The Epistemology of Kong

So, apparently this month is the twenty-fifth anniversary of Super Mario Bros. for everybody, and next month is the twenty-fifth anniversary of the Nintendo Entertainment System, which mostly matters to us Americans. Fantastic!

All I've been able to think about is The King of Kong. That may seem appropriate because the movie is about Nintendo's first super-hit arcade game, the game that was the debut of Mario, but that's not why it keeps popping into my head, so I'll explain.

I'll start by making a generalization that's probably kind of platitudinous. Do we mostly think that documentaries are educational, informative, credible, straightforward, and authoritative? I think a lot of us do make that assumption. Some documentaries are designed to inform and educate, and some of those are honest about the limits of the information they can provide.

A lot of documentaries, however, are designed to persuade or simply to entertain, and if they do inform, that is either a secondary aim or a thing that they do in service of their greater agenda. Information is filtered so that what serves the director's or the writer's argument is included, and what detracts from it is, in almost all cases, excluded. Information that detracts from the argument but is still included is almost always discredited in some way. Michael Moore's films are known for presenting commentary from or footage of his opponents that is selected to make them seem unreliable. Lengthy footage of George W. Bush being prepped by makeup and hair people is included in Fahrenheit 9/11 as if to suggest to Moore's audience that this man is surely too ridiculous (and vain?) to be leading the country. Expelled: No Intelligence Allowed, the 2008 documentary that portrays intelligent design as a legitimate scientific theory whose exponents have been persecuted by the academic establishment, seeks to create a chain of association from evolutionary theory to social Darwinism to eugenics and thus to the Holocaust. The film scores one of its points on this count by excising, without acknowledging the alteration, crucial lines about the importance of sympathy and its status as one of humanity's noblest tendencies from Darwin's otherwise clinical account of the tendency of humans artificially to resist the ravages of natural selection. The result is that Darwin seems, to those unfamiliar with the passage, to suggest that we should do away with the mentally and physically infirm.

Anyone who works in television, film, or theatre will dismiss the footage of Bush being made up and coiffed as indulgent and misleading; hair and makeup are always applied before major on-camera appearances, even in times of crisis, for better or worse. The footage therefore isn't relevant to the film's argument that he is a poor leader. Likewise, anyone who has read Darwin will likely remember his gentle and endearing observations and his obvious compassion for humans and animals alike—Darwin's capacity for finding wonder in life in all its variety was far too large to coexist with any cold tendency towards eugenics.

This is to say nothing of the inevitability that the process of filmmaking itself will come between a documentary's evidence as it was originally gathered and its argument as it is ultimately presented. All documentary filmmakers film many more hours of footage than can fit into a feature-length film, and even the most earnest filmmaker will have to edit something out, and not every piece of relevant information does or can be caught on camera, anyway. Every documentary film editorializes with its content, even if the intent of the film is purely to inform.

No documentary film is objective. When it comes to a movie like The King of Kong, it's completely obvious that we're watching a reconstructed account—not to say a reenactment, just a heavily edited remix of real footage—and I probably seem to have belabored my point. Sorry. However, what I've been trying to do is to undermine in several ways the assumption that we can rely on this genre, which we often simply take to be the film equivalent of non-fiction literature.¹

The King of Kong isn't much of a documentary, frankly. In what I think is an obvious way, it transforms actual events involving Steve Wiebe, Billy Mitchell, and the bureaucracy of Twin Galaxies into a kind of epic duel between the just underdog and the conspiratorial agents of wrong. We Red Sox fans watch this transformation happen, and we contribute to it happening, all the time.

Anyway, Billy Mitchell marches, Vader-like, to the resentful teeth-gnashing of Leonard Cohen's "Everybody Knows," through a world that seems to cater to his every whim. Steve Wiebe slumps and smirks through a world that is completely indifferent to him to a stirring little piano tune he composed himself. Everybody in the retro-gaming community falls all over himself—not to be sexist, but these dudes are all dudes, with the sole exception of the adorable Doris Self—to get his ass to Billy's lunch thing or to watch Billy's latest high score video or to type that high score into the Twin Galaxies database without a second's hesitation. Everybody aside from Steve Sanders seems content to ignore Steve Wiebe altogether or to condescend to him, notwithstanding the fact that whether he is Billy Mitchell's equal at Donkey Kong or not, he is hella better than anyone else at it;² wouldn't you agree, BRIAN KUH.

