The Stack

It turns out that Obulis makes it completely clear when it thinks you’ve won, announcing it explicitly in a pop-up and then playing a credits sequence. This is a bit of a relief, because I wasn’t entirely clear on which levels were necessary for winning and which were optional bonus levels.

The game is organized into three chapters, each of which contains multiple “paths”, or sequences of themed levels with their own graphical style, background music, and, in some cases, path-specific gameplay elements: Clock Tower levels have steadily-rotating gears that can carry marbles in their teeth, Pond levels are built out of flowering plants that are a little springy and never completely straight, etc. The maps show the individual levels as squares in a sort of board-game layout that visually differentiates between the levels you haven’t reached yet, the levels you’ve completed (which are still available for replay), and, most importantly, the levels you have access to but haven’t completed, which have a jagged nimbus so they really stand out. Paths sometimes branch a little, so that completing one path can open up access to two others. It’s all pretty typical of the superstructure in modern puzzle-games, particularly puzzle games that imagine themselves to be casual. (I personally have doubts whether a puzzle game with designed solutions can ever really fit the “casual” mold, but that takes us to the question of “What is a casual game?”, and that’s not what I want to talk about today.)

And occasionally the paths branch into optional bonus levels. Some of these levels are harder variations on levels in the path proper. These are easily identifiable by name: every level has a name consisting of its path name and a roman numeral, like “Dungeon III” or “Windmill VI”, and the harder variants just append the word “Ultra” to the end of that. Ultra levels have the same architecture as the levels they’re based on, and usually have a similar layout of balls, but are altered enough to require completely different solutions. I kind of assumed that these were optional, because they just seemed like extra content for people who want more challenges. (DROD, to name one precedent, has a few rooms that are harder variations on other rooms, and they’re invariably optional.) But there was another sort that I wasn’t at all sure about: the Inner Sanctum levels. This is a “path” consisting entirely of single levels reachable from other paths. The unique presentation for these levels is inky silhouettes; the gameplay gimmick is that the silhouettes have paths through them that you can’t see, and have to find by experimentation; the background music, rather incongruously, consists of traffic noises. And each Inner Sanctum level you complete gives you a piece of a special artifact for that map. Each artifact, when completed, unlocks some of the Ultra levels — or were the Ultra levels unlocked by collecting the medallions at the end of the normal paths? I’ve forgotten already. This stuff isn’t where your attention falls while playing.

At any rate, because I didn’t look at the manual until after winning the game, I wasn’t sure if the Inner Sanctum levels were optional or not. It turns out that they are. I went back and completed them all after winning anyway, and in the process discovered that the game is rather insistent about playing the credits sequence, which isn’t interruptible, every time you complete a level after winning. Ah well. I probably won’t finish all the Ultras: they tend to be challenges of the “get things to collide in midair in just the right way” sort, rather than the “figure out the clever tick that makes it easy” sort. If I even make the attempt, it’ll be for the Achievements. Every single Ultra level has a Steam achievement associated with it.

I don’t seem to have any big-budget titles for 2008. Just as well — I tend to think of the last few years as typified by the triumph of the indie game. Obulis is a 2D physics-puzzler of a general sort that I mostly see for free on Flash games sites. But in fact it was originally designed for mobile devices, and consequently has a simplicity of interaction well-suited for touch-screens. The goal is to get colored marbles into their matching color-coded receptacles. Some of the balls dangle from chains (or, in some cases, rigid rods). You can select chains to cut them, allowing the balls to drop. And that’s all you can do. That’s the sole means of interacting with the game.

In the simplest of the puzzles, you basically just decide on what order to cut the chains in, allowing the balls to come to a full stop before releasing any more. More advanced levels require you to cut chains while the balls are still in motion — for example, there might be a ball that drops onto a flat surface, where it needs a nudge from another ball to start moving, and the only way to apply that nudge in the correct direction is to release it in the interval between another ball rolling up a nearby slope and rolling back down again. The trickiest puzzles, the ones that I find myself having to restart over and over, are the ones that require collisions. One of the repeated patterns is to have a ball suspended by two chains in a V configuration, so that cutting either chain starts it swinging left and right, after which the exact moment that you cut the other chain makes a huge difference to its trajectory. So, it’s not just puzzle-solving in the sense of figuring out a solution, but involves precise timing as well, with small differences having large effects, the multiplier effect of an angle over distance. Solving levels like this involves more fidgeting and adjustment than problem-solving skills. Sometimes winning a level looks miraculous, the bodies in motion moving past or into each other perfectly. But then, the levels are all set up to make the solutions possible. Just knowing this, that everything was placed by the designer for a reason, helps a lot in figuring the puzzles out.

