Crisis on Question Mountain: Mass Effect 3’s Ending

The ending of Mass Effect 3, and of the entire Mass Effect series, broke the internet. I can’t call the ending ‘polarizing,’ because a polarized debate has two poles. Major gaming sites posted articles lambasting the ending, and protest groups formed to “Retake Mass Effect!” Elaborate strong misreadings of the ending were introduced to explain why Bioware was in fact engaged in an immense Inception game with players.

I finished the series for the first time a couple weeks ago, and I want to thank the folks at Bioware—if they’d set out to make a test case for endings and why they succeed or fail, I don’t think they could have done any better.

The Retake crowd, in addition to raising a number of continuity errors, claims that the ending reduces the setting’s complicated moral choices to a simple selection of A, B, or C, and that A, B, and C are mostly differentiated by the color of the explosion. Nothing is resolved, no questions are answered, and we’re left feeling hollow and betrayed.

On the other side stands Film Crit Hulk, a critic with deep appreciation for story, structure, formal experimentation, and fun. I’ve never yet been misled by a Film Crit Hulk article or recommendation. His long read on Mass Effect 3 identifies the Mass Effect series, up to and including ME3’s ending, as gaming’s Citizen Kane. That article nudged the ME series into my ‘to play’ column. I’m not generally a Wes Anderson fan but I loved Moonrise Kingdom, which I saw on his recommendation; I don’t generally sink 100+ hours into a franchise that looks an awful lot like a cover-based shooter, but the Hulk piqued my interest, and the further urging of forceful and brilliant friends sealed the deal.

In the article I linked above, the Hulk explains his love of the ending: how it resolves the series thematically and makes a powerful statement about the cyclical nature of history. To Hulk, even the similarity of the three ending cinematics is part of the text: we’re meant not to know what happens next, to stand on that alien planet with Joker and EDI and wonder what kind of strange new world our deeds have wrought. The point of inflection looks similar, but the futures look wildly different—and we’re left to wonder at that difference.

So, who’s right? The Retake crowd, or the Hulk?

Both, I think, though in different ways. (Yay, Synthesis Ending!)

(And here I put a cut, because I’m about to spoil ME3’s ending in detail, and offer minor spoilers for the ending of Hyperion and Season 4 of Babylon 5 to boot.  See you on the other side!)

Hulk’s right that, thematically and structurally, the ending works. ME3 elegantly resolves all the major conflicts in the series, one at a time: the game is all closure. We bring the krogan genophage plotline to an end; we participate in the resolution of the war between quarians and geth. We even learn the not-so-dark secret behind the asari façade of superiority. The way we resolve these conflicts drastically alters the direction we take into the final encounter, and that direction drastically alters the impact of that final encounter. The A, B, C choice is the culmination of all the choices your character has made previously in the game—in the end, the fate of the galaxy comes down to Shepard, as it has all along. The ending combines the series’ fascination with choice and its obsession with historical and narrative cycles. ME1 ends with the decision to save or destroy the Council. ME2 ends with the decision to save or destroy the Collector base—which decision, for the record, only changes the color of the ensuing explosion. And ME3 ends with the Star Child, and the A, B, or C option. Cycle after cycle, each terminated by a choice.

At the same time… Well. With all due respect to Hulk, and there’s a bunch of respect due, the guy is a wonderful writer and an insightful critic, the fact that so many people had problems with ME3’s ending indicates that something’s off no matter how perfect the ending looks on paper. To say otherwise, we run the risk of being that mad scientist in the collapsing tower screaming “But my calculations were perfect!” while our mindless cybernetic insect creations scuttle forth over the landscape, spraying acid spittle everywhere.

Sorry, analogy got away from me there. (As did the critters.) But! If the ending works on a thematic and structural level, and I agree with Hulk that it does, then the question becomes, why did many gamers rise up in fury? The answer I propose: that the problems people have with the ending arise from the final scene’s mechanics—that is, Mass Effect 3’s ending was the right ending, but that scene’s line-by-line writing and setup made it difficult for people who don’t naturally read for theme and structure (and most people don’t) to derive satisfaction from the thematic and structural resolution.

To put it another way: I think the house is tastefully built, and the living room well-furnished—only its walls are painted a bowel-shivering chartreuse.

