Skip to content
 

An improved ending for The Martian

In this post from a couple years ago I discussed the unsatisfying end of The Martian. At the time, I wrote:

The ending is not terrible—at a technical level it’s somewhat satisfying (I’m not enough of a physicist to say more than that), but at the level of construction of a story arc, it didn’t really work for me.

Here’s what I think of the ending. The Martian is structured as a series of challenges: one at a time, there is a difficult or seemingly insurmountable problem that the character or characters solve, or try to solve, in some way. A lot of the fun comes when the solution of problem A leads to problem B later on. It’s an excellent metaphor for life (although not stated that way in the book; one of the strengths of The Martian is that the whole thing is played straight, so that the reader can draw the larger conclusions for him or herself).

OK, fine. So what I think is that Andy Weir, the author of The Martian, should’ve considered the ending of the book to be a challenge, not for his astronaut hero, but for himself: how to end the book in a satisfying way? It’s a challenge. A pure “win” for the hero would just feel too easy, but just leaving him on Mars or having him float off into space on his own, that’s just cheap pathos. And, given the structure of the book, an indeterminate ending would just be a cheat.

So how to do it? How to make an ending that works, on dramatic terms? I don’t know. I’m no novelist. All I do know is that, for me, the ending that Weir chose didn’t do the job. And I conjecture that had Weir framed it to himself as a problem to be solved with ingenuity, maybe he could’ve done it.

And, hey! I finally figured out how Weir could’ve done it. As I said, the challenge is to avoid the two easy outs of a pure win, in one direction, or a failure, on the other.

So here’s the solution:

Have the spaceman get rescued—by the way, it’s a sign of the weak characterization that, even though I read the book and saw the movie, I still can’t remember the main character’s name—but have that rescue require resources that would otherwise have been necessary for the space program, so that, as a result, future missions are canceled. The astronaut then for the rest of his life has to live with the fact that all his and his colleagues’ ingenuity did manage to save him, but at the cost of ending future manned exploration of space. Bittersweet.

17 Comments

  1. Radford Neal says:

    No, your ending is too much of a downer – just a failure, really.

    How about he’s rescued, but in the process is rendered a cripple, who won’t be able to go to Mars again, even while others will. A tragedy for him personally, but not for the values he holds dear.

    The book does try to show the personal cost to the family of the astronauts who do the rescue.

  2. David L says:

    Close but I think not quite. As you point out, the strength of thy book is the cycle: fix causes problem causes fix causes problem. It’s unrealistic that that cycle just…stops; what makes that book resonant is the familiarity of the cycle. So the better ending would have been that the nature of the problem changes radically, but we can see the next problem coming.

    E.g., he lands at great expense and future missions are cancelled. So the next day, he starts figuring out how to get new funding for NASA….

    That way we end with resolution but not conclusion/finality.

    • “E.g., he lands at great expense and future missions are cancelled. So the next day, he starts figuring out how to get new funding for NASA….”

      Paving the way for the most boring sequel ever, in which we watch our protagonist working on grant proposals. Catastrophe strikes, when the typeface he’s written it in isn’t on the approved list! Which budget items contribute to the indirect costs, and how much? Thankfully, our hero is quick with math! What to do with the co-PI who contributes nothing to the proposal, but will happily suck up funds — shove him out an airlock, or just grit your teeth and press on? I can’t wait…

      • Andrew says:

        Raghu:

        Remember, I already have ideas for two movies about real-world science and academia:

        Second Chance U: a rollicking slapstick comedy in the Bad News Bears / Slap Shot genre (“they need to put together a shiny new faculty, pronto! So our hero scoops up Diederik Stapel, Frank Fischer, Michael Bellesiles, and about 15 guys from the back pages of Retraction Watch . . .”)

        The New Dirty Dozen: a science-based action thriller (“One of those cold-fusion guys is still alive, right? He’d be brought in at the midpoint of the movie, sort of a Kevin Spacey kind of thing where he only has a small role but he steals the show.”)

