Log in

No account? Create an account
Jun. 16th, 2011 @ 07:27 pm in which kevin waxes zarathustran
Current Mood: nietzsche
Tags: ,
In the month following the end of my last contract I've done a lot of recentering. Specifically, I've been trying to rediscover what it is I love about video games so much that I'll gladly work an engineer's hours in an artist's living conditions just for another shot at doing it for a living.

When you're bogged down in interface nitpicking and deadlines and cost-benefit compromises and project management issues, while under constant pressure to come up with a brilliant design on the spot, it's easy to find yourself thinking of games as nothing more than the sum of their parts. You don't see the rapture you felt when you first discovered the zen loop in Pac-Man, or the giddy look-at-me-now thrill of running World 4-1 of Super Mario Bros. in a dead sprint. You see messaging issues in the color of the "HI-SCORE" text. You see meters, gauges, ratios between player resource expenditure and strategic gain. You see points where the scaling risk-reward mechanics from Galaxian can intersect with the scaling risk-reward mechanics of unit specialization in Starcraft. You see pipes. Lego pieces. Playmobils. Prefabricated pieces to be combined, smoothed out, streamlined, made efficient, according to well-understood rules. You know vanilla tastes great and why it tastes great; you know chocolate tastes great and why it tastes great. Your job is to make a better chocolate, a better vanilla, and find new ways to make them swirl.

If you get to this point, the magic is gone. The process of game design has ceased to be a creative endeavor and has become a mere feat of engineering. When you catch yourself building games like this--and I'm sure even the best among us do--you're not making games anymore. You're just making software. You're architecting your game the same way you're building a web platform. And of course no one cares if the newest version of a web platform is exactly like the one that came before it, except easier to use and with some interesting new features--in fact, users prefer it that way. But you know what? Web platforms don't require novelty. They generally aren't designed to be fun.

And this--more than the improbable deadlines, rapidly shifting goals, and grossly underestimated technical challenges, or the fact that I wasn't even officially working on a game--is what led to the deepest feeling of existential dread I felt during my last contract. Even when I was enjoying it. Even when I was having the time of my life, moving toy soldiers on hex grids with my co-workers, tossing around ideas in bull sessions, arguing over diagrams of AI behavior, I couldn't help but feel like something wasn't right. My biggest contribution to the team, design-wise, was that I was the guy who had played virtually everything--that I could pluck one of my co-workers' suggestions out of the air, liken it to the mechanics from five obscure arcade titles from the early 1990s, describe what made it fun in those games and give an opinion on whether it would work in ours--and I wasn't comfortable with that.

Don't get me wrong, any creative endeavor builds heavily on its influences--but we were supposed to be the indies, the risk-takers, who were willing to do all the crazy stuff the triple-As wouldn't touch with a ten-foot-pole, and with the pressures of one-month deadlines and a team of four people, most of them incredibly smart designers who really understood games, the best we could come up with was "hey, let's make a game like X, except," or "let's do Y done right." (Much as we hate to admit it, this is the way with most indies. It's the reason why we have so many reinventions or slight variations on Space Invaders, Tetris, Bust-A-Move, and Arkanoid clogging up the bargain bins, across the aisle from all the top-selling first-person shooters. I mean, Angry Birds has held the #1 spot in the iPhone/iPad App Store for ages now, and it's merely the contemporary iteration of a heritage of very similar games beginning with Artillery, a game first written in 1976. It's distressing.) I imagine the everything-has-been-done-before breadth of my knowledge contributed to this line of thinking, unfortunately, and I will admit it was a problem. What kind of game do you make if the most fun game you can imagine, the kind of game that got you wanting to make games in the first place, is a game that has already been made?

