The Stack

Here’s a game that Steam has been pushing relentlessly, including it in various bundles and promotions. It’s priced at $5, but it keeps on being put on sale at 75% off and the like. I guess the main reason is that this is a year-old remake of a ten-year-old game. It’s an extremely slick remake, though. Slick and cute.

So, what we have here is a puzzle-platformer, with puzzles driven by limited use of tools, Lemmings-style. (Not that the game resembles Lemmings in any other respect.) Tools include simple things like the ability to create bridges and blocks, and also exotic mechanics like “ghost traps” that create a hole in a brick floor when a ghost passes over it. Between the ghost trap and the freeze gun, which turns monsters into blocks that you can stand on, monsters in this game are often less enemies than opportunities to alter the level. There are complicated puzzles based around getting monsters to go where they’ll be useful.

The goal on every level is to collect a number of eggs lodged in inconvenient places. As soon as you have done so, the level ends. The nice thing about this sort of goal is that the ordering of the sub-goals can be an emergent property of the level design. For example, sometimes there’s a particular egg that has to be collected last because it’s in a place that you can only leave by using up tools needed elsewhere. I’m reminded of the general “kill all the monsters” goal in DROD, which had a similar effect on gameplay. In fact, I’d call the puzzle content here overall very DRODdish in feel, even though the mechanics are completely different. At the more advanced levels, this is a game about what the DROD community calls “lynchpins”: realizations about what sort of interactions are possible.

One thing worthy of special note is the “rewind” feature, which I assume wasn’t present in the original version from 2001. You can take your actions back continuously, kind of like in Braid, except that it’s not part of the puzzle content here; it’s just a convenience for the solver, and a pretty useful one for this kind of puzzle-solving. I remember that when Prince of Persia: The Sands of Time came out there was a review that said that the ability to rewind your mistakes was such a staggeringly obvious improvement that every game from now on was going to have to have that feature in order to be taken seriously. Well, not a lot of games took that to heart, but Toki Tori is one of them. Also of note is the way that rewinding distorts the screen in imitation of rewinding a VHS tape. Now, the graphical style of this game suggests it’s aimed at children. Is the sight of a videotape rewinding something that children understand these days? Has it entered the lexicon of things obsolete but iconic, like the scratch of a phonograph needle?

It’s been a while since I bought a new A-list title. I tend to wait for the major heavily-advertised games to be remaindered or even bundled when I have any interest in them at all, which is seldom the case these days: recent blockbusters seem to all be military-themed FPSes. When I hear people around the office talking about such things, it leaves me cold. Hearing them tiptoe around spoilers for Portal 2, on the other hand, just piqued my curiosity. For Portal 2 is a rare thing: a major heavily-advertised puzzle game. I don’t think I’ve seen a puzzle game advertised on bus hoardings since the first Professor Layton. And so, after resolutely ignoring the potato-themed ARG, I finally knuckled under and bought the thing last friday, played through the entire single-player story on Saturday, and on Sunday, instead of writing up the experience, got drawn into playing the two-player co-op mode, again completing it in a single marathon session. (So I’m posting this about a week late. Chalk it up to the difficulty of summarizing the total experience of something so recently well-covered elsewhere.)

Before I start talking plot, I have some general non-spoilery observations. Portal 2 is longer than its predecessor, more detailed, and wackier. Portal wasn’t particularly wacky. It had humor, but the humor was dry, and furthermore, superficial — by which I mean, one could imagine making an alternate version of Portal that plays it completely straight without altering the plot or gameplay at all. (Not that I’d recommend doing so. Much of the game’s charm is in its piquant blend of absurdity and living nightmare.) Portal 2, on the other hand, is more of a tall tale. It makes the ridiculous central to the plot, to the point where it starts to seem strange that this is set in the same universe as Half-Life. It puts me in mind of comic-book continuities, how John Constantine shares a world with the likes of Lobo and Ambush Bug. It seems to me that this shift of emphasis is risky. A light dusting of wit can enhance any game, but in scenes where comedy is the main focus, the game is only as good as it is funny. (I’ve cited MDK2 before as an example of how this can go wrong.) Fortunately, Valve got some pretty good voice-acting talent. I don’t know how much of Stephen Merchant’s lines were ad-libbed, but he has a way of making them sound ad-libbed even when they aren’t.

