all due apologies to the theoretical naïve player I’ve postulated who never sees the rest of Doki Doki Literature Club, I do think it’s exciting that digital games are a medium where false endings can be unusually effective.

If you do start a second playthrough of the game, Sayori’s character has been deleted entirely—and now things go in a dark direction much more quickly. Natsuki and Yuri’s characters repeatedly glitch out in ways that make them humiliate themselves in front of us. And eventually it becomes clear that the party responsible is … Monika. Monika has become self-aware of the fact she’s in a dating sim, and she’s sick of being an undatable character. So she undermined the other characters by ratcheting up their anxieties, then deleting their data—first Sayori, then the others. While monologuing, Monika reveals where the character data is kept, and at this point you can force-quite the game and delete her character file. The Plus! edition of the game, released in 2021, has you do this on a virtual desktop, which—you know, I get it. They wanted to release this on consoles. But I much prefer the original 2017 freeware version of the game, where you had to actually dig through your own hard drive to do this. I’m always tickled when a game acknowledges its own status as software and makes you interact with it in a scavenger-hunt type way like this. It also just makes the game creepier. There’s that brief flicker of a moment where you have to ask yourself “wait, is my computer haunted?”

Once you delete Monika, her portrait is removed from the main menu. And if you start a new game again, Monika is gone, and every other character has been restored.

Doki Doki Literature Club is a disturbing game, but I think it fits the “repeat customer appreciation” mold. Each successive playthrough gets you closer to a satisfying off-ramp. The first playthrough ends tragically, but once you get past the false ending, it’s revealed to be just the end of Act I within the game’s larger narrative structure. In the second playthrough, the antagonist reveals herself, you learn her weakness, and can ultimately defeat her. Then your third playthrough is a dénoument—not an entirely happy one, but one that does reward your persistence. That simple act of hitting the “new game” button again after the first false ending was a sort of answering of the call to adventure, putting you on the path of vanquishing the villain.

But not every game portrays the act of replaying it as heroic. We also have the “should have left well enough alone” scenario, where curiosity doesn’t pay off, and the player is in some way admonished for their foolhardy persistence.

A good, short example of this is Kittty Horrorshow’s ANATOMY. ANATOMY places the player in an empty house, tasked with collecting individual cassette tapes from different rooms and placing them in a cassette player located in the kitchen. Together, the tapes constitute a short essay analogizing a house’s rooms to the features of a human body. [“If we were to dissect a house as we might a human cadaver, we would find ourselves able to isolate and describe its various appendages and their functions in a decidedly anatomical function.”] There’s something calmly meditative about venturing out to collect each tape, then bringing it back to that central location to listen to it. Or at least, there would be, if the house wasn’t pitch-black. In fact, the game is fucking terrifying, and it’s really impressive just how scary it manages to be given its obviously constrained resources. There are no monsters, no musical score. Nothing ever jumps out at you and says boo. There’s just a constant sense of possible threat as you’re blankly informed, again and again, that in order to collect the next tape you must move from the relative safety of the kitchen out into another dark, unexplored room. But nothing ever happens, not really. When you head down into the basement to collect a tape, it’s hard to see, but the door doesn’t shut and lock behind you or anything. It’s not until you collect the last tape that the game finally plays a mean trick and traps you—not in the basement, but in the main upstairs bedroom. Upon collecting the tape, the bedroom’s door just straight-up disappears, and you need to use a new tape player that has spawned within the room itself to listen to the last tape. But that’s the only trick. You listen to the last tape and the game quits itself, sending you back to your desktop. It’s rather anticlimactic, really. I mean … that can’t be it, can it? What would happen if we launched the game again? Might was well …

Your second playthrough is akin to a lossy dub of the first playthrough. The tapes are more distorted now, to the point where the monologue is impossible to decipher. The text instructions telling you where to go have typos in them. Even the geography of the house has started to glitch out a little, with flickering textures and doors that only lead to other doors. The final tape you find is disturbing, telling a story of a young man covered in welt-sized ticks exploring a house, much to the anger of the story’s narrator. Then, the basement door swings open. If you enter it, the game quits to desktop again.

