Home > gamethinkings > SpaceChem

SpaceChem

Spacechem is a work of dazzling originality and almost ferocious intelligence; its concept is brilliantly realised and the whole thing sparkles with love and polish. It is a game of unexpected beauty, charm and humour; a celebration of creativity and experimentation. It is a game about designing your way through sophisticated multilayer puzzles. SpaceChem’s most significant achievement is the way it is able to foster creativity and experimentation without having to sacrifice any of its complexity or subtlety. It introduced me to a compelling new skillset and type of thinking; giving me the space to learn, experiment and play with its game system. It has left me surprised, delighted, excited and intimidated – both at the demands of some of the puzzles, and at what I have managed to produce in response to them. With SpaceChem, Zachtronics Industries have created a charming, superbly rich and uniquely creative puzzler. Its rightful place is alongside Braid and World of Goo. I wholeheartedly direct you to the demo. Let’s see a video of it in action. Set the quality to maximum and enjoy this one fullscreen with the sound on.

Spacechem is a design based puzzle game. That idea deserves explaining. At its most basic, Spacechem presents you with a grid of squares. There are four special zones in this grid: two input zones and two output zones. Your job is to take certain inputs from the input zones, process them in a particular way, and then put the desired product into the output zone. You do this with the two ‘circuits’ at your disposal. Each starts off as nothing more than a ‘start’ command placed somewhere on the grid. You have to design a circuit that will do whatever processing is required to turn the inputs into the outputs. Since I started writing this article, Zachtronics industries have put a better explanation up themselves. Watch it below.

These two circuits each have a ‘waldo’, which you should think of as a cursor that moves along the circuit executing instructions as it comes to them, picking up and dropping off molecules and so on. It’s the bit that looks like a (). The commands that you can put into your circuit include: requesting a new input molecule, picking up, dropping off, bonding, splitting and rotating molecules. You can direct each circuit to go wherever you like by issuing directional commands, and can order the two circuits to synchronise in various ways. You can get it to do double loops or fancy things if you want. There are two circuits involved in each reactor you work with and they are largely isolated from each other; commands issued to one will not impact the other. You can and will have them working together : swapping molecules between them, triggering each other’s inputs and so on. You will get a feel for it. With two circuits at your disposal and free choice of of where they start, go, and what they do, you have a lot of options within this 10×8 grid. It is a powerful toolkit to play around with, and Spacechem gets an admirable amount of mileage out of it. ‘Design some circuit that will take these inputs and produce this output’ is a surprisingly deep and compelling core for a game.

With this core of mechanics in place, Spacechem opens up its complexity and depth. New mechanics and constraints are introduced, some as one-off challenges, some as extra considerations for a few levels, some as permanent additions to your repertoire. This reactor circuit work is the heart of Spacechem, but the puzzling soon becomes multi-layered. Whilst you start off dealing with one reactor in isolation, with some given inputs and outputs, eventually you are given a series of inputs and asked to perform tasks that require you to wire up multiple different reactors and get them working together. You are not just solving one ‘take these inputs and spit out this output’ problem in one reactor – there’s a meta-puzzle. You have to work out some combination of reactors performing certain tasks to get from the inputs (from atmospheric pumps, or storage tanks) to the outputs (to be dropped into a reservoir, or supply freighter). You have to assign different tasks to different reactors, wire up their inputs and outputs in a way that makes sense, and then design the insides of each reactor to actually perform the role you want them to produce.

Reactor A might break the water down into two Hs and an O, and then send the hydrogen off to reactor B to do something, and the Oxygen off to reactor C to do something else. The two levels of the puzzling interact wonderfully. Your decisions at the meta-puzzle level end up defining the parameters of you individual reactor problems. A breakthrough in the meta-puzzle can make the individual reactor circuit designs a lot easier; conversely, if you manage to make a particularly fancy circuit, you might not need so many reactors, or to hook them up in such a complicated way.


Should you be scared off by all this talk of chemicals? No. The chemistry is best thought of as a theme or a setting. Clearly, the game is made by a guy who loves his chemicals – and that love is very obvious and infectious – but you certainly do not need any chemistry knowledge to play this. Don’t think that this is a game where you do chemistry: you do SpaceChemistry, which is even sexier.

The crux of the game’s genius is the way it has executed the idea of a design based puzzle game. It taps into a potential to create things that I didn’t know I had. Each level involves creating a working circuit to complete a particular task and (after being shown the ropes) it is left entirely up to you to decide how you want to solve each challenge.