The puzzle content follows a typical pattern for puzzle games, steadily introducing new elements and exploring how they interact with what’s already been seen. (It’s what I think of as the DROD model.) The original Portal kind of did the same thing, introducing turrets and high-energy pellets one by one, and even doling out the portal gun in pieces, but that all seemed much more basic, like they could have introduced everything at once if they wanted to and they were spacing stuff out purely for the sake of spacing it out. The portal gun itself was the only real puzzle-enabling device, and everything else was just an environmental feature that provided material for portal-puzzles. Portal 2 often feels like it’s the other way around: that the portal gun is just a tool for executing gel-puzzles, laser-puzzles, etc. Crucially, some of the new elements are new means of transporting things or altering their trajectories: excursion funnels, light bridges, even repulsion gel at times, which can be both a means of transportation and a thing that needs to be transported. The original Portal had only one novel way to move objects around at a distance, and thus mainly focused on getting the player character around. A lot of the puzzles in Portal 2 involve moving objects around by novel means while you’re stuck standing on a button or something. In the co-op levels, the thing you’re transporting is often the other player, but the same principles apply.

Now to be more specific, and hence more spoilery. The game has three distinct runs of “test chambers”, bracketed and to some extent interrupted by behind-the-scenes stuff. The way that the game begins behind the scenes is a pretty big change from the enigmatic opening of the original. There, getting access to the areas outside the enumerated puzzle-game structure was the big twist, but here, it’s just part of the routine. (It reminds me just a little of Unreal, which is mainly structured around a series of building interiors punctuated by brief forays outdoors to get to the next building.) And once you have a routine, there’s a need to break it up with variety, even if it’s fake variety. Thus, reskinning! The middle run of test chambers is set in a long-forgotten section of Aperture Laboratories, implausibly deep below the surface, where we see what mad science testing environments were like in the 1940s and 1970s. This section is to the labs above what Red Alert is to Command & Conquer, replacing the gleaming engineered-looking Weighted Storage Cubes with simple wooden boxes, the glowing indicators with clack boards, and in general the futuristic high tech with precisely equivalent low — for example, the Aperture Science Unstationary Platform from the original, a levitating device that moved back and forth on some sort of energy beam, is replaced by something like a window-washer’s platform hanging from the ceiling by ropes. The very existence of low-tech equivalents underscores the tremendous wastefulness and impracticality of the whole operation. Company founder Cave Johnson, we learn, was in the habit of insisting on his own way against all sane advice, flew into rages at the least provocation (or sometimes none at all), and had enough power within the company that any half-baked idea he blurted out on a whim would be implemented at enormous expense. Even now that he’s gone, his legacy of preferring the complicated and inefficient remains.

Relics of Aperture’s past, along with recorded messages from Cave at various points in the company’s history, tell the story of its fall. Appropriately, this section of the game is precipitated by a literal fall down a shaft on the player’s part. The upper labs, on the other hand, starts off in a fallen state, decayed and overgrown, and it’s a rise up a different shaft, lined with electrical switches that are turned on by your passage, that triggers GLaDOS’s rise from the dead, followed by the gradual restoration of the facility to pristine condition.

GLaDOS herself is in much better condition than before her death, free from the audio glitches and lacunae found in the first game. Presumably such things were the result of the ethical constraint core that you destroyed at the end of the first game, or rather, of the self-sabotage GLaDOS engaged in to work around it. (Similarly, the dropping of the cake meme can be attributed to the destruction of her cake core.) She comes off as smarter, too, anthropomorphizing plainly inanimate things less ¹The Weighted Companion Cube is the famous example of specious anthropomorphization from the original Portal, but there were others, such as when she says that the hazard warnings around the High Energy Pellet installations had to be removed because they agitated the pellet. GLaDOS herself is, of course, technically an inanimate object, and seemed to regard human inability to empathize with inanimate objects like herself as a moral failing. and wasting no time on pathetically transparent attempts at deception. I suppose that’s because the time for that is over now that you’re openly enemies, but on a higher level, it’s because the role of humorously incompetent AI has been taken over by Wheatley, your sometime helper before the fall.