See? I can't even talk about the movie without getting worked up. It doesn't inform; it tells its own story. And predictably, there is a debate about the story, too. Walter Day says that the movie makes everyone in the Twin Galaxies side out to be bad people when they aren't, the movie misrepresents the facts, and so on and so on. Seth Gordon, the film's director, has said that his treatment of Billy Mitchell in the movie is in fact overly charitable. The movie even simplifies the high score race to the point of distortion: while it looks like a huge defeat for Steve when the Twin Galaxies referees reject his video, what actually happened was that a previous score of his, which had already beaten Billy's 1982 high score, remained the reigning Donkey Kong score. That controversial taped score of Billy's was also eventually rejected.

Anyway, all of this is just to point out that not only is The King of Kong a representative of a category of things that by their very nature are subjective products, but it also claims to present evidence of another reliability breakdown, this one on the part of the vaunted organization of Twin Galaxies. Walter Day spends a lot of time in The King of Kong talking about the importance of Twin Galaxies, presenting it as an indispensable piece of the video game subculture. But the film also shows Twin Galaxies' referees extending privileges, at least initially, to Billy Mitchell that they deny to Steve Wiebe, despite their insistence that they are an organization dedicated absolutely to fair treatment. The film also shows that it is not out of the ordinary for gamers to lie about scores or otherwise misrepresent their skills, through cheat codes or hardware or software alterations or video trickery. Even video games don't necessarily tell the truth!

So King of Kong is an unreliable film about an organization that is also unreliable that chronicles high scores won by people who might also be unreliable. Where do we go when we want to know who holds the high score in Clu Clu Land? How do we know that person didn't cheat? How do we know Twin Galaxies really put the screws to that person and sufficiently vetted their score recording?

We can't know anything, readers, as I discovered this week. Nintendo has released a timeline of the Super Mario Bros. series as well as a video (now fixed on the official website, those sneaky bastards!) showing footage of most of the major titles in the main series. So what? Well, every game from Super Mario Bros. 2 to Super Mario 64 is listed as being released a year later than it was actually released. Not even Nintendo can be relied upon to give us information that it should be wholly within their ability to give us! How do we even know that this year really is the twenty-fifth anniversary of SMB? Maybe next year is the twenty-fifth anniversary!

Maybe it was last year. Maybe it was last year.

¹ Actually, I don't believe in non-fiction literature, either. Surprise! I kind of sort of wrote about editorializing in the biographical genre in a previous post, and I should have pushed my point further and said that it is inevitable that all literature, deliberately fictional or deliberately non-fictional, is written by a limited person with limited access to limited information.

² Excluding of course the very charming Dr. Hank Chien, whose humility is surpassed only by his Donkey Kong score and apparently Billy Mitchell's latest score, too.

Opus of Loquacity (Part 1)

If I'm remembering correctly, Super Mario Bros. was the video game that introduced me to the concept of a level. You see, I'd grown up with nothing but the standard 1970s-to-early-1980s iterative games, the kind where you play on a single screen and, when you kill the last enemy on that screen or eat the last pellet or whatever, you're taken to another screen that's almost identical, even if the challenge is higher or the colors have changed.

My understanding at the time—please understand that I was three or four years old—was that the one screen was essentially the whole game; since the gameplay was almost identical the next time around, if you'd played one screen, you'd played them all. You could justly claim to have “beaten” the game, so I thought. This must have done quite a bit for my self esteem as a child. I can actually remember a childhood trip to a Pizza Hut where I finished exactly one screen of Centipede. When that last centipede bit exploded, I gladly turned away from the machine and beamed at my mom. “We can leave now,” I said, proudly. “I beat that game.”

Obviously, Ninja Gaiden was still ahead of me.

Anyway, since I didn't play Donkey Kong until after I'd gotten my NES, I had no idea that a game could be broken up into small segments of play that constituted only a portion of the whole. DK would certainly have gotten the message across; in its absence, though, SMB did the trick, and after I got to the end of World 1-1, slid down that flagpole, and went in the tiny castle, I was genuinely surprised to see Mario exit the castle immediately afterward and go down a pipe into another level. I mean, World 1-1 could have been the whole thing. I had no expectations for game length; as far as I was concerned, World 1-1 was quite long enough. It was certainly hard enough; my six-year-old hands had never had to work with such precision in a game. Sure, that whole “World 1-1” label in the corner of the screen should have clued me in that there might be a world 1-2 and perhaps even a world 2-1, but I guess I'll chalk that one up to the fact that it was a Christmas morning, and I was ridiculously overstimulated; not to mention, I still had this new Zelda game to play! I was in no condition to use my reason.