I mean at this point, it’s almost like it’s daring us to boot it up again, right? In for a penny, in for a pound. During your third playthrough, the geometry of the house is distorted even further. The tapes this time contain nothing but noise, and the game gives you no commands about where to go next—you just have to revisit rooms and see what you find. Finally, the game traps you in the kitchen with nowhere else to go but the basement, and if you head down there you’re greeted with a concluding burst of narration, announcing that our intrusion within the greater organism of the house has been felt, and now we will be punished for our insistent harassment. As the narrator admonishes us, teeth emerge from the floor and the ceiling, as if a mouth is swallowing us. Following this, the game quits to desktop again.

If you launch the game again, you’ll find yourself unable to move. The game simply presents a static view of the tape player as it plays one last monologue about an abandoned house that is a conscious entity, lonely, bitter, and increasingly hungry. [“While a house may hunger, it cannot starve.”] The tape ends, and this time, for once, the game doesn’t quit itself—instead, it will stay on that screen forever, until you force quit it by holding down the escape key. If you ever boot it up again, you’ll remain trapped and immobile, still staring at this tape recorder. The game is intransigent and uncooperative now, and the only way to make it playable again is to delete your save. We’ve navigated through the false endings and reached the true ending, but it doesn’t feel like a happy one. Certainly no one’s congratulating us for making it this far. It’s more a feeling that—as I’ve said—we should have left well enough alone. The game gave us plenty of chances to turn around and back down, and perhaps we should have taken one of its earlier off-ramps.

So far I’ve characterized the “repeat customer appreciation” and the “should have left well enough alone” tendencies as separate, and at first glance they would seem to be mutually exclusive. In practice, though, that isn’t really the case. You can hybridize the two, move from one scenario to the next throughout the course of repetition.

Oxenfree is an adventure game about a group of teens who run afoul of a nuclear-powered gang of ghosts. The game features a snappy and responsive dialogue system that’s a joy to interact with, with the consequences of dialogue choices ranging from mild to drastic. Maybe you’ll decide that your character, Alex, should publicly embarrass her friend, quashing his chances with his crush. [“C’mon, fess up. You wanna go out with Nona, right?”] But since the ghosts’ attempts to possess the teens result in Alex getting unstuck in time, your choices could also resurrect Alex’s dead family member, and significantly change her family circumstances. Defeating the ghosts is ultimately pretty easy. You’d have to deliberately work to get a bad ending in your first playthrough of Oxenfree. The game then ends, as so many choice-driven games in the 2010s did, with a little breakdown that shows what percent of players followed your particular route. And as is so often the case with these little graphs, you might think, “Hmm, there were more branches in this story than I originally realized. Maybe I’ll give this short game a replay.” And just as that thought may go through your head, the narration glitches, and Alex abruptly shifts from her American Griffiti-style update on these characters’ futures, to dialogue that suggests she’s skipped back to the beginning of the game’s story. [“Whatever. I’m sure it will be fun. It’s something to do, right?”] Then the credits role.

And if we do elect to start a new game, it doesn’t say “new game” at all. Instead it says “continue timeline,” so that’s weird. There’s a few new dialogue options in certain circumstances this time, with Alex seemingly realizing that this isn’t her first time through these events [“I’m sorry, but this is so familiar.’]—a situation that the ghosts later confirm, ominously warning that things won’t be as happy this time through. [“This story, it is not so sweet as before.”] It turns out Alex’s escape in the first playthrough was illusory: the ghosts have trapped her in a loop, gradually wearing her down, running her through the same traumas over and over again in order to get her to finally strike a bargain with them. [“You and us, we’ve played this parts for a long time. And we will continue to play them, forever.”] And there’s a little bit of a feeling of … did I do this? Did my desire to do another playthrough trap Alex in this loop? There’s a sense of vague—and quite unexpected—culpability in this run of Oxenfree that I find much more effective than video games’ usual hit-you-over-the-head moral dilemmas.