Of course, that’s not all Wheatley takes over. Wheatley’s conquest of the Enrichment Center — of Glados’s body, even — is the first moment that a male voice is in control, and things immediately take a turn for the worse — this is the point when both of the game’s strong female characters are literally cast down. For a while, Cave Johnson’s pre-recorded messages take over as antagonist, providing another male voice, but Johnson, as someone confident in his authority, is more of a bad father figure to match GLaDOS’s bad mother, while Wheatley is more like a spoiled kid with too much power. A spoiled pubescent kid, yet: the facility’s systems automatically give him a nagging urge to put humans through test chambers and a jolt of pleasure whenever you solve a puzzle, causing him to moan orgasmically. (I don’t think it’s a coincidence that this section immediately follows the discovery of tubes that spew viscous fluids, either.) This changes the tone of the exercise somewhat: GLaDOS hated you and wanted to murder you, but Wheatley effectively wants to rape you. The one thing that keeps this from being too horrible is that he’s so bad at it.

Finishing the remainder of the levels in a burst has left my hand worn-out. I suppose this is one of the areas where the touchscreen version is superior. You might think that the laptop trackpad I’ve been using isn’t too far removed from a touchscreen, as far as the hands go, but you use two things at different angles. Also, with a touchscreen, it would be impossible to lose track of where the cursor is, as happened to me occasionally. So I’d suggest playing this on a phone if you have any desire to play it at all. Too bad Steam doesn’t support such devices yet.

Overall, I have mixed feelings about this game. It does some good experimentation with types of challenge, but only on some of the levels. Plus, the experimental parts, like any true experiments, fail sometimes. For example, there was one level toward the end (the only no-time-pressure level that posed any real challenge) where you have to balance the bubbles in two pans of a scale — each bubble has a number inscribed on it indicating its weight. Your only way to alter the contents of the pans is of course by deleting chains, and bubbles you delete are replaced with new ones, weighing a different amount, after a substantial delay. Anyway, I never really solved this one: I was just getting the hang of thinking in terms of value differences between chains, when all of the sudden it solved itself. Some random replacement bubbles came in that just happened to match the weight of the other pan. I’d feel cheated, and want to try again and do it right, except for the fact that the puzzle wasn’t all that engaging, and doesn’t really have much of anything to do with the game’s core mechanic.

Probably my most positive experience was on the penultimate level, titled “The Mill”. Here, bubbles fall in batches into the buckets of a three-lobed whirligig, which spins slowly to spill what it holds. The goal is to delete 300 without losing 20. Frequently it’s impossible to nab the last crumbs in a batch. This is a time-pressure level, but it strikes me as having just the right degree of pressure: the mill rotates slowly enough that I wasn’t just frantically trying to pick things off as quickly as possible, I was thinking about optimization. Also, my own victory here was particularly dramatic. There came a point when I was very close to making quota, but the unchainable residue in a batch was going to put me over the limit when it fell. I hadn’t been paying much attention to the nearly-emptied buckets after I nearly-emptied them, so I wasn’t sure exactly when in the cycle they dumped their load, relative to when the new batch came. When the new batch did in fact come just in time for me to win before I lost, it felt like the cavalry had arrived in the nick of time.

Levels in this game end with a peculiar lack of fanfare. There’s always an on-screen display of your progress, but it’s small and doesn’t draw attention to itself, and finishing a time-pressure level usually requires enough concentration that I don’t go looking for it. (The Mill is one of the few that permitted that luxury.) So winning doesn’t involve a great sense of anticipation, and can feel abrupt: you’re in the groove, making chains fast enough to keep pace, and then things just stop. A little text message appears in the center informing you that you’ve unlocked the next level, and would you like to go there now? And that’s all the recognition you get of your accomplishment at that moment. Beating the last level at least takes you to the credits screen, which thanks you for playing.

The credits screen is interactive, by the way: the words “THE END”, made of movable letters, are in a bin of bubbles, which you can chain to your heart’s content. There’s no goal here, and anything you delete is replaced from the top, but you can at least do things like undermine the bubbles supporting the letters and make them fall over or out of place.

Overall, this is a very elegant game. I mean this both in the visual sense — it has a clean, simple aesthetic that I find quite attractive — and in the mathematical sense — this is a game that gets a lot of mileage out of very simple rules. I think I liked it more than I disliked it, if only because it did manage to develop its core mechanic in unanticipated directions.