But as it turned out, Mario's adventure had just begun, and so had mine, and yes, that is the way I'm opening this thing.

So I learned about levels with Mario, but it really wasn't until a few years later—1990, in fact—that I learned that game levels could actually communicate ideas. If you think about Super Mario Bros., this isn't surprising; what does the level design in SMB communicate? SMB's level design doesn't communicate a whole lot. There's a clear sense of progress, paradoxically reinforced by the game's repeated use of four-level cycles, and there's a fairly vague sense of movement, as if Mario were traveling from province to province within the Mushroom Kingdom, but that really is about it.

I know I'm oversimplifying, but I'm going to try to boil level design for 2D platformers down to just two things, which I am lazily going to refer to as challenge and aesthetics. By “challenge,” obviously I mean the kinds of things that are purely related to gameplay: platforming challenges, enemy placement, that sort of thing. “Aesthetics” covers a broad range of considerations, but the ones I'm mostly interested in are those that involve the kind of real-world sense the level makes to the player experiencing it. Could this level exist in the real world, and if so, how would it work? Both of these elements are necessary, but the degree to which either one is necessary depends entirely on the kind of game a designer wants to create. Something like SMB needs just scads of challenge, because frankly that's what SMB is all about. The need for that real-world sense is way less important, since SMB is essentially a virtual playground with environments that don't need to (and don't!) make much sense at all. If the player can tell what's solid from what's not, and what's harmful from what's safe, that should just about do it. Anything else—end-of-level castles, occasional weather effects, all the other little things SMB happens to do—is just icing.

When you think about a game like Castlevania, the importance of the two elements is almost reversed. Thinking back on the level design for the first Castlevania, I don't remember anything like the sort of sophistication on the level of that “challenge” element, though CV is, duh, hella harder than SMB. It just isn't nearly as involved as a platformer. Thank god for that, too, because Simon Belmont is very nearly immobile; he might, in fact, be gaming's first asthmatic (or perhaps pneumonic) protagonist. On the other hand—and as so often I owe this basic observation to Jeremy Parish—the environments in Castlevania make perfect real-world sense. Think about it: there are so many games from the 8-bit era that are basic platformers. How many actually made those platforms make sense? Castlevania did just that; as Parish points out, “whoever crafted the levels actually took the time to make the platforms logical -- something of a first. Whenever there was a hole to leap over, it was clear by looking at the background that the gap was caused by crumbling masonry as the castle rotted with age.”

Castlevania's great achievement is that it marries these elements of challenge and aesthetics. Challenge, yes, is at a minimum, if only because the Belmonts suck at walking, to say nothing of jumping, but almost all of the challenging aspects of the game are justified by the game's strong aesthetic sense.

It's a brilliantly designed game, but it isn't the focus of this series of posts. I enjoyed the design of Castlevania when I was a child, but I also took it for granted; I could never have told you how well the game was designed, because it never occurred to me how much thought must have gone into crafting a game like that. However, there was a game that shook me from my slumber, not because it made me take notice of the way platforms were thought out, but because it got me thinking about the way levels were strung together to create a world.

This could have happened several times by 1990, but it didn't. In fact, Castlevania III: Dracula's Curse was the first game that really made sense to me as a cohesive collection of levels.

This isn't really too surprising; CVIII takes the already brilliant inclusion of the inter-level castle map from the original game and pumps it up a bit by adding the element of choice. Now, the player could take one of two routes to get to the innermost reaches of Dracula's castle. Further, Dracula's castle was much more varied now, with the environments covered in the first game comprising only the last quarter or so of the game. Dracula's castle was now more of a ranch, really, complete with swamps, dismal underground caverns, and even a pirate ship (!). The way that Trevor and his spirit buddies started out in a creepy ghost town in Transylvania—perhaps the same town haunted by the crazy old man right on the outskirts of the castle in Simon's Quest—and slowly made their way towards the castle, working their way in through miles of hostile territory, just made perfect sense to me. It also set Dracula's castle inside a world that felt real and varied. It wasn't just a castle that sat at the end of a tree-lined path, as in the first game, and its surrounding environs weren't the repetitive towns and wilderness areas of the second game. It was a fortress built on an island surrounded by inhospitable countryside, and suddenly, it made perfect sense to me why a game character would go through that opening town and attempt to make the straightforward crossing at the bridge only to double back to the swamp when the bridge collapsed.