As you continue through your second playthough, it becomes clear that at least this version of Alex is doomed. She’s not going to escape the loop. The best she can hope for is to send a warning back in time that might help another instance of Alex, somewhere out there in another timeline. And if you play the second playthrough right, that’s what happens: This Alex remains trapped, but there’s a short epilogue showing that her warning worked, and somewhere out there, in some splintered-off timeline, there’s an Alex that never went to the island, and therefore never got trapped in the loop.

So there’s a bit of both tendencies at work in Oxenfree. Your second playthrough is darker than the first, fully revealing the hopelessness of these character’s situation in a way that’s only teasingly hinted at in the final seconds of your first playthrough. The game sets things up so that you feel like an agent of these character’s terrible fate, simply by choosing to play through it a second time. But, on the other hand, only by playing your way through that profoundly pessimistic second playthough in its entirety can you emerge into that epilogue, which is arguably the game’s only good ending.

A sequel to Oxenfree released last year, and Night School Studio faced an interesting conundrum while making it: which playthroughs should they acknowledge? Presumably, not everyone who liked the original game took the time to play it twice. The two easiest options would be for Oxenfree II either to focus on different characters and Alex and her friends never show up at all, or to retcon the first game in some way. Honestly, if I were designing the thing, I would have gone with one of those routes, probably the first one. But Night School surprised me: Alex and friends show up in Oxenfree II, and it’s the version of them that’s been stuck in a loop. In fact, it gives this version of the characters a happy ending, as well, by providing them with the means to finally escape that loop. Oxenfree II didn’t hit me quite as hard as Oxenfree—it lacks that sense of deep, sinking pessimism of the first game, which was so surprising that it stuck with me over the years—but I do appreciate how seriously Night School approached the task of sequel-making. The first game had already used seriality in an interesting way, focusing on the tragic possibilities of replaying a game, and they used the sequel as a chance to do something different, let these characters finally escape seven years later. I don’t fault them for going with something new—not when they had already repeated themselves, to great effect.

And speaking of repetition and sequels, it’s time to move on to my final case study. When you first boot up and play NieR: Automata, you take the role of 2B, a soldier in a centuries-long conflict between androids (that is, humanoid robots designed by humanity as a tool of war) and machines (who are also robots, but are less strictly humanoid, and who were designed by an extraterrestrial invading force as their tools of war). There’s a wide breadth of warfare on display in NieR: Automata: the game has a habit of switching on the fly from an action RPG to a shmup, or twin-stick shooter, or side-scrolling hack and slash. The dynamism of the game’s mechanics may distract you from the fact that the campaign’s overall shape is kind of a mess. There’s a lot of repetitive running back and forth between a small handful of locations at the behest of voices on the radio. The shapelessness of the story fits 2B’s character, in a way—she’s a soldier, who takes orders well, and prides herself on efficiency. [“YoRHa aren’t allowed to be emotional, remember?”] She’s basically incurious about the world—a trait she shares with the android military operation as a whole. [“Those Council broadcasts are always so stiff.”]

There is one major plot revelation in this campaign: the aliens that originally created the machine army are long dead. As a result, the machines have become directionless. And they’re taking it pretty well, actually. They’re learning about humanity, creating new artificial life forms. Some are experimenting with pacifism. Others are inventing religion. It’s not all great—the religion, for instance, is a death cult. But in the absence of their creators they’re freeing themselves from the burden of being instruments of war. By comparison, the androids seem unable to adapt, rigidly welded to a military purpose that not longer makes sense.

Eventually, the campaign comes to a close, in a way that frankly feels rushed. You kill the mysterious humanoid machines Adam and Eve without ever learning much about them, or the motivation behind their creation. In fact, multiple times throughout the campaign you’ve been introduced to characters who just disappear from the story, with nothing having come of their introduction. As the credits role, the whole ordeal feels unsatisfying. But then the game throws up a splash screen, announcing that the campaign you’ve just played was only the “A route,” and there’s now more content waiting if you start the game over again.