Knowing the physics helps, too. Collisions between balls here are perfectly elastic, like in a Newton’s Cradle: when a moving ball hits an unmoving ball on a horizontal surface, it’ll transfer all of its kinetic energy, coming to a full stop — but in a slightly different place. Slight differences of placement can be important if you’re about to drop another ball on top of that spot.

After less than a day, I find myself more than halfway through the game, and with more than half of its Steam trophies. I suspect that I’ll either finish this game very soon or not at all.

The basic injustice of this blog is that the amount of attention a game receives is proportional to the time it takes me to finish it, not how significant or interesting it is. In fact, the less interesting a game is, the more likely I am to play it in short sessions, dragging it out. Thus, a sprawling mediocrity like The Humans gets two full weeks of posts, while its fellow puzzle-platformer Braid gets only one. But I have vague plans to return to Braid someday, to give it amore thorough analysis, whereas I’m pretty sure that this is my last post about The Humans.

Towards the end, the Witch Doctor pretty much disappears, presumably because it makes things too easy if you can have a rope at the very beginning. The rope is definitely the one tool that has the biggest impact on where you can go in this game. Most challenges can be reduced to alternately stacking Humans, and using the rope to haul everyone up to the platform that the stack let you reach (or, conversely, using the rope to let everyone down to a lower platform, then forming a stack to get the rope bearer down.) The only thing wrong with this approach is that it’s so time-consuming, and many of the later levels attempt to create difficulty by making their time limits low.

In fact, I found the hardest level in the game to be level 159, the second-to-last one, simply because it had such a tight time limit. The last level has a limit of 9:59, the maximum the game is capable of displaying, and seems to be included more or less to close things off with a nice pleasant experience, a reward to the player for coming that far. 159’s limit, even on Easy difficulty, strains things enough that I honestly wonder if it’s a mistake. There’s an obvious route to victory, and I suppose it might be possible to execute it in time given lots of practice and clever optimizations, but this would make it take time and effort way out of proportion to the rest of the game. Instead, I hit on the gimmick of leaping halfway into a bush from a platform slightly higher than it; once there, I could walk the rest of the way through it, which spared me the time I would have needed to fetch a torch. There was an earlier level that involved jumping over a bush in a similar way, but there, I cleared it completely. The messiness of winding up embedded in an obstacle makes it seem like a glitch, but who knows? It could just be the game’s last mean trick.

And yes, the needs-more-QA moments continued to the end. There was a level that exhorted me to “find the idol”, but which instead used the queen as a reward. There was a level that ended with the pet-dinosaur cutscene, despite not having any pet dinosaur in it — this was the only cutscene I had seen in the Jurassic levels, so it seems like the mistake is that it’s there at all, rather than that it’s the wrong one. Also, there’s one major glitch I failed to mention in my post on glitches: on Swamp-themed levels, where the floors are uneven, one sometimes falls through them when jumping. I wonder if the Amiga version had these problems too, or if it’s all a matter of hasty porting.

Speaking of cutscenes, there is in fact an outro scene, despite my doubts. Predictably, it riffs on 2001. And, having seen it, I’m well and truly finished with this game. It’ll be a while before I get it out of my head, though; I’ve played enough of it this week that it’s started haunting my dreams. I’ll have to try to choose as my next game something capable of muscling it out of my head.

Christ on the crapper, he thought. Africa. For the ghosts of dead tribes. Wiped out to make a land of — what? Who knew? Maybe even the master architects in Berlin did not know. Bunch of automatons, building and toiling away. Building? Grinding down. Ogres out of a paleontology exhibit, at their task of making a cup from an enemy’s skull, the whole family industriously scooping out the contents — the raw brains — first, to eat. Then useful utensils of men’s leg bones. Thrifty, to think not only of eating the people you did not like, but eating them out of their own skull. The first technicians! Prehistoric man in a sterile white lab coat in some Berlin university lab, experimenting with uses to which other people’s skull, skin, ears, fat could be put to. Ja, Herr Doktor. A new use for the big toe; see, one can adapt the joint for a quick-acting cigarette lighter mechanism. Now, if only Herr Krupp can produce it in quantity…