Look at the most anticipated triple-A games announced at E3 this year, the ones that are virtually guaranteed to be top sellers. It doesn't take an insider to recognize that the industry has hit a state of everything-has-been-tried creative apoplexy. Two of them are the same gritty, post-9/11 first person shooter--each of them merely the newest iteration, version n+1.0, of the same game that has been winning awards for the past ten years--itself Tribes 2 with Quake Team Fortress-inspired classes, Deus Ex skills, Gears of War cover, and maybe six or seven hours of flashy cutscenes. One of them is a not-much-different sequel to a game exactly like Grand Theft Auto 3, except with AI followers and better shooting and driving controls. The remaining two are a driving game and a soccer game. Even the indies are disappointing, for all their visual marvel and attempts at deeper meaning--I mean, I love existential tug-of-war games and beautiful, immersive, gimmicky platformers as much as the next guy, but come on. Enough reinventing the wheel! Is this what I have to look forward to, career-wise?

Fuck that shit. I want to be a composer, not a DJ. I'm going back to basics.

After all, look at Shigeru Miyamoto, designer of some of the best-loved and most frequently imitated video games ever. What did he draw upon for inspiration in the thousands of mostly terrible, industry-ending Atari and Colecovision titles that preceded his work at the beginning of his career? Platform jumping mechanics from Pitfall, maybe. Silly hats. A score counter at the top. That's it. All the other crazy amazing stuff he came up with was built out of a deep, very fundamental understanding of play, fun, and imagination. Donkey Kong was inspired by an iconic photo of New York construction workers eating lunch on an I-beam of a half-constructed Empire State Building. The bad guys in Super Mario Bros. were inspired on the movements of creatures he saw while walking in the woods near his home. The jumping-on-bad-guys-to-kill-them mechanic was based on wheeeeeeeee. And one of the main reasons why these games succeeded even as the Western game industry crashed was because Miyamoto, like many of his Japanese successors, built his games purely around the joy of play. He didn't try to squeeze meticulously balanced board game mechanics, hyper-specialized for an experience involving four people sitting around a table scheming to turn those mechanics against each other, into a game for one person sitting in front of a television. He didn't fall in love with the complexity of his own simulations, like Will Wright would, or confuse detail with depth, as the early flight sim folks did. He didn't see a video game as a vehicle for narrative power fantasies like being a millionaire or an action hero or the biggest kid on the block, knowing that such details would get lost in the representational abstraction of early game graphics, nor did he build his games around clumsy abstractions of single moments that sound awesome but play terrible, like punching a minotaur in the crotch.

What Miyamoto made were games. Not board games. Not tabletop games. Not sports games. Not toys. A category of games in their own right, in the same way a game show is a game, in the same way politics is a game, in the same way an obstacle course is a game. He didn't have an agenda in the social construction of games to be something they weren't. Go play Super Mario Bros. again, in its original NES/Famicom incarnation--note how opaque the messaging is, how inscrutable the rules of the game often are (I can never remember how to get fireworks at the end-of-stage flagpole), how the game actively encourages game-breaking exploits like the ability to trap a shell between two blocks and hop on it repeatedly for infinite lives, or how you can be an avid player for twenty years without knowing that Bowser is vulnerable to fireballs. Note the little engineering compromises the programmers had to make--the bushes being a palette swap of the clouds, the graphical glitching that occurs if you earn more than 128 lives, the inability to scroll the screen to the left. As a piece of software, it's a beautiful mess. As a game, it's simply beautiful. Every moment of development time that should have gone into making the game less technically embarrassing, more smooth, more elegant, more understandable, more balanced, more fair--it went into the wheeeeeeeeeeeeeee.

And what did we learn from him, we Americans? We learned that jumping on things is fun. That tight controls are important. That players like whimsical characters. And then we went on flooding our games with unnecessary complexity, because we've learned from board games that complexity is depth, and it was only a matter of time before we had variations on Super Mario Bros. gameplay with hit points, elemental resistances, guns, dice rolls, dozens of special abilities, twenty powerups that do the same thing...