I was starting to think I had underestimated this game. It looked like a match-3 with a gimmick, but it was actually turning out to be a puzzle game in the truest sense of the word: something where you have to figure out solutions. The randomization means that the solutions aren’t move-by-move precise like in a chess problem, of course. They’re tactical solutions, approaches that yield better results. But they’re still things that you have to figure out. For example, in the “Coathanger” level that I mentioned last post, the key things I figured out were: (a) it’s important to alternate sides so that the coathanger doesn’t start swinging too wildly and throwing things off, (b) it’s better to make chains on the inside, where there’s pressure keeping the bubbles from rolling away while you’re trying to link them, and (c) contrary to what you might think, it’s better to delete stuff on the side that’s currently on the upswing, because that way the loose bubbles roll away from the edge.

But then, the moment I decide that this game is richer than it appears, it starts repeating itself, and worse, starts giving me tactically simple levels, with stable architecture, where the only possible approach is just finding chains as quickly as possible. This isn’t the sort of challenge I want, so I finally give in and drop down to “Beginner” difficulty level to get by them.

It’s actually pretty hard to notice exactly what the difficulty level changes. Turning the difficulty down definitely makes things easier, but it doesn’t do it in the more obvious ways, like making the bubbles fall slower or reducing the number of bubble colors. I’m pretty sure it affects the maximum link-length — that is, the easier the setting, the farther apart the bubbles in a chain can be. This alone makes a tremendous difference: far too many times have I wasted precious moments trying to make a chain whose third link turns out to be just out of reach.

I’m not completely sure of this, because my attention while playing is mainly on pursuit of goals, but it seems to me that easy mode also makes the bubbles exert a slight magnetic attraction on one another, so that they cluster together more easily in nice tight packs. This is hard to judge because gravity tends to do the same to them, given enough time to work. But it’s definitely the sort of thing that would help, keeping bubbles in easily chainable formations instead of drifting apart. But it’s definitely a subtle effect. It makes me wonder a little about Medium difficulty. Am I receiving help there that I’m unaware of? If I switched to Hard now, would I start noticing it?

Another day spent in Chains. After spending a good long time on the Stream, and almost but not quite convincing myself to drop the difficulty down to Easy just so I could get past it, I finally had a breakthrough. My winning technique was a bottom-to-top approach: I start at the bottommost clog on the board and delete the largest, most space-filling chain I can see, then work my way upward, doing the same, creating pockets of emptiness that bubbles can fall through to create new chains for my next sweep. I still don’t know exactly how the game measures “flow”, but this approach at least gets things moving all over.

Having done that, I zoomed through the next six levels in short order. They continue in the same pattern of alternating levels with time pressure with levels without. I’m finding the levels without to be trivially easy — you’d think that the extra thinking time would allow for tricky puzzles, but I’m finding that it requires more thought to discover efficient tactics for the time-pressure levels. Perhaps this game was calibrated for a different sort of player than myself, but it seems unbalanced to me. Again I think of the choice to present it as a linear series of levels, rather than having them all unlocked from the beginning: the disparity of difficulty probably wouldn’t bother me so much if they were presented as mini-games selected from a list, or a Wario Ware-style grid.

So far, I’ve spent more time on that Stream level than on the rest of the game put together, but that could change: level 11, titled “Coathanger”, looks like another toughie. It pours bubbles constantly onto a two-sided platform that swings left and right, and asks you to prevent more than 30 bubbles from falling off. Previous challenges to keep stuff from falling off were more absolute about it: if anything left the screen, it was an immediate loss. On the Coathanger, the level designer knows that this is an unreasonable demand. A certain amount of slop is inevitable. And by allowing it, they gain the freedom to make the level that much harder.

Too weary to launch into a major game on getting home from work, I instead look for something simple and casual that I can play on a laptop on the bus. I think I got Chains as part of the same Steam indie pack as Obulis, and, like Obulis, it seems designed for playing on phones — although I think I vaguely remember playing a PC demo for it some time back, or at least for something very similar, involving colored bubbles of varying size.

The basic mechanic is that bubbles fall from the top of the screen, and the player can delete them by tracing a chain — a path that joins together three or more like-colored bubbles. The chain can zigzag arbitrarily, but each link mush be sufficiently close to the last. I suppose it’s in the general family of “match-3” games like Bejeweled, but it differs from most in that it’s continuous rather than grid-based. I suppose Zuma fits that description as well, but there’s a little more physics going on here. Bubbles have weight and momentum, and bounce off each other somewhat.