— Philip K. Dick, The Man in the High Castle

I said before that, past a certain point, the puzzles in The Humans amount to a straightforward application of principles already learned. This isn’t completely true. One thing muddies the waters: lack of knowledge of the levels. You only see a screenful of the level at a time, centered around the human you’re currently controlling. Unless there’s something I’ve missed — which could easily be the case, given what else the docs have failed to disclose about the UI — the only way to scroll the view is to move. Thus, it’s frequently the case that I have a choice of available paths and no idea which to tackle first. Very likely one will hold the key to progress on the other — a pressure plate, a torch to burn a bush, an extra spear to kill a dinosaur — but both will require a significant investment of time to explore, requiring stack-building or repeated spear-passing. There have even been cases where there’s a torch in full view, tempting the player to waste time retrieving it, that turns out to have no use on the level.

Thus, the levels take on aspects of maze. Once you have a rope, you can pretty much always move as a team. But sometimes it’s worthwhile to just scout ahead with one guy. Indeed, some levels do this for you, placing humans at various places, places that make them useless physically, unable to leave the platform they’re on until the others bring a spear or a rope, but useful for the intelligence they bring about their position. It’s a little like a scout unit in a RTS, dispelling the fog of war.

Another thing: When you switch control between humans, the view glides smoothly to the new one. This lets you see everything between the two. There are times when I’ve deliberately taken advantage of this.

I’ve just been through level 117 — that is, level 37 of the expansion. Thanks to that Wikipedia article, I had advance warning that this level was bugged and impossible to complete. I didn’t really trust the article, though, and had to give it a try for myself. It is in fact possible to reach the red floor tile on this level — in fact, it’s possible to reach it much more efficiently than the designer probably intended. There’s an obvious route that goes clockwise around the entire level, picking up an extra human and killing a dinosaur along the way, but once you have the rope, you can take a shortcut straight from the start point to the exit. I’ve been encountering a lot of that lately. It doesn’t do much good here, though, because the exit doesn’t work. Even when you have multiple humans standing right on it, the level doesn’t end. So I’ll take what I’ve done as good enough and skip ahead to 118 with the level password.

The reason I didn’t trust Wikipedia on this matter is that I had independently seen someone on a web forum (which I can’t find now) complaining that level 39 of the original levels was impossible, which it wasn’t. It was probably trickier than intended, though. The accusation was that you couldn’t make a particular jump on the loathsome wheel, and indeed I was unable to do so — but with a little luck, I managed to work around it. This is not the only place where wheel jumps seem to be more difficult than intended. I suspect the increased framerate is to blame, but it could just be a glitch.

It wouldn’t be the only one. I’ve repeatedly had trouble riding pterodactyls: a lot of them seem to be flying just a little too low to allow you to step off onto the platform they’re obviously taking you toward. In some cases this may be deliberate: the platform at the end of the pterodactyl’s route is a red herring and you’re really meant to step off and drop onto a different platform in the middle of the route. In other cases that’s definitely not the answer. I’ve found that it’s actually possible to make stacks on the top of a pterodactyl’s back, allowing the guy on top to step off onto a higher platform — it’s a weird thing to do, but when you come down to it, a pterodactyl is just a kind of moving platform in this game. And I’ve seen one level where it seemed to be actually necessary to do this. But the consequences are blatantly bugged: whenever I try to unstack humans on a pterodactyl’s back, they vanish and go somewhere they’re not supposed to be — offscreen somewhere, or in the middle of a wall or something. Often they immediately die, which isn’t so bad: as long as I still have humans in my pool, I’ll just get a new one standing on the last stable ground that the dead guy touched.

The pterodactyl-unstacking bug seems like a serious problem. I don’t see any other mention of it on the web — not that this is an easy title to google for — so it could be specific to fast systems, or to DOSBox. But I can believe that it’s mere carelessness on the part of the developers. If I understand corerctly, the game was developed primarily for the Amiga, so the DOS port may not have received the same attention. But then, the way that so many of the puzzles can be short-circuited suggests a lack of attention in all versions, as does a certain comment found in the startup sequence of the Amiga CD release:

the game to me is fucking crap – and i’m doing humans 2 now as well and guess what???
its FUCKING CRAP TOO!!!!!

The thing is, the mistakes that make the game harder and the ones that make it easier kind of cancel out. I’m not exactly playing the game that the designers had in mind, but sometimes it seems like I’m playing a more interesting one.