(Yes, I am saying that the Retake crowd may be wrong about what’s actually wrong with the game. This happens all the time in editorial discussions: A first-pass impression that “there’s not enough X” often means there’s actually too much Y, or even too much X. “Mary’s motivation makes no sense” can mean John’s motivation doesn’t make sense. “You start this scene too early” sometimes means you start the scene too late, or this one scene should be two scenes, or that the scene isn’t even a scene. A friend of mine is a classical portrait painter, and every time I tell him something seems wrong with an elbow in one of his paintings, he’ll come back a week later and tell me the problem was actually the shadow on the model’s opposite shoulder. Awareness of a problem does not equal ability to diagnose and correct the problem—obviously, or else there’d be no doctors.)

This is where spoilers fall hard and heavy, so if you’re worried about that sort of thing, stop now.

Let me paint you a picture.

Shepard wakes at the end of time.

Her last memory: crawling across the floor of a control room in the Citadel space station. Almost made it, too. Built the superweapon to end the Reaper threat, found the secret power source known as the Catalyst—only nothing happened. So, bleeding, burned, broken, she crawled across the control room to try, one last time, to save the world.

She didn’t make it. Strained for the controls. Collapsed.

And she wakes, at the end of time.

She hasn’t been here before. She stands beneath the superweapon itself, in the vacuum of space, somehow alive. Above, a battle rages between the Reapers and the fleet she’s assembled to defeat them. Cruisers and destroyers spark like fireworks and fade, brilliance to cinder in seconds. Ships burn fast in space.

In front of her stands a child, translucent hologram, shining with starlight.

The child opens his mouth, and this is where everything goes a little wonky.

The child tells Shepard that it is both the Catalyst Shepard has been searching for—the secret ingredient for the superweapon—and the intelligence secretly animating the Citadel space station, and the guiding mind of the Reaper fleet.

This is a perfect time for us to meet a character who could explain (or fail to explain) the context of our actions—structurally and thematically, we’re golden. But this is a character the audience has never seen before, and we’re being asked to accept a great deal about him in the last few minutes of a game. On top of the three different identities thing, we’re forced to change our answers to two important story questions (Who do the Reapers work for, the answer for which was previously ‘Harbinger’, and What is the Citadel, which answer was previously “a Mass Relay”) in ways that don’t quite make sense given what we already know about the story world.

The Citadel, we know, is a Mass Relay (Stargate-like device to enable long-distance space travel). In the first game, a Reaper advance scout tried to take control of the Citadel in order to help the main Reaper fleet shortcut into the heart of the galaxy. If the Citadel were the guiding intellect of the Reapers, why was that necessary? In fact, in the first game we learned that the Citadel itself was hacked to keep it from obeying the Reapers’ commands. How would that even be possible if the Citadel is the guiding Reaper intelligence?

It’s possible to construct answers for these questions. Perhaps the child only lives on the Citadel when the Reapers control it. Perhaps it sleeps while the Reaper fleet sleeps. And perhaps the child isn’t just in the Citadel—it’s actually part of the entire Mass Relay network, which functions as a sort of quantum-entangled hypernetworked brain. But the part of the audience’s mind that would answer these questions is directed toward other, more immediate story questions. Those questions don’t go away, though, they’re just hidden, swept under the rug, raising the background level of confusion in the scene.

“Okay,” says the audience, “Reaper / Citadel / Catalyst child, I’m with you. Why are you trying to destroy all organic life in the galaxy?”

To which, the child’s answer: “I’m trying to save organic life from inevitable conflict with robots. Turns out the best way to do that is by archiving entire civilizations in the form of godlike insect-robots, with the individual members of those civilizations trapped for all eternity in a single scream of exquisite timeless agony.”

“Um,” says the audience.

“But at least they aren’t exterminated by robots!” says the child, helpfully.

Does this make sense? Sort of. I can imagine a “mad” neural network seeded with a very specific goal (‘find some way to stop organic life from being exterminated by robots’) that would come up with this sort of exceptionally inhumane, yet (in its mind) ‘efficient’ solution. After all, twisted immortal cyborgs are still part organic!

But questions remain: “Is robot conflict really inevitable? I just spent this entire game matchmaking between organic life forms and robots. My pilot’s dating a robot. I ended a three hundred year blood-and-oil feud between some robots and their masters back on Rannock.”