Given the above mechanic, what does the game do with it? One can imagine various different goals and rules that work with deleting chains, and rather than choose one such, the developers here apparently chose to implement everything they could think of. There are 20 levels, and rather than just varying the board layout, each level is effectively a different game. When you complete a level and unlock the next, it asks you if you want to go on or stay on the same level. This would be a strange thing to ask in most level-based games, but here? Here it’s more like asking “Do you want to keep on playing Bejeweled or would you like to switch to Tetris instead?” I feel a little like structuring the game as a sequence of levels is unnecessary, that it would be better to just have them all unlocked from the beginning so that you can just choose whatever game you feel like from the main menu.

Incedentally, I choose the particulars of that simile — Bejeweled and Tetris — because those are more or less the paradigms for the first two levels. In the first, the bubbles are all the same size and contained in vertical columns, forcing a grid-like formation, and additional bubbles fall in only to replace the ones you delete. There’s no time pressure here, and no real risk of failure: you just keep making chains until you’ve deleted 100 bubbles, at which point you win. (At one point, I thought I had run out of possible moves, but that was because I hadn’t yet realized that you could connect chains across the solid-looking column walls.) Level 2 isn’t nearly as Tetris-like as level 1 is Bejeweled-like, but it shares the fundamentals: things piling up inexorably over time, with the player frantically trying to delete stuff before it piles up too far. The model here is that bubbles are constantly falling into a bin with a hinged bottom, held shut by a pulley and counterweight. If the weight of the bubbles exceeds that of the counterweight, the bubbles start to leak out the bottom and you lose. Which seems like a rather fancy way to say “don’t let things pile up too much in the bin”, but if you’ve already got the physics, why not?

Level 3 is a quick one: it gives you several columns of bubbles in different sizes and point values (the value being determined by the size), and asks you to make a chain with a specific value. You might need to delete some stuff to make this possible, in which case, as in level 1, more bubbles will fall from the top to replace the ones deleted, but it’s basically a simple exercise in addition, similar to making change. The level after that is another time-pressure one, and difficult enough that I haven’t got past it yet. In a way, it’s the opposite of level 2: instead of keeping the bubbles on the screen, you’re told to “keep the stream flowing for five minutes”. The “stream” in question is a gently-curving frame with a couple of small barriers suspended in the middle to create blockages, which obviously you have to clear. The big difficulty with this is the ambiguity about what it means by “flowing”. Does it judge it by the rate at which things are entering from the top? The rate at which they’re leaving the bottom? The average downward velocity among all the bubbles? These are different goals, that require different approaches. There’s a (subtle and easy-to-miss) meter on the screen giving your current rate of flow, and you lose when it reports a flow rate of 0, but it seems to lag behind events a little; sometimes I lose at a moment when I’ve just cleared a major blockage and things are moving freely past both barriers again.

So far, both of the time-pressure levels are much more difficult than either of the others. What’s more, I don’t think the time-pressure levels are as well-suited for my purposes. Part of the appeal of casual games is the idea that you can pick them up and put them down whenever you feel like it, with no significant loss of state. The Stream level here requires you to play it continuously for five minutes. Sure, that’s not a large time commitment, but if I’m playing while I’m waiting for the bus, and the bus arrives four minutes into the level, I feel a little put out. Sure, you can pause the game, but that interferes with flow.

And Yet It Moves consists of three chapters, an interactive credits sequence, and a bonus level. The first chapter is set in a cave, the second in a forest, and the third, after starting in the forest, goes all trippy and stops pretending to be representational. The scraps that form the world suddenly take on a brightly-colored pattern like wrapping paper, and the set-pieces become more elaborate and more gameish. None of the game tries to be particularly realistic, but here at the end, the designer seems to feel freer to just do whatever he finds interesting.

Objects grow and shrink, or have textures that move completely independently of their real motion. Some areas rotate continuously on their own — compensating for this with only 90-degree turns is difficult enough that it seems like these bits in particular have to be easier on the Wii. There’s a repeated gimmick of colored platforms that appear and disappear in time with the background music. The background music doesn’t usually have a very strong beat, but for these segments, it changes. I really don’t care for the music in this game — it consists mainly of random Seinfeld-style mouth-pops and samples of someone saying “Doong” — but in the these segments, it becomes more coherent, and thus more tolerable. In fact, it reminds me of the music sections in some of the Rayman games.