So let’s start a “B” route. This time, we’re playing through the same events from the perspective of 2B’s chatty support android, 9S. And since we’ve seen these story beats before, it is rather repetitive—but the repetition is broken up at least somewhat by new elements. Gameplay-wise, 9S is more of a hacker than a brute-force fighter, so during combat you can switch on the fly from your swords and guns to playing a twin-stick shooter hacking minigame. The Simone boss fight is a good example: when you played as 2B, you were just whacking the thing with your sword, but when playing as 9S, you’re dodging attacks one moment then diving into the minigame the next.

There are other differences, as well. When playing as 2B, you saw a brief flash of some sort of scene between Simone and another machine as the boss died. When playing as 9S, you see much longer versions of these scenes, and a full inner monologue exploring Simone’s psychology and motivations. This isn’t the only time, either—during the “B” route, things are constantly being interrupted by little picture book illustrations telling you more about this world. More sidequests open up in this campaign than were present in the “A” route, partially because 9S’ hacking ability allows for richer interaction with the world, but also because he’s just more curious, drawn to question aspects of this world and its machines.

And, eventually, 9S discovers something that marks the major plot revelation of this campaign: Just like the invading aliens, humankind is also extinct. They were extinct even before the conflict began. The entire idea of androids defending Earth on behalf of humans is a ruse. This empty shell of a conflict has continued on for centuries based on nothing but habit and momentum, with neither side actually having any higher purpose to fight for.

The “B” route ends, and now it feels like, despite the repetition, we’re getting somewhere. The bare facts of this world are depressing, but the more the characters learn about them, the more hope there is for a good ending, right? During the credits for the “B” route, the game announces there’s a “C” route. So, we hit start again, perhaps expecting a third repetition of the same events from yet another perspective, but that’s not what happens—instead of looping back, we lurch forward in time now, into a new chapter of the story—one that’s filled with a lot of defeat. Major characters die, sympathetic factions suffer devastating atrocities—the whole “C” route is just an unrelenting series of losses. As we limp our way to an ending, we ultimately have to choose one of the two major player-characters still left alive to ally with, and fight the other one to the death. Choosing one character gives us Ending C; choosing the other gives us Ending D. And if we do one, then reload and do the other, the game asks if we would like to take the final option, Ending E, which requires deleting our save file. And if you take it, the game really does just that: it deletes your save file entirely. [“In exchange for all of your data… I will convey your will to this world.”]

This is not the first time the NieR franchise played this particular trick. The previous game, NieR Replicant, had also had a system where players could re-start the game after seeing the end credits, cycling through this process over and over until they were ultimately asked to delete their save file. It’s a trick that’s so memorable that it feels odd to repeat it twice—you would think there would be diminishing returns. However, I actually prefer Nier: Automata’s execution, for three reasons.

One is tedium. [“This feels alarmingly familiar.”] NieR Replicant is interminable to play through. Like Automata, it does insert additional scenes—or even in some cases just subtitles so that you can understand what the enemies are saying—as a way of deepening your understanding of this world and its various perspectives over repeat campaigns. And it also has little bits of levity that acknowledge your multiple playthroughs. [“Don’t we already have this weapon?” “Shh! These things happen the second time around.”] But these bits don’t do nearly enough to stave off repetitiveness. There’s no introduction of entirely new fighting styles, like 9S’ hacking minigame in Automata. Mechanically, the only difference between the at-least-three, and possibly-up-to-four consecutive playthroughs you end up doing of the game’s second act is that you’re continuously leveling up, meaning that each run is more and more of a speedrun. The game encourages this type of fast play by giving you achievements for things like beating bosses in 90 seconds or less, but it’s simply not enough to overcome the tedium inherent in the process.

Two is a lack of conviction. NieR Replicant originally released in 2010, with an enhanced re-release in 2021 following Automata’s success. The enhanced re-release improved the combat and expanded the map a bit, which is fine, but it also added a fifth playthrough that restores the save data that was supposedly deleted at the conclusion of your fourth playthrough, which sort of misses the point. It undoes a sacrifice that’s supposed to be meaningful. [“This is the world with the people we cherish.”] The 2021 edition features a new final scene where everyone breaks character and thanks you for playing as a way of letting you know that yep, you’re really done, for real this time. [“Thank you for playing!”] It’s an embarrassing step down from the previous version, where you knew the terminus was absolute because you simply had no more save file to continue.