In other words, the series of levels in CVIII told a story. It wasn't an involved story, but it was one that made a certain amount of sense. In games that were roughly contemporary, like Ninja Gaiden, characters bounced from place to place, and even the cutscenes that ran between levels did little to make these levels feel connected. CVIII, on the other hand, did next to nothing by way of explaining the transitions from level to level, but the game's world felt organic and real nonetheless. In fact, CVIII feels like the perfect run-up to the current “Metroidvania” school of CV design; replace the map screens (and the level breaks) with transitional areas and make it possible for Trevor and friends to backtrack, and all of a sudden there's little separating CVIII from a modern installment in the series. It isn't surprising at all that after the refinements on the CVIII formula in Dracula X: Rondo of Blood, the next CV game happened to be Symphony of the Night.

I have much more to say, some of which might actually be of interest, and I have some observations on genre I'd like to explore, but excuse me, because A) it's getting dark and bad things are out at night and B) my laptop's battery is about to go dead.

The above image comes from retrojunk.com

Space and Structure in the Super Marioverse, Part 1

Since it was released, I've been trying to decide what I think of Super Mario Galaxy. I know that it's at least a very good game, and it exhibits the usual creativity and quality of the Super Mario games, but is it maybe even something more? Is it actually a revolution in the platform genre, as a number of reviews have said—and as the game's Famitsu (38/40), Metacritic (97), and Game Rankings (97.4) scores may suggest?

If you follow this blog, then you know that I'm probably not going to spend a whole lot of time talking about Super Mario Galaxy without first discussing its aged forebears, so I'll dispense with the pretext right now and get on with what this essay is really about, which happens to be Super Mario Bros. SMB certainly wasn't the first platform game—the first, um, good one was Donkey Kong, which followed a couple of less remarkable pioneers—nor was it actually the first scrolling platform game (and let's just leave it at that), but it was the game that happened to popularize the genre, and it has the added distinction of making it obvious to the American business world that home console gaming was a viable, permanently marketable, um, thing. As much as I do love the VCS, it's important to note that Atari had not managed to convince people of this by 1983.

I've written about this before, and I've said that I think a lot of this had to do with the way that early NES games distinguished themselves from the arcade experience in crucial ways, though admittedly I focused more on aspects of play control than the trajectory of gameplay. But let me reiterate that what post-SMB console games offered was not the arcade model, the iterative, short game with levels that the player experienced repeatedly and, most importantly, with no end to the procession of levels, unless it was a hardware-imposed kill screen. SMB was the catalyzing agent that popularized a new model of game design: games with actual goals. We might as well call these teleological games.

As is completely appropriate for an early representative of this "paradigm," SMB exhibited the marks of where video games had been as well as where they were going. While SMB had thirty-two distinct levels, grouped into eight worlds of four levels each, it is true that the levels were only slight variations on five basic types: simple outdoor levels, underground levels, high outdoor levels with advanced platforming challenges, water levels, and castle levels. As such, there was some repetition, but it was not the same precise iteration seen in titles like Donkey Kong or Gorf. And whereas these games would simply repeat the cycle of levels without any resolution (DK did have a minimal storyline, but this only played out in the first two levels), SMB was plainly over after the last of its thirty-two levels were completed. There was even a mechanism that marked the final castle level out as something different: rather than the Mushroom Retainers who appeared seven times before—and my six-year-old self, clueless as to the number of times I'd have to find out that the Princess was in another castle, came to hate these guys—Mario discovered Princess "Toadstool" herself, and that was that.

This became, I think, the prevailing paradigm for game designs throughout the console's life, and I think it's safe to say that, for the most part, this has been the case ever since. The titles most often associated with the NES, including the Castlevania, Mega Man, Legend of Zelda, Dragon "Warrior," and Super Mario Bros. franchises, as well as Final Fantasy and Metroid, followed this design model.

What this meant, though, was that games had to show not just the distinction between beginnings and endings but the whole spectrum of locales in between these two endpoints. SMB might only have had five distinct level types, but it used those types in ways that suggested change and progress: outdoor levels were often given a black rather than a blue background, suggesting a transition from day to night, and level 6-3 even changed the color of everything including the bricks, the tufty raised platforms, and the castle at the end, to shades of white and light gray, suggesting a thick snowfall; levels 3-1, 3-2, 5-1, 5-2, and 7-1 had trees with white-colored tops and white pipes that suggested, I guess, a lighter dusting of snow. All of this made levels seem more distinct during the experience of gameplay, and it gave context to the consistent four-level cycle that always culminated in a (futile) attempt to save the Princess from another castle.