And there’s a motif, used once per level towards the end, of doubling the player character. You hit a checkpoint in what looks like a dead end — it should be noted that the checkpoints look like sketch-people similar to the player avatar, who stand still and point in the direction you should go next, like a guide — and suddenly the world changes into an enclosed space with two such guides, one color-inverted, white-on-black instead of black-on-white. Another sketch-person stands there, and you’re in control of both, but they move in opposite directions. The only way to continue is to get them both to their opposite guides at once. It’s reminiscent of Scott Kim’s Double Maze, except taking place in a single space.

At the very end — and into the credits and bonus level — the color drains from the world, leaving it unmarked white, with occasional crumples and creases. It’s sort of a larger-scale version of what happens at the end of every level: your sketch-guy finds a white space with a black silhouette in the shape of himself and fits himself into it, restoring it to its pristine condition. Unusually for a platformer, the game doesn’t even address the question of why the player character wants to do this. You could interpret the whole thing as a metaphor for transcending the world of appearances (the photographs and other markings) and achieving awareness of the world as it is, which in this game means just paper. Except of course that the papery appearance is itself artifice. It’s all just bits. When the image on a “scrap” moves independently of its edges, it makes it clear that these aren’t even digitized versions of things that ever even existed as scraps in the physical world.

I can’t really back this up, but I get the impression that And Yet It Moves is best-known as a Wiiware title, even though it was released for Mac and PC first. I suppose that’s just the nature of the market right now. Even ignoring the popularity of the Wii, Wiiware is an effective tool for making games visible to people who wouldn’t be exposed to them otherwise. But also, even though I haven’t tried the Wii version, it sounds like a better game. I mentioned that there are rotating-world games that give you continuous rotation, rather than the four-sided stuff I’m seeing here. The Wii version of AYIM has that, with multiple ways of accessing it from the controls. I have to wonder if the puzzle content had to be redesigned at all to accommodate continuous rotation or if it was just left alone, and if the latter, whether it makes alternate approaches possible.

One description I’ve read says that continuous rotation makes things more difficult, but it almost has to be easier to at least do what you intend most of the time. With a keyboard, you have the left hand on WASD and your right hand on the arrow keys, although only three keys of each set are used: A and D to move left and right, W to jump, Left Arrow to rotate the world counterclockwise 90 degrees, Right Arrow to rotate clockwise, and Up Arrow to do a 180-degree flip. The problem with this is that the directions of rotation aren’t very strongly associated with left and right. Half the time, I wind up pressing the wrong thing — which, given that the world takes a little time to rotate, and doesn’t freeze while it’s rotating, can be enough to kill me or otherwise make me start over whatever I was trying to do. (Checkpoints are plentiful, at least.) At first, I tried to remember that the left/right arrow keys indicate the direction the top of the screen moves in, but this is a difficult thing to apply in the heat of action. After a while, I instead tried thinking of it as pressing the key corresponding to the direction that I want to become down — a rule that happily applies to the up arrow as well. I think this is a little easier to apply, but I still wind up making a lot of mistakes.

The times when I’m least likely to make mistakes are the more intense stretches, when I’m rotating the playfield a lot. I don’t even think about it in terms of absolute directions then: I just know that I have to rotate the world opposite to my last rotation, or in the same direction again, and that’s an easy thing to communicate to my fingers. The game seems to be making this kind of quick flurry or rotation more and more necessary as the game goes on, replacing the more conventional platforming, which could have the ironic effect of making things easier for me.

Another update and suddenly And Yet It Moves is working for me. This is a 2D puzzle-platformer that, like, Braid, is based around building puzzles around one unusual ability. In Braid, it was control of time. Here it’s control of gravity — or, equivalently, the ability to rotate the entire world. I’m told that there have been other games since that explore this idea more thoroughly — including things where you can rotate the world freely by any angle, instead of just in 90-degree increments as is the case here. There’s a whole mini-genre, apparently. There are also antecedents, like the Shift series, which lets you simultaneously flip the world upside-down and reverse figure and ground.