I’ve already described the tools available in The Humans. I haven’t gone into detail about what you’re ultimately trying to accomplish with those tools. That’s because there isn’t a lot of detail to go into. The overwhelming majority of the levels are all about reaching a given point on the map. Usually it’s just a spot marked with a red floor. Sometimes they dress it up a little and make the goal to rescue something: another human for your pool, a pet baby dinosaur, a golden idol, a captive queen. (So female humans do exist!) Only the first of these makes a difference to gameplay. The earliest levels made a goal of obtaining the tools for the first time: “Discover the spear”, etc., as if technological progress were a matter of picking up ready-made items.

In the first half of the game — that is, in the 80 levels of the original floppy-based release — all of these special goals have cutscenes associated with them. When you rescue a baby dinosaur, for example, you get a brief full-screen low-framerate cartoon of the dinosaur leaping onto its rescuer, knocking him flat, and licking his face like a dog. These cartoons are completely horrible — predictable slapstick without any sense of timing. And they loop, which gives them the tiresome aspect of a repeated punch line, a joke retold by someone who doesn’t understand why we didn’t laugh the first time. We were probably supposed to be impressed with the fact that they were doing full-screen animations at all, given the technology of the day, but I remember thinking they were horrible at the time as well.

It should be noted that, unlike in Lemmings, all you need to do to pass a level is get one human to the goal. Sometimes the red-floor goals are flanked by special flowering bushes that give you bonus points if you park a human in front of them before claiming victory, but I haven’t been paying attention to my score. Occasionally I’ve gone for the bonus as an additional challenge, but usually it isn’t challenging enough to make this worthwhile. If you can get one human to the goal, you can get others there by the same means, and so, like the stereotypical mathematician, I consider the problem solved and move on.

There’s one sot of goal that’s different: killing a dinosaur. You kill dinosaurs by throwing spears at them, and the number of spears required varies from one to three. Only the hardiest dinosaurs are used as level goals, so these levels amount to “Find three spears and bring them back to this point”. Except that you don’t necessarily have to bring them back to the same place: spears can be thrown from either side, or even dropped onto the dinosaur from directly above. Complicating matters, you don’t necessarily know that this is the goal. When you see a red floor or a golden idol, you know that’s the goal, but when you see a dinosaur, it might just be an obstacle, or even a red herring. (The same goes for captive humans, although rescuing them is pretty much always beneficial.) Also, recall that spears are necessary for jumping gaps, so throwing them away isn’t always a good idea. With the other goals, you pretty much go for them as soon as you’re able, but with spear-throwing, one holds off.

So basically what I’m saying is that the dinosaur-hunt levels are the most interesting ones, and if one were for some reason going to make more levels for the game, one would be well-advised to use more dinosaurs. Which may go without saying. What game wouldn’t be improved by adding more dinosaurs?

I’ve once again taken to playing on a laptop during my excessively long commute. This is something I was doing for previous games this year, despite the inconvenience it made for mapping, but I had stopped with The Humans for three reasons. First, there was the installer, which, as you may remember, gave me problems. I couldn’t run it inside DOSBox, and that meant I couldn’t run it on my Macbook. But this was easily solved by installing it on my desktop machine and copying over the directory it was installed to — since it’s a DOS game, it doesn’t have registry entries or shared DLLs complicating matters.

Secondly, there was the key disk copy protection. This wasn’t a serious problem. The game recognized the CD in my laptop just fine. I just don’t like using CDs in my laptop. My reasons are minor and irrational: concern for battery life when moving parts are involved, fear of something untoward happening when the bus jostles the machine while it’s reading, distaste at the sound it makes. This was solved by copying the CD to my hard drive. (Which also has moving parts and is probably more vulnerable to jostling than the CD drive, but is also easier to forget about.) I couldn’t make an ISO out of the disc with standard tools, presumably because it has some kind of copy-protection bit set. I could probably find a more hackerly tool, but in fact it was sufficient to copy the files off and make a fresh disk image with the same volume name as the CD.

Thirdly, there was the sound. The sound works fine, but the bus is not the ideal place for it. Well, actually there are no sound effects in this game, so sound is not crucial. The only audio content is the background music, which I do think is an important part of the experience of the game. But after a week, I think I’ve heard enough of the background music. Even if the game is silent, I can hear it in my head.