There is a vast number of different possible circuits that will get you through a given level. Spacechem is not about you finding the answer, it is about you making your answer. Because there is such a wide range of possibilities, you do not feel constrained or hemmed in. That’s why it is a design based puzzle game, and this is how creativity comes into the picture. Make no mistake, this is a challenging, deep and complex puzzle game – when I finish a circuit and watch it run successfully, it is one of the most satisfying things I’ve ever achieved in a game. The solution I arrive at for each level feels much more uniquely mine than any game where there is only one prescribed path through. With no unique, right or optimal solution to a level, there is room for style and individuality, choosing one’s own approach. In contrast, in Braid, Portal or World of Goo, I solved things in much the same way as anyone else; I might have executed things slightly differently, but I will have been pursuing one of at most four or so different routes through.

I had two big fears about this game.

  1. I would never be able to produce some of the frightening looking circuits that I had seen in the previews, or anything of that calibre.
  2. The game would end up being a chore of optimisation – one where I would feel that my lack of prodigious mathematical and spatial analysis ability would be holding me back from playing the game properly, that the whole thing would give me a headache and feel a bit too much like work. Work that would much better be done by a computer.

On the first point, I shouldn’t have worried. I always find myself pleasantly surprised when I look back at some of the things I have produced. Some of them rather intimidate me. (Quintin Smith noted this very nicely in his RPS review.) No doubt you too will surprise yourself with what you end up creating.

It is not necessary to see the an entire design solution in your head at once: most of my progress has been made by taking the input molecule and seeing how I could make it a bit more like the required output one operation at a time. I would gradually build up the circuit in little stages like this. Work out how to unbond it first, then work out how to rotate it, then work out how to drop it over there, and so on. Step by step tentatively designing towards the goal, running the circuit for a bit, building the next stage as I could best see it, and so on. Every now and then I would tinker, go back and pursue a different line of thought or inspiration. Through no more complex process than this, I have been able to produce some rather impressive results. It is made very easy to tweak, play back, revise, tinker and shift your circuits around when you are making them, so you can locate faults, and see exactly how things are working. Things get a little trickier with multi-reactor problems, but not before you are ready.

SpaceChem has an excellent answer to the second worry, and it is worth looking at closely. Forget the idea that puzzles only have one solution – or one optimal solution – right now. Any design that can deliver the right amount of output molecules completes that level. That is a totally binary thing, success or failure, 1 or 0. Hundreds of designs can complete each level for you. With multi-reactor problems, where different configurations of reactors are encountered, the number of possible solutions to each level increases massively.  I will use two examples of my own to show how two radically different designs can complete the same production task.

Your solutions have the following statistics reported when you complete a level (and compared to a distribution of other poeple’s designs):
Number of cycles taken (how long your reactors take to produce the stuff).
Number of symbols used (how many components your circuits needed)
Number of reactors used (how many reactors you needed to have working to solve the problem)

In gameplay terms, these are meaningless. There is no in-game reward or benefit for doing it in one way or another. By the time you see these, the game has already opened up the next level for you. So, the game does not force you to optimise your solutions on these counts. But, if the graphs are there, might we not be tempted to slavishly optimise all our designs until we sucked the fun out of the game? SpaceChem cleverly ensures that we are unable to get too hung up on optimisation. How it does this is crucially important. Beyond a certain point, it simply is not possible to optimise your solutions; this is because almost all the time, each of the statistics you might want to optimise are in tension. You are prevented from chasing a ‘perfect’ circuit’ because no such circuit exists. Having a faster circuit often requires more symbols; using fewer reactors requires that the ones you have work harder and use more symbols, and perhaps take longer. Because there is a trade-off between these different statistics you cannot chase all of them to a maximum, so you don’t have to worry about the burden of doing so.

Sure, your design may have been relatively slow, but you probably used fewer symbols. That guy who used even fewer than you? He was probably slower. Looking at the statistics is entirely optional. The question is less about making the single perfect design, and more ‘can I make what I have done better by the standards that I think are important?’ How you want to refine or optimise your circuits is up to you, the game does not mind either way. I found that I created circuits that were slow, but used few resources, and enjoyed trying to make them as economical as possible. The narcissist in me loves the idea of having my own particular SpaceChem style, and unique portfolio of intimidating SpaceChem designs.

