The Stack

I’m a couple of days late with this — it’s a busy time, and I seem to have entered one of those can’t-start-writing zones — but the latest news is that I’ve made it through level 53, “Mount St. Monday”. The level purports to be created by one Ken Megill (yes, Icebreaker takes the unusual step of identifying the designers of individual levels, and why not?), but I can only assume that this is a pseudonym for the Marquis de Sade, or possibly Torquemada. Maybe it’s just me, but this level seems difficult out of proportion to the levels around it on either side.

It’s all down to the terrain. Icebreaker has several sorts of terrain hazard: stones that block movement, pits that kill anything that attempts to move onto them, slime that kills you but doesn’t affect the enemy, swamp that slows you down, ice that makes you skid. Some types of Seeker pyramids know how to avoid specific sorts of obstacle: the basic ones always charge at you directly and either fall into pits or get stuck on rocks, but there are ones that will go around pits and ones that go around rocks and ones that do both. The area containing the stationary pyramids is a rectangle diagonal to the grid-lines, so that its edges are corrugated; around it is a very large margin of a default tile type, usually something navigable, providing the player with a place to do an end-run around a pack of pursuing Seekers.

Level 53, now. Level 53 has a volcanic theme: the normal ground tiles are blackened and cracked, the pits are pits of lava. At least half of the playfield is lava or rocks, and the margin is pure slime. You’re effectively trapped in a maze. To traverse this maze, you frequently have to cut across corners. All terrain hazards fill the entire map tile they’re on, but if you’re careful, you can move between two tiles that touch at a corner without incurring the hazards in either of the adjacent tiles. The player has doubtless learned this by now, but in the past, this has usually meant edging between two similar hazards. Here on level 53, we get the peculiarly cruel scylla-and-charybdis combination of corner-movement with a pit on one side and a rock on the other. Aiming for the middle and being slightly off on the rock side means you get stuck, but unsticking yourself by edging slightly pitward can spell death. And when you die in this game, you have to start the level over from the beginning. This happened to me a lot.

Icebreaker provides multiple difficulty levels, so I had been thinking for a while that if I ever found a level that was too tough for me, I’d just dial it down. But that’s not much help here, because the difficulty only seems to affect the number of enemies, and the enemies aren’t really the problem here. Or so I thought, until I cleared all the stationary pyramids for the first time. In order to win a level, you have to destroy not just the stationary pyramids, but the Seekers as well. As long as there are stationary pyramids remaining, dead seekers simply respawn out in the margin, but once the land is cleared, you can blast them for good. On most levels, this really just amounts to a brief victory ritual in which you turn around and open fire at the crowd on your tail. In level 53, there are a smattering of stupid pink Seekers that get stuck on rocks — specifically, having spawned on the outside of your maze, they get stuck on the rocks around the outside, in all directions. So after you’ve successfully navigated all the pinch-points once, you have to go back through them again, trying to reach the open spaces on the periphery from which you can blast the last remaining Seekers. Once, I got to this point only to realize, to my horror, that one of the Seekers had managed to get stuck on the outside in the wedge between two rocks, a place inaccessible to my fire, and which I couldn’t lead it away from. I had no choice but to start over.

Anyway, I’m past it now. But if this sort of thing becomes the norm in later levels, it could take me a long time to finish this game.

I’ve finally conquered the tower of the Cult of Kefka — not the tower of Kefka himself, but a lesser imitation, which can actually be climbed. It provides a nice bit of variety by changing the way combat works: within the tower, neither you nor the monsters can perform any attack other than casting spells. A largish fraction of the monsters seem to have the Reflect effect on them, too, even if they don’t explicitly cast Reflect first. This means that you can’t rely on direct-damage spells. At least, not targeted ones — area-effect spells do fine, and that includes most Esper summons, which count as spells. Alternately, you can cast Reflect on one of your own guys, and then cast direct-damage spells at him, reflecting them back at the enemy. (Spells can only be reflected once.) My favorite tactic here is to summon Carbunkle, which is the equivalent of casting Reflect on everyone in your party at once. Then you can cast a whopping big direct-damage spell like Fire 3 on your entire party at once, splitting the reflected effect four ways — and, of course, get the added advantage of complete protection from the enemy’s direct-damage spells while you’re at it. So, basically, most of the encounters here are a breeze once you figure out these tactics, as long as you don’t run out of mana — Carbunkle is one of the cheaper Espers to summon, but I still had to bring a load of mana restoratives in with me, and used most of them. Which wasn’t strictly necessary: Osmose, the mana-leeching spell, works really well here, if you can bear to waste valuable attack opportunities on it.