        I honestly think that both these movies could be great. Then again, I don’t know anything about movies.

  3. A lot of it was him alone, written first person. That doesn’t leave a lot of latitude for reinforcing names. But then I can never remember the names of people in books or movies or even the names of people I’ve met only once or twice.

  4. “Captain Blondbeard” is a Perfectly Memorable Name.

  5. Jonathan says:

    Problem with your ending is unreal economics are in the premise: they have set up rockets dropped in advance so I conclude they have the resources to divert. As to an end, one way to approach that problem is by looking at problems with the beginning: that a dust storm in negligible atmosphere not only has the same impact as a similar storm in a denser atmosphere, that we’re really bad at designing structures which survive and, the hidden flaw, that if storms could do this then how has any rocket dropped in beforehand remaining standing? There’s some interesting ‘space’ in that, by which I mean cognitive space that illuminates – since this is a statistical blog – statistical thinking.

    I mean the guy is stranded because a rocket that’s been sitting there for some long period – greater than a year, it appears – which must have been through the gamut of Martian weather suddenly decides to tip over AND the guy is able to leave Mars because there’s yet another rocket which has been standing through the Martian weather. To skip ahead, this implies that the purely logical answer would be for another dust storm to come up as he’s getting ready to leave Mars and for that to either tip him over or off course, etc., meaning any negative outcome desired, or escape by the very milimeters of his suit, meaning any positive outcome. To get there requires rigorously somewhat more space than I want to take today but which I can describe using basic function concepts: there is a set of functions that generate negative and positive outcomes relative to what is considered neutral and ‘neutral’ is a multi-dimensional space that renders as a series of planes tied together by rotating them so a perfectly straight line disappearing into the z-axis is actually a series of infinitely stacked planes which essentially open up or ‘unfold’ to you as you count along z. As a note, this allows you to measure off the ideal, including in geometries and thus shapes, particularly when you define the ideal and draw the shapes which fit the ‘anti-function’, which is a complex idea for the functions that generate other than in the ‘pro-function’ (by which I mean the equations, given or derived, that generate observable or measured – or expected, inferred – results in that idealized frame).

    In this case, you derive the balancing equations which you obtain by finding the endpoints and then testing the paths between for the best fit to identity. So you start with the end, which asks about the beginning, which has the interesting effect of discarding the middle, and that leaves ‘he was left behind by a rocket’ and ‘he leaves on a rocket’, and that, for example, cuts out ‘left behind’ and ‘leaves on’. Or identify the identity and look deeper at ‘rocket’ and its attributes. That x-y plane quickly focuses you on falling or standing up because the crew left him behind because the rocket was tipping over. The identity in this would be something like ‘rocket’ and a statement about both were standing – see how ‘now one isn’t’ is now ‘were’. I could two ways here: I could investigate rocket and make a movie in which the company built a defective rocket or NASA sent this one up knowing there was a 1 in 3 chance the strut would fail in mild weather, etc. That would be a different movie and that says I should try the other side because the goal is not to make an entirely different movie but this one. I could go into the inability to design or the actual mild impact of Martian wind but that’s also not the movie. What is the movie then? Well, there’s a rocket standing at the end and one tipping over at the beginning so we balance those by adding tipping in at the end (because we can’t take it out at the beginning because, again, that’s not this movie). I’m not saying this is the only solution that fits, just that it fits really well, partly as I’ve explicitly described but also because it embodies the cool and the stupid in the movie by encoding the ‘maguffin’ of the rocket tipping into both ends as evenly as you want. You could, for example, have him show up and the rocket is partly tipped and, voila!, that makes rescue easier (as you realize with joy as the calculations are hurriedly processed on screen). You could do anything because you’ve now made a really good inference engine!

    About inference, you should be able to see there is a chain which encloses chains and, when without understanding the underlying model, you see that can at least be read as inference (forward and backwards). The neat thing with the Martian example is you can define a middle which you can largely eliminate or can put in with great detail and you can see that all the choices made could have been made differently and some would hurt survival and some would not – like the contrived potato disaster put in because a guy actually surviving on Mars while the earth tries to rescue him needs, you know, some spicing up for interest – but they all generate endpoints to an endpoint which is the end of the freaking movie. Note that making another movie means you shift the orientation of the z-axis: you can make another movie with minor alterations that respect the constraining endpoints or you can make an entirely different movie. Like a sex farce or Jerry Lewis appears and talks to the Martian and you realize he’s not only dead but he’s actually living in a 1950’s TV show.

    The processes of inference become clear: they are directional from any perspective and direction reduces visualization or morphism onto an x-y plane (like more or less, like observable or not, or any other existence statement) but gets highly complex when strung together along the z-axis. Note how when you plot this way you can more clearly see the contribution of a certain depth of plane to an idealized view, that being when z is hidden, and thus how you plot. Interestingly, the confounding point is really close; two steps and it isn’t that the couplings become too numerous but that the couplings among the couplings confound so you can’t reliably fix sign or magnitude or other measures of counting value from your observational perspective. That’s the problem with social science research in a nutshell: it’s imposition of orderings on an associative set overlaid on a counting set, which is the same as saying it’s imposing uncountability on a countable set while expecting to know which branch of countability you’re in. Imposing means you carry out some process, even the most basic iterative one, and to get deeper into that I’d have to discuss the halting problem. I’m avoiding Riemann already. It’s interesting to stick with Cantor and note you use uncountability to define the reductive ‘process’ which enables countability within uncountability. I believe George would say you can diagonalize anything. Anyway, an explicit ordering includes implicit orderings, an easy diagonalization unless you artificially restrict or constrain, and these implicit orderings have weights in any explicit ordering, as any API enables a set of functions on some variables within the context in which that API has functional meaning. I love how counting set layers becomes functions.

    BTW, is it what Cantor did that gave him mental problems or did he do his work because of his instability – the ‘balanced’ answer to someone unbalanced. I find versions of ‘if you look good, are you good” and “are you good if you look good” to be huge fun because they embody George’s essential concept, though he never realized it, and I assume that was because it was hard enough for him to maintain thought in the then extraordinarily thin air of forms of infinity made somewhat rigorous and the super heavy, almost immovable rock of set theoretical descriptions of whatever we can conceive and label. Seriously straining stuff: pushing way up into the thin air as you describe the literal edges of big and little while simultaneously considering how that organizes (into a specific field, btw). I don’t see how he could have thought of the existence and process statements inherent in what he was observing because it was so very hard to describe what he was observing!

    Look at poor Goedel: so consumed by his statements demonstrating the inability to contain that he, if the stories are remotely true, actually concocted beliefs that a ‘flaw’ in the Constitution could lead down a specific set of statements – that he has laid out in perfect order – to a dictatorship. Can’t contain it! Must follow it out! Next step leads to the next step and, look: they’re all in perfect order so far! I assume Einstein used to remind him Alexander the Great cut the knot with his sword, meaning there are far easier, far less remote, for more plausible ways to reach the same result. It’s great example of not recognizing how inferential chains have endpoints that may be reached through better or worse paths. Talk about your garden of forking paths: each choice has directions which confound only a second step in and which may flip one step in as the confounding levels change in associative relationships. See Riemann again. This becomes path-integral stuff really quickly and I don’t need to go there. I do sometimes imagine Kurt and Albert talking about relative limits. Those would have been some of the best conversations in human history. Albert truly understood that relativity is not just physical and not just fundamental in a physical stuff sense – his letters are clear – and Kurt worked at the limits of relative existence within other relative existence.

  6. Jeff says:

    The story revolves around Watney* and how he survives each successive challenge with his personal brand of reckless ingenuity. Even when he gets help, as when his team comes back to rescue him, the plot unfolds so that their presence creates a situation in which he is able to take decisive action to rescue himself.