For me, then, the move from "world" to "world" felt like a progression from region to region (province to province?) within the Mushroom Kingdom. Mario traveled gradually through the four stages of each "world," and when he reached each castle, always inexplicably in the "world's" most remote location, he discovered that, yes, he had to keep moving. The large castles that always appeared at the beginning of levels numbered 1 (excepting 1-1 for obvious reasons and, bizarrely, 5-1) gave a sense of continuity to the proceedings, and it reinforced the impression that Mario was on some sort of quest that, for all its repetition, was advancing in some sense.

So Super Mario Bros. was something of a tenuous first step. Only minimal differentiation existed between levels, but there was nevertheless a great deal of difference between Super Mario Bros. and earlier level-based games like Donkey Kong. Super Mario Bros. also distinguished itself from its few earlier competitors in the side-scrolling platform genre, like Pac-Land, by exploiting a repetitive pattern in such a way that its ludic and (admittedly thin) narrative qualities appeared to blend seamlessly.

Still, the world that Super Mario Bros. took place in was a product of cooperation, a construct of the player's mind just as much as it was a part of the game's program. It would still be some time before the Super Mario Bros. games managed to present a world that made sense on its own terms, and it is the progress towards this point, and the movement away from it that I think has occurred since, that I've set out to explore throughout the present series of essays. The next part will continue with the next two games, and there are a number of other issues I'm looking forward to bringing out, too. Please anticipate it (and, in honor of Chris Kohler's wit: bow).

All images come from Ian Albert's brilliant webpage, where the author presents several video game maps created from in-game graphics. Just like Nintendo Power, only all nice and stuff.

Manufacturing Enjoyment? (with apologies to Walter Lippmann)

So, this morning I woke up and came down to campus, and what to my wondering eyes should appear but the second minor snow New York's gotten this year.

That honestly didn't rhyme the way I originally typed it, but see what happened? Sorry about that.

And I'm only partially sure about this "second snow" thing. We had this bizarre out-of-nowhere flurry in late October, I think, and that's the only other one. This one seems to have spent itself for the moment, but looks like we're due for more "rain/snow" this afternoon. Meaning flakes + wet, gross sidewalks. Oh, New York, how I love you!

This is all almost beside the point, because what I'm really interested in this morning, after a thread at Gamespite's forums sparked a few thoughts, is this question: why are there so many ice/snow/winter levels in video games? Super Mario Bros. 2 and Super Mario Bros. 3 and arguably the original SMB all have one, as does almost every other Super Mario Bros. game. So does Teenage Mutant Ninja Turtles 2: The Arcade Game for the NES. Sonic 3 has one; Donkey Kong Country has one; Metal Gear Solid basically is one. Lots of adventure games and RPGs have an ice section, including the ice cave shoehorned (sorry; it was!) into Ocarina of Time and the incredible Snowpeak Mountain Mansion in Twilight Princess, the aptly named Winters in Earthbound, Phendrana Drifts in Metroid Prime, and Death Peak in Chrono Trigger.

I'm tempted to assume that it's just an easy way to throw some visual variety into games that would otherwise be dominated by green and brown, and honestly, that's probably not inaccurate. Mario games in particular are known for their almost elemental stages: desert levels, water levels, fire levels. Again, this could be a matter of visual interest: desert levels use lots of yellow; water levels use lots of blue; fire levels use lots of red and orange. These swaps of huge swathes of the palette do create what amounts to an artificial feeling of variety; they also come with all sorts of associations that make for easy categorizing and theming. But come on, it's so much more fun to dig a little deeper and come up with a more interesting if not necessarily more correct answer. Let's just assume that it's not just convenience. What would that mean?

Well, I kind of suspect that if we deny ourselves the obvious assumption, winter levels may be a clever, perhaps even a subconscious, way of playing on the ways we think about things like color and time. I've mentioned that there are these "elemental" levels in the Mario games; the 3D Zelda games have much the same thing going on; the first Metroid Prime does this too. There's a lot of emphasis in all of these series on the tie between colors and certain kinds of environments, and indeed we might take these connections as suggestive of particular elements--not quite the ancient elements of earth, fire, air, water, but something fundamental. Certainly they seem to suggest certain crucial ecosystems. For instance, both Ocarina of Time and Twilight Princess start off with a trio of dungeons that have something important to do with, in order, plant life, fire, and water. Ocarina of Time hammers in the connections to these elements later on with yet more forest, fire, and water temples, all explicitly and chromatically associated with their respective elements, but the early dungeons in Twilight Princess form a clear progression that's synced up with a color palette that moves from green to red to blue. Thinking back to Ocarina, Link is in the first three dungeons collecting gems colored green, red, and blue. Even Wind Waker has a forest dungeon and a fire dungeon.