Shepard doesn’t ask any of these questions, instead opting to make a little noise about free will. The child, for his part, asserts the inevitability of robo-pocalypse. Again, we could answer these questions ourselves, again—perhaps any peace between organics and synthetics is doomed to be short-lived. Perhaps leaders like Shepard are less than one in a quadrillion. Certainly if Shepard hadn’t spent the last 120+ hours of gameplay trying to save the galaxy, these robots would still be firmly in the ‘Kill all humans’ camp. But we move on, full clip, to another operational question, leaving all this stuff unanswered.

Note that I’m not saying every raised question must be answered! But, sweeping these issues under the rug along with the earlier questions about the Catalyst, we start to see the lump in the center of our rug. We grope queasily for solid information.

“So how do I stop you?” Shepard asks.

And the child tells us: now that you’ve plugged in the superweapon, you have choices. You can do A, B, and C, all of which stop the Reapers from destroying all organic life in the galaxy. I’ll get to A, B, and C in a second, because a doozy of a question arises at this point.

Namely, Why should I believe anything this kid says?

The child is, self-professed, the Reaper Overmind. It’s spent the entire game trying to kill you, through various agents. It’s destroyed your planet. Its mind-controlled servants tried not ten minutes ago to mind-control and murder you, to stop you from activating the superweapon. It is 100% convinced of the rightness of its path. Never once gives a hint that it finds the ‘Reaping’ distasteful or regrettable. Lawful Terrifying alignment in a nutshell. And Shepard is about fifteen minutes away from dying from her injuries.

The kid could just have not manifested, waited for Shepard to die, mopped his holographic brow, and said, ‘whew, that was close! Back to the reaping!’ Yet he not only talks to Shepard, he tells her how to fire the superweapon—in fact, each option, A, B, or C, involves a non-intuitive action, every one of which looks like it might break the thing, kill her, or both. (For those following along at home: we shoot a vital-looking conduit filled with caustic gas, or grab two prongs of something that looks an awful lot like a charged capacitor, or jump into a beam of superheated plasma.)

One frustration I’ve read that Retakers have with the final scene comes from the fact that you the player can’t take an option other than those the child offers. We want to create our own future! I think this is sort of misplaced. Luke, at the end of the trench run, has a torpedo, knows (on some level) he’s about to kill a few million Imperial soldiers, and has to decide whether to fire or not. He doesn’t get to create another option just then.

But—he believes that shooting the torpedo will win the war. Because someone he trusts told him it would. Several someones, and they checked their work. He knows that he is making a meaningful decision based on reliable information.

We don’t have the same trust in the Reaper Overmind. Quite the opposite, in fact. It has every incentive to mess with our heads, and no incentive to tell the truth. And I think this is the reason people want another option: they feel the child is selling them a bill of goods.

Again, a question easily answered with half a line of exposition. The superweapon, in fact, is more of a clever hacking device—the child is being forced to tell us these things. Or the superweapon’s hack has already convinced the child it is wrong, and it needs our help to reprogram itself. I’m sure you, Dear Reader, could come up with an answer or two more. But none of these questions are answered, and so they go under Mount Rug with all the rest. By this point Mount Rug is more Mount than Rug, and our questions are spilling out from underneath. Bow before Mount Question!

And then, two final points of mechanical confusion. First: the superweapon can, depending on player choice, either: destroy the Reapers, control the Reapers, or make the Reapers irrelevant by creating a synthesis of organic and technological life. (I’ll get to that in a moment.) Imagine, if you can, a reason to build a device that does all three things. What engineer would, designing a weapon to destroy the Reapers, build in a remote control for them, not to mention a bonus function capable of rewriting all life in the galaxy? Or, say we’re going to control the Reapers—why build something that would blow them up, or do whatever synthesis does? Who drew up these plans? Can you imagine the dev meeting? “Basically we think our one-shot ultraweapon should do one of these three things, depending on the whims of the person who happens to be standing at the control panel.”

Not very compelling, right? But, fine: maybe the device was actually built to Control, but by sabotaging it you can Destroy the Reapers. Maybe the device was actually built for Synthesis, but can be jerry-rigged to Control or Destroy. Again, a half-line, but without that line, Question Mountain grows. Variables tangle and recombine in our heads. Do we know anything at all? Have we known anything since Shepard woke up on this strange space platform?

And then: Synthesis. Which, I mean. Speaking of questions.

The following quotes are verbatim from the child in the original ending. The Extended Cut doesn’t change much here.

“The chain reaction will combine all synthetic and organic life into a new framework. A new… DNA.”