The one thing that really distinguishes AYIM from the likes of Shift is that your rotations affect more than just the player avatar. Boulders tumble from their now-horizontal holes, falling water drops do sharp mid-air turns, bats are dislodged form their perches and fly up to the new ceiling. There are bits where the focus is entirely on making some inanimate object fall the right way, although you still have to make sure that the avatar doesn’t fall too far and die in the process. Still, the most-repeated material is all about gravity-control-enhanced navigation: jumping off a cliff and then quickly turning the cliff wall into a floor, for example, or extending the length of your leap by falling part of the way.

I’m a bit disappointed about how little of the levels I can see at once. Surely the re-orienting of the world would be more impressive if I could see the world? But then, there may not be much of a coherent world to see, the levels being patched together out of bits that only make sense locally. Certainly they’ve picked a graphical style that suits such a design. This is a world of collage, made of ragged scraps torn from photographs. The really interesting thing is that the pictures in the scraps sometimes waver relative to their frames, or lag behind their movement a little, suggesting that the scraps are windows, or pieces torn out of windows.

Rushing through this game is probably not the right way to play it, but that’s what I’ve done, just to free it from the Stack. Since it’s the only game of its kind, it’s reasonably likely that I’ll go back to it at some point, to try to perfect my performance. There’s an Achievement called “Perfection” for 100% completion, and another called “Nirvana” described as “Learn to let go after attaining Perfection”, which I assume means getting 100% completion and then wiping your progress. Although it would be kind of funny, in a cruel way, if it didn’t mean that.

Not that we know for sure right now. Looking at the global stats for Zen Bound 2, it seems that no one has actually achieved Perfection. Or at least, few enough that it rounds down to 0.0% of players. Even the achievement for getting a minimal passing grade on all the levels has been claimed by a mere 0.9%, which is oddly low for something so easy. As a point of comparison, 13.9% of players have got the “Unwinder” achievement, for using up all the rope and then reeling every inch of it back. There is no reason to do this other than for the sake of the achievement. If you want to start a puzzle over, you can just reset it. So 13.9% of the players are achievement-mongers, but only about 6.5% of those people thought it worthwhile to grab the much rarer achievement for finishing the game. What gives?

I don’t think it’s likely that anyone gives up on this game because they’re stuck and unable to make progress, so people must give up on it because it isn’t to their liking. Well, okay, it’s an oddball game. To me, its mere oddballness is appealing, but the reason we see so many games rehash the same ideas is that those are the ideas that a lot of people like. But I think I can identify another reason why people wouldn’t like it: it is a fundamentally awkward game.

I know I spoke in praise of its 3D rotation UI, but that’s just half the story. You control the object pretty perfectly, but you don’t control the rope, or at least not directly. You control the rope by catching it on the object. You spend a lot of time trying to nudge it off corners and into crevices, or angle it so that it doesn’t come off said corners. Sometimes you accidentally tie it off before you intend to. Very often you can’t quite see what you’re doing because the free end of the rope is on the underside. And this is a crucial part of the game’s challenge. The awkwardness of the rope isn’t a flaw in the game, it is the game. And it’s impossible to play without thinking about how much easier this would be in real life, where you could just take the object in one hand and the rope in the other for greater control.

Contrast this to more mainstream games, which often go out of their way to present a fantasy of being more competent than you are in real life. I tend to downplay the role of the fantasy element in games, but I’m not just talking about the fiction, I’m talking about the interface. First-person shooters have auto-aiming. Brawlers execute complicated combos with the press of a few buttons. Guitar Hero simplifies its riffs.

And in the end, what’s the result? I don’t think there can be much disagreement that you’re making the sculptures uglier with your efforts. They start off rather elegant, they end up ruined by a clumsy tangle and, unless you managed 100% coverage, a splotchy and incomplete paint job. The game does its best to pretend that it’s pretty, but it’s like a parent’s praise of an attempt at art by a toddler, which is frankly what your efforts resemble. And that’s ignoring the often-unwelcome symbolism. Many of the sculptures are representational, so there’s a sense that you’re putting animals and people into snares, making their condition worse. I suppose the game is kinder to its subjects than a more conventional game, which would just have you kill them. Still, as in Shadow of the Colossus, your goal is to take things of beauty and ruin them. But at least SotC made it difficult. Here, you wind up with an ugly mess regardless of whether you pass the level or not. Because of your clumsy awkwardness.