    It’s a common notion that even a brief mission orbiting the earth changes your perspective on humanity, and our guy has just spent a few years being a glorious caricature of self-reliance. The movie ending suggests but doesn’t dwell on an obvious challenge for him: now’s he’s on the ground, part of a bureaucracy, teaching a survival course? When I saw the movie I couldn’t help but feel some sympathy for him, though I don’t think that was the intended reaction. It wouldn’t take any major plot changes to frame his re-entry into society as his next major problem.


    *I can’t escape the feeling that I should have preceded this with a spoiler alert.

  7. Xi'an says:

    I literally fainted [in a plane to Boston] watching the beginning so the ending will forever remain a mystery for me. Ideal, no?!

  8. Emily says:

    Most films about people marooned follows the usual character development that ends in the third act where the protagonist achieves “a higher level of self-awareness, which in turn changes who the character is becoming”. Example like Gravity, where the the tension lies in the protagonist’s will to live which ended satisfactorily with her survival after she decided that she wanted to go on with her life. In Cast Away, the first act, the character is established as the familiar ambitious, successful and stressed high-performer. It ends with him potentially finding a calmer and happier life. The problem with the Martian as I remember it was that the protagonist’s character arc was not as compelling. It was a series of problem-solving tasks, on both sides of Mars, which really didn’t end with anyone becoming a different person due to the struggles in the second act. Watney was established as a brilliant botanist and exited as an even more brilliant botanist (with tenure), which is pretty unsatisfactory. The drama in film, just as in a statistical graph is the presence of tension and release. (Preconceived notion vs Data; Two opposing trends etc.)

    The best resolution in The Martian would therefore have been for Watney to completely relinquish his self-reliance at some point to be able to be saved. Feeling helpless and frustrated that he couldn’t solve a problem, just to realize that sometimes, a person has to rely on other people and just let it happen.

    • Andrew says:

      Emily:

      Good points. I thought the last part of Cast Away (after Tom Hanks gets rescued) wasn’t nearly as good as what came before, but it was good enough, certainly much better than had the movie ended in the style of The Martian, with people around the world cheering the hero.

      Also, +1 for your analogy of dramatic tension to the statistical issue of posterior predictive checks, Popper/Kuhn, etc.

  9. Mitzi Morris says:

    have that rescue require resources that would otherwise have been necessary for the space program, so that, as a result, future missions are canceled

    The book did acknowledge this problem – remember the failed attempt to send a couple of tons of food using a Chinese rocket that would have otherwise been used for an unmanned flight towards the sun? I found this to be one of the best and most dramatic parts of the book.

    having read the book first, I saw the movie a few months later in 3D, and even in 3D it was far flatter and much less fun than the book. the differences between the book and the movie are a nice case study in what happens when you try to mainstream geek fiction: the telling technical details are replaced by generic plot heroics. In particular, the endings of the two are pretty different.

    In the book, the Martian dude launches himself towards the mother ship and a member of the crew floats out and grabs him. In the final paragraphs he concludes that he was rescued because we humans are generous and can’t help ourselves from helping one another.

    In the movie, the Martian dude launches himself towards the mother ship. The captain gets her diva on and overrules protocol to be the one to do the big rescue spacewalk, which takes far too long – as if there were some chance he might never be rescued at this point? (OK, director was Ridley Scott. Maybe the Alien could have showed up). The final scene is set at the astronaut summer camp and the Martian dude turned teacher makes a speech about how it’s up to each individual to have the skills to rescue themselves. Rah rah the space cowboy.

    In sum, I liked the book just fine, and don’t really see your objections to the ending. As for the movie, meh.

  10. Noah Motion says:

    I’m pretty sure China diverts resources to support his rescue that prevent, or at least delay, future Chinese space exploration. So it kind of did end that way, just not for NASA specifically.

Leave a Reply