Wait, what? DNA isn’t just a word, right, it’s a molecule that behaves in a specific way and actually about half the aliens in the setting have different ‘DNA’ from humans, and anyway machines don’t have DNA, and what exactly do you mean by chain reaction anyway coz there’s, you know, a lot of empty space in the galaxy which makes it hard for there to be a chain reaction with, well, okay, maybe you’re being metaphorical and I can dig it, let’s hear what else you have to say—

“Your organic energy, the essence of who and what you are, will be broken down and then dispersed.”

Huh? My ‘organic energy’? If we’re doing anything on a galactic scale we’re talking energy expenditures that dwarf anything Shepard could add. Even if she turned into 100% raw energy via antimatter annihilation, you’re not going to get close to that level. Plus, then you’d have, you know, energy, which is fungible and interconvertible with matter in the same way as any other energy because there’s only one kind unless we’re talking about dark energy which we’re not I think. But, oh, maybe, you’re talking about my soul, which would be great except we’ve seen no evidence in this game that such a thing exists, and in fact the entire setting’s gone out of its way to be monist and hard-SFnal, even to the point of saying “Telepathy is impossible” while making up weird quantum entanglement technobabble to explain telepathy. Mind control is explained as electromagnetic interference with the brain. So, okay, I guess souls are a thing now, is there an afterlife then? Is Thane there? Is Legion? Do robots have souls? Do organic and robotic souls behave differently? Question Mountain rises.

“The energy of the Crucible, released in this way, will alter the matrix of all organic life in the galaxy.”


“Organics seek perfection through technology. Synthetics seek perfection through understanding. Organics will be perfected by integrating fully with synthetic technology. Synthetics, in turn, will finally have full understanding of organics. It is the ideal solution.”

….Actually that bit’s fine, though we’ve been here before: Shepard’s a decent chunk cybernetic, and every Biotic (magic user, sorta) in the game has cybernetic implants. Meanwhile, the synthetics who hang around organics for a while come to understand them more or less perfectly.

So we have a hilariously improbable device built to do something incomprehensible. And, again, we could buy it if we wanted to. There’s been equally silly technobabble throughout the series. But… well, by this point Question Mountain towers Everest-like over the conversation, our rug covering its utmost snowy peaks. There’s so much we want to know. We’re talking to someone whose existence, motives, and ethics are each gigantic question marks, inside a question mark device, the function of which is an immense question mark that raises fractal little question marks.

And yet, and yet, and yet… it’s thematically excellent. We choose between two untenable alternatives and the Better Way. Any path destroys the world, while promising a new life, a new chance, for those who remain. We have met the Man Behind the Curtain, alone, and he has asked us to choose the future.

Structurally, too, it works. The series starts with Shepard running and meeting her mentor, Anderson. The series ends with Shepard’s last conversation with Anderson, and, in the end, with Shepard running. We build a squad, a team, and we bid them farewell. We see those we’ve lost one last time before the end.

But Question Mountain looms. These issues, little and big, undercut our sense of closure. Most subsets of these problems we could ignore, or explain away in true grand No-Prize fashion. But all together, we’re left groping for a coherent narrative universe even as our eyes are supposed to be opened to the universe of possibilities.

And it’s not like it had to be this way. A slight redesign of the superweapon, and we have a device that requires an organic being, untainted by the Reapers, to die, in order to use it. Excellent safety measure. And, when used, its energies are entirely at the mercy of the person in the driver’s seat. She decides what happens next. A simple visual redesign eliminates most of the questions about the device, and goes a decent way toward resolving the Synthesis tangle.

With only one point of interface, it’s obvious what Shepard needs to do when she arrives in the final room. Use the superweapon. Walk into the light. She tries to get there. The Reapers manifest in front of her as a child—because they want a form to which she will be receptive, a form they can use to plead their case. Because she’s safe from them now, and can decide their fate. We’re no longer asking ourselves why we should trust the Reapers to tell us how to stop the Reapers. We know how to stop them, and we know what the child’s goals are—to present its view of the world and ask for our help, or at least our mercy.

Why are they afraid of us? Because when they’re in our galaxy all their energies and thought processes run through the Mass Relays, which in turn connect to the Citadel. Our superweapon has backdoored us into their brainstem. Which resolves our first questions about what the Catalyst is, and how it makes sense for the Citadel to be the Reapers’ heart.