Anyway, the whole experience is a nice rules-puzzle. Encountering the Reflect-enhanced creatures for the first time, my reaction was basically “Aaaaah! What do I do? I can’t hurt it with spells, and I can’t take it down with a melee attack, like I’d normally do to something that I can’t hurt with spells!” But really, there are quite a few things you can do, once you think of them. You just have to get out of the rut of thinking like you do in normal encounters.

Even having mastered all that, though, I wound up basically playing through the whole thing three times, because of the tower’s boss. It’s not that he’s hard to beat — he has randomly-changing elemental resistances, but by this point, my entire team had mastered some non-elemental damage spells. I trounced him handily on first encounter, only to find that my entire party somehow perished during his death throes. On my second attempt, I was careful to keep everyone at full health and have some protective buffs on at the end, but the same happened. I resorted to hints to find out what was going on: apparently dying there is inevitable, and the only way to continue is through the Life 3 spell. Life 3? I had that spell, but hadn’t used it — generally speaking, the resurrection spells are ones you want to avoid needing to use. The in-game description of the spell was “Protects from wound”, which didn’t seem to justify its insane mana cost: there were other spells to protect you from damage, and other spells to heal damage as well. What I had forgotten is that “Wounded” is the game’s name for the status I had been thinking of as “Dead”. Life 3 is a preemptive resurrection, like the Ozmoo spell in Enchanter. Cast it on someone, and they’ll be automatically resurrected after the next killing blow.

I’ve been told by now that the edition I’m playing is not a very good translation. I keep finding more and more evidence of this. Some of the creatures in the tower had a spell called “Merton”, which seemed to be an area-effect heat-damage spell, judging by the graphics. Merton? I finally figured out that it was probably a bad re-romanization of “Meltdown”, and a glance at Wikia confirms it. But that didn’t impede my ability to play the game. This confusion over the meaning of the word “wound” did. I suppose “death” isn’t really a good description either — this is an effect that can be cured by a stay at an inn. “Unconscious” or “Knocked Out” would be good, and apparently some games in the series use the latter term. Maybe even the better translations of FF6 do. But I’ll keep playing the one I have.

Icebreaker is not a game about precision aiming. No quick mousework is required, and nothing dodges your blasts. It’s controlled by an eight-direction digital D-pad or its equivalent, and the things you’re shooting at are mostly either slow and predictable, or completely stationary and arranged in neat lines. And yet, I find I miss fairly frequently. Why?

For one thing, collision detection seems to be somewhat buggy. Or maybe not; I’m having difficulty making up my mind about this. It’s definitely true that when a seeker pyramid is approaching me from due north, and I shoot due north, my shot will occasionally go right through it, possibly destroying a red pyramid directly behind it. But understand that seekers don’t simply glide rigidly from place to place. Rather, they trundle and sway, wobbling like they’re made of rubber. Sometimes they look like they’re actually leaping. I can almost convince myself that they’re leaping over my shots, or, more likely, swaying out of their way — the oblique isometric view means that the grid-lines are diagonal, so what looks like moving straight down could be a series of diagonal tacks. Or it could just be a glitch. The framerate and the speed of the shots are such that the shots actually do skip stretches of pixels between frames, visually if not internally.

But there’s another factor, and one which has got to be deliberate. I said that the grid lines are diagonal. And I said that you could fire in eight directions, including diagonally. But diagonal shots don’t follow the grid lines. They’re at a slightly different angle — off enough that you can stand still and blast at a row of three red pyramids, and destroy the first two but not the third. I’m not entirely sure if the same applies to diagonal movement, because movement typically involves frequent small stops and adjustments to deal with attackers, but it probably does. This difference sometimes allows the player to pull off shots that would otherwise be impossible — for example, shooting a red pyramid on the other side of a green one that would otherwise block the shot. But mostly it’s just something that you have to get used to.