So now that my installation is perfectly bussable, how is the experience? The graphics certainly don’t lose anything — if anything, they look better on my laptop’s screen, smaller size making the pixels finer — and, lacking mouse support and being essentially designed for a digital gamepad, the controls are basically unchanged. The one inconvenience it poses is that it’s realtime. It only occasionally requires any precision of timing on the player’s part (those occasions being the bits where you have to leap onto flying pterodactyls or make a jump on the loathsome wheel), but the clock is always ticking down. My commute is full of micro-interruptions where I look out the window to see how close we are to my stop (or even just to enjoy the scenery), and the typical level takes between five and ten minutes to play to completion, or even to fail. Really, it seems to me that the ideal commuting game is turn-based. It makes me wonder how Tetris on the Gameboy became such a hit.

But then, in a way, it’s the ideal kind of game for a bus, or a waiting room, or any other situation where there’s nothing else to do, but you don’t want to get so involved that the interruption at the end will leave you unsatisfied. Once you’re past the point where the designers run out of new tricks, gameplay is very methodical and by-the-numbers. It engages the higher brain just enough to distract, and not enough to enthrall.

Past level 100 now. Things are getting pretty repetitive. It looks like there are only so many tricks that the toolset here supports, or at least only so many that the level designers found, and they ran out a while ago. On most levels, there are some obvious sub-goals, such as getting to where the torch or rope is, or picking up an extra spear for the dinosaur blocking the way. Once these sub-goals are identified, all you need to do is find a way past the obstacles to achieving them, which may involve more sub-goals.

The only thing that interferes with this pattern is red herrings. Sometimes there are bushes that don’t need to be burned, dinosaurs that don’t need to be killed. In fact, there was one level where I spent a while trying to puzzle out a sequence that would make things possible, but it seemed like every crucial tool was locked away behind another tool’s obstacle in a kind of deadlock. It seemed completely impossible until I noticed that I in fact had enough humans at my disposal on that level to build a stack tall enough to reach the level’s goal without any of the tools.

The thing is, I’m not convinced that this is what the designers had in mind. It’s entirely possible that they intended some more conventional solution, albeit one that depended on a trick I hadn’t figured out yet, and that they simply failed to notice that their cleverness could be short-circuited. It’s hard to tell with the sort of puzzles found in this game. It’s not like a cryptic crossword, where each answer is self-evident when found. Perhaps the game only seems so repetitive to me because I’ve found general ways to avoid the clever solutions. I doubt that’s actually the case, but some lesser version could be.

I started this weekend somewhat over halfway to level 80 in The Humans. New tricks were coming less frequently. The only unknown functionality I discovered lately was that you can change the angle at which you throw something by pressing the left/right buttons between selecting the “Throw” action and actually throwing it. This isn’t mentioned in the manual at all; I checked. But you pretty much have to figure it out on level 51. Aside from that, there was one new non-obvious technique to pick up: getting an entire group of humans down from a platform by using a rope, then building a stack under the guy holding the rope to get him down with everyone else rather than leave him stranded. Intuitively, one thinks of the human-stack as a means of ascending, not descending, so there’s a bit of a mental trap there.

So most of the second forty levels were a straightforward application of techniques already learned, and the only real challenge was in execution — particularly on levels involving the loathsome wheel. So I made good progress, and, as is often the case, the perception that I was close to the end made me redouble my efforts. Even the level passwords started ratchetting up the sense of anticipation as level 80 neared: “getting there”, “now its done”, “im out of here”, to name a few.

And then, after level 80 — a suitably climactic design, using every sprite in the game — just when I was wondering if there was going to be any sort of special victory sequence, something peculiar happened. There was one more level. It had the same password as level 1, so for a moment I thought it was just looping back to the start, like an old coin-op game. But no, it was a new level. A bonus level? Would the game end after this one? I played through it, and found myself on level 82. The game was clearly ending at level 80, but at the same time showed no sign of stopping.

Confused, I turned to the internet. That’s where I had learned there were 80 levels in the first place. The documentation doesn’t say anything about how many levels there are. (The game comes with a nice thick manual, but it’s mostly useless, and consists largely of pseudo-humor about cavemen.) I found the explanation at Wikipedia, of all places. Apparently the CD-ROM version isn’t equivalent to the floppy version after all. It contains both the original Humans and an expansion pack, variously titled The Jurassic Levels or Insult to Injury depending on country, which adds another 80 levels.

And so I end the day as I started it: a little more than halfway through the game as I know it.