The Reaping / archiving is as sensible or not as ever, but that doesn’t matter because you’re not asked to trust or like the child. You already know how to use the device. Its A, B, or C design doesn’t seem as weird when presented as, say, dialogue options appearing once you’re inside the device, or as Interrupts. (And wouldn’t that be cool? The sudden appearance of a Synthesis Interrupt at the last possible moment?)

And Synthesis, well… For that one I think have two paths: either we need radical conceptual clarification, or we just leave it as it is—a thematically appropriate, and symbolically beautiful, hot mess of a hard SF explanation for an event that should, and does, feel mystical and redemptive.

With about two thirds of the narrative questions rendered irrelevant, the remaining ambiguities feel more numinous than frustrating. In the end of Hyperion we don’t understand the Time Tombs or the Shrike, really—but we know enough about their world to understand that they make sense inside it. Midway through Babylon 5 season 4, we still have oodles of unanswered questions about Shadows and Vorlons, but we know the important stuff, so we can let the rest go.

This same principle extends to the rest of the ending. Maybe if we weren’t already so staggered, we’d be able to answer the question of “wouldn’t the Mass Relays blowing up destroy the galaxy” with a sensible observation like “you can blow up a nuclear bomb with c4 without detonating the nuke.” We could accept that Joker’s running away from the blast because, shit, giant explosion. Our squad members’ escape is harder, but maybe we could have swallowed that too. Because we were watching a story about sacrifice and rebirth, about cycles and the ending of cycles. Because Shepard died so we might live, so that we may be born again.

But the ME3 ending’s troubles with information release, I think, obscure its excellent thematics & structure. The ending accomplishes its most difficult goals, but falls down at tasks that, while less challenging, are so fundamental they are hard even to identify, let alone diagnose. It is, in short, brilliant, and flawed, and as a result, makes an excellent test case for the analysis of endings and what they mean.

At least, that’s my take. What’s yours?

2 Responses to “Crisis on Question Mountain: Mass Effect 3’s Ending”

  1. ralph

    I agree with what you say about questions, but I don’t think the star child narrative couldn’t have been saved, for one thing I was never emotionally involved with that one nameless kid that died in the beginning of the game 3 I saw the nightmares that shepard was having about him boring, had you changed the nightmares to be about a fallen squad mate instead (thane for example, who I was emotionally attached to and who died at the end of the second game for me) maybe I would have seen those nightmares to mean something other than a nuisance to get to the next sequence.

    Another thing wrong is the A B C choices, this game was marketed based on having an ending based choices, so much so that bioware specifically said publicly that it wont be an A B or C ending, what many ppl dont understand is many fans had beaten ME1 and ME2 in multiple ways in anticipation that ME3 would deliver on its promise of multiple endings based on choices, so the A B C ending just seems like a betrayal to me since I wanted to earn my ending and the realization that all six of my shepards get the same ABC ending is just bull

    • max

      I took the kid who died in the Prologue to be a representative of all the people Shepard had lost—when I was running after him, IIRC, I heard the voices of dead party members and characters. I only heard Ash’s voice at first, because I kept everyone else alive, but Mordin and Thane showed up eventually as they passed on. That worked better for me than showing the individual squad mates, which would have felt too much like a reunion, when I was supposed to feel as if I were chasing the parts of the world I’d lost.

      As far as the ABC choice / marketing argument, well, I played these games late, and I never pay much attention to marketing, so I never felt as if I was being lied to. That said, the ending in which (spoilers) Shepard cures the Genophage, builds peace between the quarians and geth, and so on, then destroys the Reapers, sacrificing all her synthetic buddies, is very different from the ending in which Shepard, say, cons the krogan into believing the Genophage was cured, shoots Legion in the back of the head, and then destroys the Reapers. (Just to pick two noteworthy plot points.) The different paths you walk make the same final cutscene mean dramatically different things. I like the message there: we all die eventually, the question is what we do on the way. And your choices in previous games do determine the game’s end-state: kill Wrex and he’s not around to lead the Krogan in ME2 or ME3, which might lead to Shepard making a different choice about the Genophage in ME3; if Kasumi’s not around you have to let the saboteur get away with murder (or kill the colony), kill the Thorian colonists and they never form their psychic death squad, etc.

      If that’s not what you expected when going into the game, though, I can see how you would be frustrated.


Leave a Reply