Something must be said about the game’s origins, about Andrew Looney and his obsession with pyramids. This is all well-documented elsewhere on the web — that is, after all, how I came to know everything I’m about to say — so I’ll be brief.

It started with a self-published novella called The Empty City — the full text is now available online, if you’re curious. In this story, Looney described a tabletop game called Icehouse, and the ethos of cool that had developed around it. Icehouse, as described in the story, was a peculiar thing: a board game without a board, a strategy game without turns. If you saw an opportunity in the way the pyramids were arranged, you grabbed it before someone blocked it. Understandably curious about whether such a system could be made to work in real life, Looney decided to develop the in-fiction descriptions into a game that people could actually play. And thus began his career as a game designer.

But not, it must be said, a videogame designer. Icebreaker was and remains his only credit on Mobygames. He mostly does card games — his best-known work is probably Fluxx, a game where the basic conceit is that the cards you play change the rules (albeit only in specific ways, like how many cards you draw at the beginning of each turn and which combination of cards you need to win). I’ve played much of the Looney Laboratories catalog, but I have to admit that his games generally aren’t what I want from a game — too much alea, not enough agon. Usually the winning move comes as a surprise, which means there’s no opportunity to strategize against it. But tastes differ. Some prefer the beer-and-pretzels school of design, and I’ve noticed in particular that the people who like Fluxx the most are people who don’t usually like games. Anyway, Icehouse doesn’t fit this pattern at all. I find it almost unbearably stressful to play. Perhaps this is part of why people who bought Icehouse sets immediately started inventing other games to play with the pyramids — although aesthetic appeal of those pyramids also played a role, of course. If there’s one thing that the original Icehouse has going for it, it’s that every session results in a unique tableau that looks like the skyline of a Martian city.

Knowing all this, Icebreaker feels a bit like a game from an alternate universe where Andrew Looney’s life went differently. But my first exposure to the game came years before I had any other knowledge of the man or his works: I saw it reviewed in a gaming magazine or two on its initial release, where it was praised as new and different, but apparently not considered important enough to merit anything more than a few sentences in a sidebar. I remember seeing the comment in Electronic Gaming Monthly expressing confusion over the fact that you’re a pyramid blasting other pyramids, and thinking what a weird thing that was to find confusing. I mean, there are plenty of games where you’re a spaceship blasting other spaceships, right? It’s true that pyramids in real life don’t usually come equipped with blasters, but then, neither do real spacecraft. (Come to think of it, the ships in Spacewar are about the same shape as Icehouse pieces. Perhaps they were really pyramids all along!) But I suppose the confusion is more understandable given the blurb in the manual:

Icebreaker is about destroying pyramids. Pyramids are bad. They are evil and nasty. You’re outnumbered and alone. All you’ve got are our wits and cunning… Oh yeah. And a real big plasma blaster.

That’s as much story as you get in this game — yet another way it resembles the coin-op games of yore.

Some time after this, I learned of Icehouse and became intrigued enough to try it. When realized that Icebreaker was by the same person, I naturally wanted to try that too. And so, when I found a bin full of original Icebreaker boxes at a computer show, selling for cheap, I snatched one up. I really should have snatched up more than one, for distribution to the Interactive Fiction community, because the disc contains, as an easter egg, a text-based adaptation of the game by none other than Andrew “Zarf” Plotkin, author of such works as So Far, Shade, and Spider & Web, and a personal friend of Looney. It’s not much of a game — more of a joke, really — but it’s a text adventure, by a prominent author no less, published on CD-ROM and sold in stores, and that makes it a rarity. Really, I think more games should ship with text adventures as bonus items, and there are people who agree with me and are willing to make it happen. I suppose the biggest obstacle is getting approval: games are big business these days, and big business doesn’t like content that hasn’t been vetted by legal.