This is what American developers are infamous for doing wrong (don't think Japan doesn't satirize us mercilessly for it). Not that it's always wrong, of course. For genres for which complexity does indeed make play deeper and more interesting--real-time strategy, for example, or sim games, or roguelikes, or any multiplayer game intended to have a tournament scene--this leads to expertly balanced, repeatedly astonishing, Swiss watch-like masterpieces. But does a first-person shooter really need fifty slightly different guns? Is a shmup with eight powerups all that different from a shmup with nine? Why create both an archer class and a mage class for a dungeon crawl if they play exactly the same tactically, aside from that one uses arrows as ammo and the other uses MP? (Don't give me "flavor" as an answer; that's nonsense. Flavor only differentiates those two classes for the five minutes of play, before you get used to the different sound and graphics sets and realize you are doing exactly the same thing.)

I am an American developer. My understanding of games is rooted in chess, Monopoly, Magic: The Gathering, and Dungeons and Dragons; I understand depth by way of complexity. I grew up playing video games, and, of course, all my childhood video game fantasies were built on the games of my youth done better. Sonic with rocket jumping! Platformers with robot protagonists with destroyable limbs, which could be replaced by defeating monsters and grafting them onto your body! RPGs that were part Zelda, part Final Fantasy, part X-Wing Alliance and part Chip's Challenge! Puzzle games with elaborate branching narratives based on your choice of solution! And it took me twenty years to realize this, but while all of these made great daydreams, they would have made terrible games. For one, the sole purpose of those games was to find creative ways to relive gameplay experiences I had already had. For another, they were designed around my playstyle and my experience as a player--not the curious thumbs of a gamer who had wandered in, curious, determined to figure things out for herself and wanting to play her own way. Depth through interesting permutations of mechanics is only interesting to players already familiar with them.

Don't get me wrong, the lessons we as an industry have learned about gameplay mechanics from the early NES days are important. But gameplay mechanics alone are not gameplay. A video game isn't Parchesi, it isn't a novel, it isn't a bunch of guys in wizard hats rolling dice on a table screaming for a 20. It is a thing in which you press buttons and stuff happens on the screen, and yes, we do understand why that is fun. We don't have to speculate and say, no, we don't know why that's fun, but here's some stuff that is, so we'll streamline it and make it part of our game. We need to put down our Warhammer figs for a while, enjoyable as that is, and ride a roller coaster. We need to put away our baseball bats and rock climbing gear for a weekend and watch an ant struggle up a leaf. We need to sit at a desk during a boring meeting with nothing to entertain us but a rubber eraser, and watch what our fingers do with it. We need to get back in touch with the wheeeeeeee.

(Epic has the right idea with Bulletstorm. It's not a game about shooting things. It's a game about bouncing stuff around, with guns merely being the instrument by which that is accomplished.)

This alone is where new, innovative game mechanics arise. Not by playing Super Mario Bros. for the millionth time searching for an as-of-yet-unresolved imperfection or unmined tidbit of genius, not dicking around with variables in the physics engine of a first person shooter until something fun happens, not sitting on a couch with the rest of your team trying to wring out one brilliant idea on the spot as you watch the budget dry up. We have to stop trying to come up with a truly original video game idea just by playing video games.

Come. The weather's great, the sun's shining, and the park is open. Let's go where no gamer before us has dared.

About this Entry
[User Picture Icon]
Date:June 16th, 2011 11:50 pm (UTC)
(Permanent Link)
Nicely written. :) I disagree with a fair bit of your conclusion, but then I like the highly detailed intricate swiss watch games, and want to build those.
[User Picture Icon]
Date:June 17th, 2011 08:41 pm (UTC)
(Permanent Link)
I believe it's a matter of genre. I think one of the reason why a lot of early first-person shooter maps were so tedious was because early designers saw their maps as Parchesi boards pretending to be space stations, not architecture someone would actually have to navigate. (Hence lots of elevators, colored keycards, backtracking through foyers, elaborate 2D symmetries in the overhead map that were invisible in 3D...) But something like Advance Wars--well, that IS a board game.