Each major Zelda installment (and I'm only somewhat unfairly excluding Majora's Mask when I say this) seems to be playing more and more with the series' set of conventions, to the point that many have said that Twilight Princess is essentially a remake of Ocarina of Time. I really do think a lot of this perception comes from the games' persistent connection of color to concept. And these concepts that are deployed in the games work well because we already believe in the kinds of worlds Nintendo's able to concoct. A desert world should be predominantly yellow, mostly rounded because of all the sand dunes, should have a bright sky, should feel dry, etc. An ice world should be predominantly white, might have some craggy rocks or ice crystals or perhaps some rounded snow drifts, should feel dry but crisp, etc. A forest world should be mostly green, have lots of dirt and stumps and trunks and branches, should have dim lighting because of the forest canopy, and should feel somewhat moist. All of these ideas come out in many of the desert, ice, or forest worlds we encounter in modern, sophisticated video games.

But where things get kind of strange is in games where these ideas come into close contact with each other, which I suppose means pretty much any game. Something like Metroid Prime seems almost odd, because you can take perhaps five minutes and travel from the lush, green overworld of Tallon to the fiery underworld of Magmoor to the wintry landscape of Phendrana Drifts. I realize that the overworld of Twilight Princess is supposed to be huge, but if you're thinking too literally about the game's spaces, it can feel quite strange to travel only a short distance and go from a beautiful, quiet lake scene to a cold, harsh alpine environment. The environments in Mario games have always been connected arbitrarily; if there's one message you can take away from the way Mario moves from world to world in one of the 3D games like Mario 64 or Mario Galaxy, it's that this is a Mario game, and Nintendo doesn't particularly care (and hasn't ever cared) if you think there's a natural progression to the levels or not.

Still, I think we sort of need these things in our games. I realize that I've brought the point back to where I started, and I'm running the risk of creating a circular argument wherein the very fact that these differentiated worlds exist means that we need them. Yes: what I seem to be setting up here is a theory of manufactured enjoyment, an internally defined and enforced regime of comfort gaming propagated by Nintendo and other publishers. But that's not what I'm saying. It's always possible to get out of these circles by saying that the system came about and continues because it works. Yes, its continuation depends on its original instantiation; it wouldn't be here if Nintendo (et al) hadn't decided it was a great way to design games. But that doesn't mean it's not an aesthetic decision that resonates strongly with Nintendo's audience. Nintendo probably would not exist today if it weren't for the staying power of its major franchises, and I suspect that a lot of those games' success relies on their ability to touch their audience. Furthermore, I don't think Nintendo would have stuck with this wonderful system, seen first in Super Mario Bros. 2 but really raised to an art form in Super Mario Bros. 3 and later games, if it hadn't been for the fact that this progression of colors and elements and concepts didn't appeal to something embedded deep in the human psyche.

So here's the thing. Basically, a good game takes a fair amount of time and effort to finish, right? Something like Twilight Princess takes, what, sixty hours or so for the average mortal to complete? But even something like Super Mario Bros. 2 felt long in its proper context, and that game, of course, is intended to be finished in a single sitting. These games take commitment, particularly a time commitment. We need these sign posts of advancement, these markers that show a movement from the part of the game world we're leaving behind to the part we're entering. These connections of colors and ideas, these webs of association, work the way they do because in some sense we need a game to feel like real life; we need games to have visual cues tied to progression, the same way that the natural world does. Most of us use manmade rhythms to understand the passage of time. Many people are students or educators who use the school calendar as an index of movement. Some people are in business and rely on the fiscal year calendar to understand progress (or a lack thereof). But almost all of us (inhabitants of equatorial or polar regions excepted, I guess) understand the world's natural rhythm of seasons, a rhythm that is itself a mixture of colors and associations. The brilliance of winter levels in video games is that they represent a piece of the actual cycle of seasons we live within, and so they produce a significant link between the artificial world of video games we are trying to understand each time we play and the real world we understand better.

Which might just explain what's so great about playing winter levels, if you're into those.