But the most important reason is the third: theme. If a game is going to ask its players to bear with it through multiple repetitive campaigns and false endings, terminating in a final fourth-wall-breaking request that you delete your save file, there should be a reason. A point. Some sort of thematic resonance that motivates such an unusual, repetitive structure. NieR Replicant sort of gets there. The major revelation of the game is that you and all of your friends are clones, created by a previous generation of people that died in a plague, but whose consciousnesses have lingered on after death. The original idea was to preserve the souls of the dead and implant them into new clone bodies, but in the intervening years the clones have become self-aware, fully-formed people, whereas the previous humans’ consciousnesses now have degenerated into violent and animalistic shades. The overall point is that this new generation of replicants has a right to self-determination. The past has no claim upon the future. And I guess if you were being generous, you could say that the multiple playthroughs are a way of portraying the accumulated weight of that past, and deleting your save file means ceding your control of this video game character, just as the shades of the old humans should let go of the dream of inhabiting the replicants. I guess. It’s a stretch.

Whereas in NieR Automata, theme matches structure in a very clear and clean way. It’s revealed at the end of route “C” that 9S model androids have repeatedly learned the truth about human extinction. 2B was a special executioner unit, assigned to kill each 9S whenever they inevitably discovered this. [“It always…ends like this…”] Then a new 9S would be manufactured, and the cycle would continue. You can meet another one of these executioner units in a sidequest, who was driven mad by having to repeatedly kill friends, and ultimately erased her own memory for relief. It’s one of several plot moments in the game where memory erasure is depicted as a viable escape from trauma. And the whole setting of the ongoing war is just one big cycle of meaningless trauma. Both sides enact violence as a rote automatism, despite having no higher reason to fight. Machines that fall in battle are used as parts to build the next machines, and it’s later revealed that androids themselves are built using harvested machine cores. It’s all just the same bits of metal, re-configured in an endless cycle of harm. Endings C and D are a little bit optimistic, hinting at the cycle potentially being broken. But the only ending that truly ensures that the cycle is broken is Ending E, because it forces you to stop playing the game. [“And so, we must say goodbye.”] The very first words spoken in the game’s first campaign are: “We are perpetually trapped in a never-ending spiral of life and death. Is this a curse? Or some kind of punishment? I often think about the god who blessed us with this cryptic puzzle … and wonder if we’ll ever have the chance to kill him.” That god is you. The player is that god. You’re doing this. And if you want these characters to finally rest, you need to erase your data and stop making them kill each other.

Now, normally with these videos, I try to end on a “so what.” Or, if not that, I cheekily end them in a way that reflects the overall structure I’m commenting on. But I can’t do that here. I can’t make a video you finish once, but the next time you hit “play” YouTube will serve you a slightly different video, one that remarks on the fact that you’re re-watching it. And I think that is the so what. It speaks to how unique this particular possibility of game storytelling is.

For much of human history, it has been not narrative but music that makes the most use of repetition, teaching its listeners patterns to recognize and remember so that the variations become meaningful. When these sort of repetitious forms have shown up in other art forms like cinema, it hasn’t usually been for narrative ends—instead, it’s usually the farthest edge of non-narrative experimentation. I think that it’s because games are already so repetitive, based on recognizing patterns and trying again and again until you succeed, that they’ve inculcated their players with the necessary patience and attention to tell stories in the way we’ve examined in this video. (And it helps that single-player games have normalized the practice of multiple off-ramps, cascading tiers of closure for players who have different levels of investment.)

And then there is, of course, as I said at the beginning, the humble save file. It is the save file that makes possible these convincing false endings and re-starts, that play between reprise and variation. Cherish your save files. Hold them close, keep them safe.

But when the time comes, you should also be ready to delete them.

Leave a comment