It’s a little eerie how I was led toward this obscure title by three different channels — computer game magazines, tabletop gaming, and IF. Or was it only two? I don’t remember where I first learned of Icehouse; it could have been from the IF community.

So, what sort of game is Icebreaker? One that doesn’t really fit into a genre category narrower than “action”. People have stretched this as far as “strategy/action” and “puzzle/action”, which I suppose is necessary to distinguish it from mindless action, but neither description really fits — the level of thought is more tactical than strategic, and the only reason anyone would describe it as a puzzle game is that they classify anything sufficiently abstract that way. If you ask me, the genre it has the most in common with is classic arcade games, things like Space Invaders, Pac-Man, and Breakout: games with simple controls and world models, where the player is always bent on the same goal. That goal being, of course, to destroy everything. To clear the screen, by shooting, eating, or bouncing things off of everything, until there’s nothing left.

The things you destroy in Icebreaker are pyramids. Steeply acute ones. ¹This could mean either “ones that are acute in a steep way” or “ones that are acute and resemble steeples”. Either description fits. The main playfield is an isometric grid of regularly-spaced pyramids, the “seekers” that chase you through this grid like Robotrons are animated pyramids, even the player’s avatar is a pyramid on its side. This is basically “programmer art” that stuck — the author’s website describes how the entire game started out as a programming exercise that took on a life of its own. Now, how do you destroy the pyramids? Do you shoot them, or eat them, or bounce things off them? A little of each, it turns out — at least, if I can extend “eat” to mean “collide your avatar with” and “bounce things off of” to cover any sort of induced collision with objects not under your direct control. There are three basic colors of stationary pyramid: red, green, and blue. Red pyramids are deadly to the touch, but can be destroyed by a blast from your cannon. Blue pyramids are cannon-resistant, but shatter when you ram them. Green pyramids you can’t destroy at all yourself, but crumble on contact with a seeker. It’s really the green pyramids that save the game from being trivial. When people call it a “strategy/action” or “puzzle/action” game, they’re mainly thinking about the need to lead the enemies to specific places instead of just shooting them.

There are further complications as you go along: you get obstacles like walls and pits, terrain like slippery ice, smarter enemies (the basic ones are prone to getting stuck), and new colors of stationary pyramid with different properties — purple pyramids that turn into pits when shot, stone pyramids that have to be shot ten times, rainbow pyramids that pick a color at random when rammed or shot. But the three basic pyramid types have a special relationship that these advanced types do not: they occasionally change color, cycling from red to blue to green to red. (This ordering is important, because it prevents blue pyramids from turning into deadly red while you’re charging at them.) The changing colors keep the action from being too predictable, even on boards that start out very regular, and also serve to upset equilibrium. Blue pyramids that are hard to reach eventually become shootable; a seeker stuck behind pyramid need only wait for it to become green to get through it.

Aside from the tutorial, there are 150 levels. A big level grid shows you which ones you’ve completed, and at which difficulty. There’s no unlocking of levels — you can access them all from the very beginning. However, the “next level” button on the victory screen makes it slightly easier to play them in order than to not play them in order, so that’s what I’m doing.

1995 was an epochal year for the PC: with the release of Windows 95, we suddenly had 32-bit addressing, true preemptive multitasking, and, most importantly for gaming, genuine hope for hardware-independent code in an increasingly unwieldy world of semi-compatibility. The installers for DOS games of the time presented to the user long lists of all the graphics, sound, and input devices they supported, and asked the user to select IRQ settings and other such arcana. 3D graphics accelerators were still a speck on the horizon, but the age of the CD-ROM multimedia extravaganza was here, and with it, long-since-forgotten extravagances like MPEG decoder cards. The new Windows Games SDK promised to simplify things by putting a layer of indirection between the software and the hardware — an indirection layer that, in a tremendous feat of denial and marketing spin, was dubbed “DirectX”. But none of this happened immediately, and PC game developers continued to primarily target DOS for a while. After all, not everyone had Windows 95 yet, and why limit your potential audience? Besides, Windows was reputedly inferior as a gaming platform — Windows 3.1 functioned as an abstraction layer too, but tended towards lowest common functionality.