Because because efficiency over and above completing the objectives is optional and – in terms of gameplay consequences – meaningless, because there are no stakes, no rewards or incentives involved there is no external reason to pursue it. You do it -if you do it at all -for its own sake; it is rewarding in and of itself. With no pressure either way I found that I wanted to have a quick tinker, especially when I cast my eye over some of my earlier designs. Making playing with circuits enjoyable in this way is a master stroke of design. Not introducing aggregate scores is a crucial part of what makes SpaceChem fun to play. I have seen some suggestions about introducing this sort of thing, but I hope it is clear why this would strike at the core of the game’s Fun. Pressured optimisation would creep in, and I’d feel like a maths program should be playing the game instead of me.

Beyond the bare minimum of completing the output quota is the room to breathe, experiment, be creative and unique. That is the crucial space that SpaceChem has created. That is the genius of design puzzling in SpaceChem- it is a uniquely creative form of puzzling, especially given the multi-layered structure of many of SpaceChems levels. The parameters are kept open enough to allow circuit creation to be genuinely free and enjoyable for its own sake. In SpaceChem, this sort of puzzling is delivered in a lovely package and polished to a shine.

You will find yourself much more talented at playing the game by the time you reach some of the more tricky looking puzzles. Again, you should not be intimidated by this game. I got better at SpaceChem as I played it. I have a better feel for the sort of difficulties I come up against, and have a better grasp of how I can work round them. Whilst a lot of my play revolves around incremental trial and improvement, I do have occasional flashes of insight and eureka moments, and I’m able to read things a bit more sensibly in advance than I used to. I’ve dropped some odd assumptions I had since the opening tutorials. (This wasn’t entirely my fault: see the supplementary post). Going back to the earlier puzzles shows me just how much I’ve learned, and it is great fun to go back and remake them.

Beyond the superbly designed mechanics of its central puzzling, Spacechem does a lot more to draw you into the game. Short written interludes break up the puzzling and sketch a story with a gentle sort of charm and wit. The music is superb; I was certainly not expecting to find music in this game that I would happily cruise to in my car. As Quintin Smith highlighted on RPS, this music is very important for morale, strengthening your resolve, and salving any potential feelings of frustration. The art direction is fantastic, the GUI is excellent and the game is tight and precise like it should be. The circuit designing is very responsive and intuitive – it is very easy to follow up an idea without having a hard time translating it into action. Components clunk into place in a very satisfying way. Whilst I have one or two minor suggestions that I will post later, SpaceChem is superbly made.

Spacechem is a work of dazzling originality and almost ferocious intelligence; its concept is brilliantly realised and the whole thing sparkles with love and polish. It is a game of unexpected beauty, charm and humour; a celebration of creativity and experimentation. It is a game about designing your way through sophisticated multilayer puzzles. SpaceChem’s most significant achievement is the way it is able to foster creativity and experimentation without having to sacrifice any of its complexity or subtlety. It introduced me to a compelling new skillset and type of thinking; giving me the space to learn, experiment and play with its game system. It has left me surprised, delighted, excited and intimidated – both at the demands of some of the puzzles, and at what I have managed to produce in response to them. With SpaceChem, Zachtronics Industries have created a charming, superbly rich and uniquely creative puzzler. Its rightful place is alongside Braid and World of Goo. I wholeheartedly direct you to the demo.

http://store.zachtronicsindustries.com/product/spacechem

Quintin Smith did an excellent review of SpaceChem on Rock Paper Shotgun.

http://www.rockpapershotgun.com/2011/01/10/wot-i-think-spacechem/

Advertisements
  1. March 7, 2011 at 3:23 pm

    Minor Quibbles:
    This game has been excellently supported since release: I saw several patches in my first few weeks of play, mostly adding and tweaking minor things. I was going to ask for another introductory video to explain the mechanics a bit more, but that has been done. Still, I have one or two suggestions for those at Zachtronics. I’d like to see more use of Zachtronics forum space instead of reddit for discussing SpaceChem, simply because it is easier for people to find ideas and more professional.

    The SpaceChem digipack provides a nice resource for planning multi-reactor puzzles and I’d like to see something like that included in the game. That would help reduce the current problem of the meta-puzzle getting cluttered when you have multiple output notes all over the place. Perhaps some way of making them more compact, or a better way of hiding/showing them on the fly could also help.

    SpaceChem would benefit from an increased ability to save not just reactor designs, but to save reactor states, so you could experiment off in one direction, and then load a previous idea, etc. The same should go for the meta-puzzle, allowing you to re-load previous configurations and so on. Just keep making it easier to actually hone in on the juicy puzzling.

  1. No trackbacks yet.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: