42: SpaceChem

SpaceChem is a puzzle game at heart, where you are placed in the shoes of a reaction engineer and told to make the required chemicals from what you’re given and nothing more. You use reactors to do this, with each reactor holding two waldos (circles that read and execute instructions) that proceed along a track that you’ve defined, doing the operations that are specified there, like grabbing a molecule, or breaking its bonds, or rotating it 90°, and so on. So in this basic one below, the red waldo is grabbing an oxygen molecule, and so is the blue one, and when they get into the bonding area they’ll be stuck together and then kicked out of the output tube.

The best description for what you’re doing is a sort of visual programming, really. You have an input, a required output and a bunch of tools with which you can manipulate the input to get there. And while that is the best description, it’s not a sufficient one. The relatively small amount of programming I did in university was a source of immense frustration for me, I could never quite get in the right frame of mind to do things efficiently or well, and there seemed to be this vast database of stuff that I wouldn’t know about, that would make everything so much easier. SpaceChem is better than that.

The superb sense of context is the main tool for easing you into the trials ahead. While this is obviously not what happens in laboratories, there’s a substantial illusion that it totally might be what they do. You actually need to use stuff like atomic weights (for proper fusion) and the maximum number of bonds an atom can make, all the molecules in play are real ones and correctly represented. There’s a periodic table that you can just open up and browse to look extra-nerdy, too (and get that weight/bonding information off). Everything is presented in a realistic seeming way and, added to the visual nature of what you’re sticking together, I found it much easier to get in the requisite groove.

Like a streamlined lizard, there are no more frills than are necessary. It doesn’t hold your hand, just tells you what everything does and retreats to the observation deck to see what you’re going to do. There’s no painstaking tutorial, no handy hints, no simplifying of the concept. The reactor screen is bold and clear, and is perfect for communicating the information they want to get across – everything is on one screen, there’s no visual clutter and it’s obvious what waldo is doing what, and when. The production screen has some pretty enough backgrounds but is essentially a grid upon which you place reactors and pipes. What story there is, is told through single screens of text with maybe an illustration but SpaceChem does a good job of imbuing this potentially sterile task with a dash of personality and humour.

Your tasks are diverse enough that you don’t ever get bored, too. Roughly, they split into research tasks, where you use a single reactor-space to build a single type of molecule, and production tasks where you’re allowed to chain together up to six of the bastards to create a whole bunch of molecules. The production tasks are more free, in the sense that there’s a lot of possible intermediate steps between what you have and what you need. They take a lot longer though, so it’s nice that the research missions break them up. On top of that you have boss battles where you have to create certain things within a time limit before the eldritch horror destroys your base.

An often noted feature of Spacechem is the difficulty. Every person, every single one, who plays this game will hit a wall, a point past which they cannot proceed. There will be some ridiculous mission where the input makes absolutely no sense and the output is ridiculous and you’ve got to do it all on a space the size of a postage stamp. The thing is, they’re all possible, and constantly being forced to confront the limitations in your playstyle really appeals to me. I’ve hit several walls during the playing of this game: all crumble with time and enough thought, and it’s thrilling to come up with some new process that the game hasn’t told you.

It’s not a game you can just solve intuitively, not after the first couple of planets. Though some people will be better than others, someone who’s determined to burrow deep into SpaceChem’s carapace is going to need to plan. In my case, I’ve had to make notes on my conditions and diagrams on how it’ll all fit together and ratio work on the proportions of chemicals going in and out and I love that, too. It’s not just testing your ability to reason, it’s testing your ability to plan and manage and be efficient, all skills that a game rarely requires you to be that good at.

But it’s not just that it makes you think through the stick of being difficult, it allows and positively encourages you to think through the carrot of feedback. When you finish a level, it gives you a list of your friends scores (how long their reaction took, how many symbols they used) as well as a graph of how everybody in the world did and where you place on it. There’s also an inbuilt feature to record your solutions so they can be sent to youtube and spread knowledge to all. If ever you feel a task is insurmountable, the tools are in place to enable you to succeed. If you’re bored of the tasks in front of you, there’s a whole internet of extra challenges out there available for free, a spacechem newsletter built into the game with particularly awkward problems to keep your noggin ticking over.

Importantly, though, there is never just one solution to a problem. Because of the modular nature of what you create, because paths and tracks can criss-cross anywhere and because the end conditions are simply defined, there are untold thousands of ways for you to solve a problem. Some will be more efficient than others but you don’t have to be perfect to get through, you just need it to produce ten of whatever you are after without breaking down. Quite often you’ll end up with pipes snaking across the landscape holding the excess that you can’t figure out how to get rid of, or reactors whose entire point is to keep things in time. I had a problem where the sheer amount of possible inputs was giving me a headache, but I figured you could just ignore than and fuse everything you needed out of just the hydrogen inputs.

I remember taking this to extremes on one of the later boss battles, where you had to fuel a laser and periodically alter its direction to aim it at asteroids smashing into your space base (told you the context was awesome). I’d spent ages carefully arranging the proportions of stuff I was sending through a packed network of pipes so everything was running smoothly with no wastage before I realised I had four more hydrogen molecules than I needed appearing in the reactor which changed direction. I couldn’t just let them pile up and I was in no mood for reworking the whole system, so I decided I would just fuse them together as they came out and end up with some ridiculously heavy molecule as a waste product. This worked well up to a point, until it became apparent that in the runtime necessary to beat the asteroids, I was going to run out of possible atomic weights and everything would just explode.

I was still committed to keeping my original design, this had taken some hours already, so I jury-rigged in another system that would occasionally move along the budding super-atom before it got too super, and stick it in a growing line of massive atoms. Thing is, the only space I could find to add this additional system in was the button which controlled turning the laser anti-clockwise. Every time I turned the laser that way, the line of massive atoms would shift to the right and eventually it would hit the side of the reactor. In effect I had built a laser that worked perfectly, with the minor flaw that you could only ever turn it anticlockwise eight times before it exploded. Thank God asteroids don’t circle-strafe.

The point is, it worked for long enough for the game to pass me. This is not a game of perfection, no matter how much I go on about the constant striving for self-improvement. This is a game where you’ve got to be good enough, and the threshold for good enough keeps moving. And with me, this adds up to something incredibly addictive. I poured hours into this game over the weekend, progressing in fits and starts but desperate to play and see what I could do. When I wasn’t playing, I was thinking about playing, solutions kept striking me throughout the day while I was making sandwiches or fighting dogs or whatever it is I actually do. When I closed my eyes, I saw waldos moving across the back of my retinas, tracing out thin luminous paths. I’ve never felt so, well, rewarded by success in a puzzle game, so desperate to just do one more, so undaunted by repeated failure…maybe it isn’t much to look at, but the game’s deeper than a really deep thing. The demo is out there and covers the first planet or two, so give it a shot.

I haven’t finished the game, there’s still two of the nine planets to go and with the difficulty continuing to increase I doubt I’ll earn my winner’s medal any time soon, and I haven’t even attempted the researchnet challenges or the special campaigns, but I believe I’ve seen enough at this point to call it played and just plug away at those remaining challenges over time if they’re still interesting.

5 stars, 33 hours played and counting