So why, in 1995 of all times, would anyone release games for Windows 3.1? It seems like the worst of both worlds: limited adoption and lagging behind the cutting edge. But apparently it was a convenient platform to port things to — Myst, for example, never saw a DOS port, presumably because Windows 3.1 was a better fit to the original Macintosh code. Today’s selection, Icebreaker, was originally written for the 3DO, and, if I understand correctly, ported to both Windows and Mac simultaneously by a third party.

Installing Icebreaker on a modern system is a bit of an adventure. I’ve run it on a win32 system before, and I know from experience that it has overzealous copy protection that demands that you insert the CD even when you already did. The game’s author, Andrew Looney, has gone on record encouraging the use of a no-CD crack. Possibly related to this, I have never managed to get the game to play its intro, outro, or between-levels movies. But that’s not such a big deal: they’re not an essential part of the experience, and besides, they’re all stored as ordinary AVI files, watchable from the desktop.

A more serious obstacle is the palette requirement. Icebreaker will only run if Windows is set to 256 colors, neither more nor less. Windows apps in those days didn’t know how to change the color depth on their own — this is one of the many reasons why DOS was considered a superior gaming platform. The problem is, my current system doesn’t do 256 colors. 32-bit color it can handle without problems, but 8-bit, once the mainstay of VGA, isn’t even an option. It’s true that I’ve run other 256-color games lately, and even 16-color games, but only through an additional indirection layer — specifically, DOSBox. DOSBox is certainly capable of emulating 256-color mode on a more capable display, but unfortunately, it only runs DOS apps, not Windows 3.1 apps.

I was about ready to give up and pick a different game, when I realized that Windows 3.1 itself is a DOS app, and can be run inside DOSBox.

Thus began the second round of installation fun: locating Windows 3.1 device drivers that behave correctly under DOSBox. None of the built-in graphics drivers supported 640x480x256, but I managed to find something that worked just as well, given a little help from Vogons. It took me a few tries to find a Soundblaster driver that actually produced sound. But now, I have a convoluted-but-functional Windows 3.1 gaming system that, as an added bonus, works on my Macbook, which I really wasn’t expecting when I got started.

Tomorrow, I suppose I’ll try to describe the actual game.

[ADDENDUM] Looks like I could have just installed it under XP and checked the “Run in 256 colors” setting in the “Compatibility” tab in the shortcut properties. But that wouldn’t have helped me play it on the Macbook.

To judge by my last few posts, you’d think that I’m on the verge of completing the second half of Final Fantasy VI. And I am. But it’s a very wide verge. I spent about three months in a similar state at the end of the first half. Admittedly, that was because I wasn’t actually playing for most of that time. But the reason I wasn’t playing was that I had grown impatient with the game: I felt so close to the momentous transition that all the mopping-up I felt compelled to do before taking the plunge became burdensome. As much as I want to face off against Kefka — recently named 18th greatest videogame villain of all time by IGN, right above M. Bison — I also want to see the rest of the game.

Rushing through the game is no way to play it, if only because the game makes it impossible. Sometimes you only get to take a couple of steps between random encounters. Some games in the series have a rare and special piece of equipment that decreases the rate of encounters, or even eliminates all encounters with anything other than bosses. If such a thing exists in this game, I have yet to find it. And if you’re approaching the game from a position of impatience, these constant interruptions will only make it worse. I wrote before about the annoyance of all the system’s little delays when working under a time limit. My self-imposed time limit of two weeks is no exception.

So, on to 1995. I’ve already started on my next game as I write this. But I intend to keep dipping back into FF6, in small sessions, for however long it takes.

I began this weekend hoping that I’d have just one more post to do on FF6, but after multiple hours of play, I still haven’t made a serious attempt as Kefka’s junkyard-like tower of magically-attracted debris. Oh, I’ve visited it, and I think I might even be able to conquer it at my current level, but it’s going to take more time and preparation than I felt like giving it at the time. You see, it splits the party in three, and that complicates things.

This isn’t the first time the party has been split. Way back near the beginning, there was a part with three sub-scenarios that I had to play out with different characters, but that was different: the scenarios were self-contained and independent from one another, and were played out in sequence. More recently, the descent into the treasure cave to find Locke involved splitting the party however you like into two groups, then switching to control whichever group you like at any given moment. And you couldn’t just use one group and leave the other alone: every so often, each group would run into an obstacle that could only be cleared by having the other group stand on a pressure plate somewhere.

Kefka’s tower works like that, but with three groups, which makes it a lot harder to decide how to split things up. Through most of the game, you get your pick of four characters out of the entire party roster, so it’s easy to take your choice of combat specialist, your choice of mage, and your choice of guys with weird special abilities, and still have one slot left over for whoever you’re trying to level up. With three groups and 14 playable characters, you don’t get much choice of who to take. You just get to choose who to partner them with — and my experience is that some combinations have a much easier time surviving than others. It isn’t just a matter of taking one from each of the four categories I just mentioned — you have to take into account that characters are going to be killed or disabled sometimes, and get some redundancy in there, like a mage who can fight in a pinch. (Or, I suppose, you could just grind until everyone is level 99 and not worry about it, but I want to enjoy playing this game.)

Then there’s the equipment. The very best armor and weapons are, of course, not available in stores: you have to find them locked away in dungeons or loot them from bosses. There’s an item called the Atma Weapon ¹Another questionable transliteration: some versions call it Ultima Weapon. But Atma fits pretty well too. It’s a glowing lightsaber-like thing that visibly grows with the wielder’s experience level, as if responding to the strength of your soul. that simply does way more damage than any other weapon I’ve seen. I try to always have it equipped, but there’s only one in the game, which means only one of the three groups can have it. I suppose I could just unequip it whenever I switch control to a different group, but that starts to get cumbersome. And it becomes even more cumbersome when you factor in the Espers. I frequently swap those around between characters even when I’m dealing with only one group, to make sure everyone gets a chance to learn their spells, and also because many of them grant permanent stat increases when their wielder levels up. I’ll probably have to just abandon that habit in the tower if I don’t want to spend 90% of my play time in inventory menus.

Or, like I said, I could stop optimizing and do more grinding. I think it was Ben “Yahtzee” Croshaw who said that Final Fantasy gives you a choice at the end: you can spend 40 hours building up your characters to the point where you can take the end boss easily, or you can spend 40 hours fighting the end boss. This kind of applies to the entire final dungeon as well.

The story in the second half of FF6 is all about finishing things, tying up loose ends from the first half. And it’s funny, because a lot of those loose ends are things that I don’t remember until I see them tied up. There was even one major loose end — the Terra vs Phunbaba ¹A bad transliteration: it should be Humbaba, the demon from the epic of Gilgamesh (and in some later editions, it is). Mistakes like this happen a lot in the translations of the Final Fantasy series, due to their predilection for throwing in random elements from diverse mythologies that the translators aren’t necessarily familiar with. arc — from the early part of the second half that I didn’t remember. Things I don’t remember are things that I can’t pursue as goals. But that hardly matters, because implicit goals are provided by the game’s very structure: you visit every dot on the map, talk to every NPC, and explore every dungeon, and in the process, you wind up completing the story.

But there’s one other set of major goals the game has provided for me: finding and slaying the great dragons. There are eight of them. I know this because the first time I killed one, I got a message telling me I had killed 1 of 8 dragons. Re-exploring the parts that Celes passed through alone a year ago, I find there’s an NPC who explains how the dragons were released by the cataclysms or something — I don’t remember the details, but there’s some kind of reward for killing them all — probably some magicite yielding Bahamut, the dragon-king summonable from previous games.

I’ve racked up 5 dragons already without really trying, because they tend to show up in places where you’d go anyway: slightly off the main trunk of a dungeon, for example. One of them was even squatting in the opera house. Unlike random encounters, you can see the great dragons as you wander the area: they show up as single-map-tile sprites just like heroes and NPCs, and they look misleadingly cute in that form, like geckos with little wings. So you know when you’ve found them.

Nonetheless, I’ve pretty much run out of places to look, and I’m still short three. I suppose I should recheck the places that I visited with only Celes and Sabin. I would have been avoiding optional boss fights at that point, so I might have passed a dragon by. And after that? I’ll just have to recheck everyplace else. This is basically the stage of the game where it all comes down to grinding: I’m preparing to assault Kefka’s tower, but I need to be stronger before I can make a serious attempt at it. The dragon hunt at least turns the final grind into something purposeful. It gives you something to do other than just wander back and forth and wait to be attacked.

Final Fantasy VI really is the pinnacle of its form, but that shouldn’t be surprising, considering that it’s also the last of its form. The next game in the series shifted to blobby low-polygon-count 3D, and, while that style has its charm, it required different techniques than the rest of the series up to that point. There’s a real craft to storytelling via tiny pixelated sprites, and it was pretty well-developed by now. The human figures have a large library of emote animations for use in dialogues and cutscenes, some of them quite expressive despite differing from the neutral expression by only a few pixels — although others involve running back and forth or leaping in the air several times the character’s height. Everything I said about the theatrical gestures in Police Quest 4 applies even moreso here.

Cutscenes are the obvious place to show off sprite animations, but there’s even more impressive work in the combat. The most noticeable part of this is in the special moves, such as when Cyan swoops into the midst of the enemies, with a comet-trail of desaturated afterimages behind him. The movement there doesn’t doesn’t look at all natural, but then, neither do fireworks. Personally, though, I’m more impressed with the subtler touches, like the alteration in posture to indicate each character’s state. Someone who’s been ordered to cast a spell, for example, will bow their head and make muttering motions until it’s their turn to act. This actually provides useful feedback about what’s going on, whereas the flashier attack animations are just a matter of showing off. The one disappointing thing is that the monsters aren’t animated in combat at all. Certain monsters — mainly bosses — have a fully-animated sprite representation that’s used in the main movement-and-exploration mode, but during combat, all monsters use larger portraits that just stand still. Presumably it would have been prohibitively space-consuming to include animations for every monster type in the game within the constraints of the SNES — even using still images, most of the monsters are palette swaps of other monsters. I suppose the shift to 3D in the next game helped there: suddenly animations were relatively lightweight, consisting of differences in vectors instead of a full copy of the bitmap for every frame.

Today’s indie game developers are in love with minimalistic pixel art, partly because it’s the aesthetic of least effort. But that’s certainly not the case here: the artists put in loads of effort and want you to know it. To the extent that it goes all “less is more”, it’s a product of systemic constraints. Considered purely in terms of style, the closest recent game is probably Braid, which similarly tries to be as evocative as it can with a super-deformed sprite with a limited number of cels. But even Braid was deliberately retro, and there was nothing retro about FF6 at the time of its release.

On the other hand, the game’s biggest reach beyond sprites is something I regard as its biggest failure: the character portraits. In my posts about FF5, I described the concept art created by Yoshitaka Amano, and how little it resembled the stuff in the game. FF6 puts a closer approximation to the concept art in the party stats screens, fitting in greater detail by showing just the head. There’s just something off about these portraits. Some of the faces are just ugly in a way that their sprites are not: Gau looks misshapen, Setzer has scars that you can’t normally see. But even the pretty ones look very wrong to me. I don’t think this is the famous “uncanny valley” effect — even the portraits are too far from human for that. It’s more like Scott McCloud’s famous observation that it’s easier to identify with simplified and cartoony characters than with highly-detailed ones. So anyway, here’s a case where I think it would have been better for the art to be more minimalist than it is, and therefore for the artists to choose minimalism rather than do as much as allowed by the medium (and the budget, and the deadlines).

I’m kind of wondering now what FF7 would have been like if the series had stayed 2D. Would it be a better or a worse game? It was astounding at the time, but today, I tend to think that the primitive 3D of the day has weathered worse than well-crafted 2D of the same era. A remake of FF7 in the style of FF6 seems like such an obvious fan project that I’d be a little surprised if it doesn’t turn out that someone is already working on one. But a quick google only yields rumors and arguments about Square going the other direction, doing a HD remake for the PS3. I guess we all at least agree that FF7 in its current form falls short of ideal.