Less obvious are the games that exist somewhere in the middle, like RPGs. On one hand, Diablo II. Wouldn't work at all without the boardgamey elements. Complex skill trees and forcing the player to do lots of arithmetic to make decisions are half the game, and the results of those decisions profoundly and rewardingly affect the more visceral experience of slicing things apart and blowing shit up in the other half. I used to criticize those kinds of games because five minutes at the level up screen is supposed to deliver enough enjoyment for each hour and a half of killing the same horde of zombies over and over, but some people really enjoy it, and I can respect that. Yeah, so the maps are just randomly placed tiles on the square grid, preventing any non-uniform challenge or true sense of exploration. So what? That's not the most important part of the game anyway.

On the other hand, JRPGs like Final Fantasy VII. Reviewers criticized its predecessors for creating elaborate castles and sprawling overworlds where players got frustratingly lost or didn't know where to go, so FFVII's dungeon designers built almost every non-town map out of single hallways with an entrance at one end, an exit at another, and some jump pads and ladders in between, with a high chance of a random encounter every clock tick. Maybe a fork with a treasure chest for someone with the patience to take the long way around. Each one of these maps is a little game of Snakes and Ladders. It's the Game of Life. It's got a clear objective, a set of meaningful challenges (never the same twice) that set you back if you fail, and a reward at the end. Would be a blast to play with some dice and a bunch of drunk friends at a party. But when you're sitting alone in your room with a soda and you're still wiping the tears away from Aeris's death scene, the last thing you want to do is kill the same three Bombs over and over again trying to get to the end of the next staircase. It checks off all the ticky marks for good game design, but for the wrong kind of game.

I've come to the disillusioning conclusion that the sole reason why I enjoy complex games is because I've already played so many simple ones. Gameplay elements have diminishing returns, and complex games have more gameplay elements. They don't offer more "meat on the bone," they just get me bored less quickly because they switch between a greater variety of challenges I've already mastered--and the joy of designing those challenges for someone else is diminished by the fact that I can no longer even imagine what it is like to encounter them for the first time. I don't think I can create an experience that is new to players unless it is also new to me. To me, "make a D&D dungeon for a low-HP sorcerer character that throws 3x3 AoE fireballs" is a solved problem. Generations of designers have taken it upon themselves to do that, and they've done it very well, but I for one am just so damned sick of throwing fireballs.

If I make a complex game, I want it to be one so outside the realm of typical gaming experiences that even I don't understand how to balance it. And I want it to be balanced by my testers breaking it with moments of absurd utter genius.

Edited at 2011-06-17 08:43 pm (UTC)
[User Picture Icon]
Date:June 17th, 2011 09:11 pm (UTC)
(Permanent Link)
Actually the random maps in Diablo are intentional and in Dave Brevik's mind incredibly important. He feels very heavily that the randomness is critical to creating that true sense of exploration. The world IS what you go and discover, not something that you can look up a map of online. That wasn't a minor part of his design, it was a crucial element. :)

Complexity is indeed a reaction to people having mastered simpler ones. Genres tend to grow in complexity over time as once you've established the language of a particular game type you can express more complex ideas within that genre and the players demand increased complexity to keep purchasing it. This can ultimately lead to the demise of a genre as complexity spirals upwards, driving off new players while retaining an ever shrinking set of genre devotees (because you always have lossage over time).

This is part of where the platform reset of Casual Games and then Social Games was useful, it allowed lower complexity games to gain a foothold amongst new cohorts of players. It is part of why games can acquire new players down at the very young end of the market too.

Good luck with your goal! It is certainly a challenging direction to try and go. Experimenting and trying to find gameplay is hard and expensive. We came up with Iron Construct that way, and there were some interesting experiments along the way. It would have been neat to see how the game could have been if we'd been able to get funding and build the whole thing out. :)

For me, I'll have fun making games I love that are elegant and fun to play, that's what I look for in my personal projects. :)
[User Picture Icon]
Date:June 17th, 2011 09:44 pm (UTC)
(Permanent Link)
Elegant and fun to play is what I'm shooting for too. :] And I am frequently impressed by how even games with well understood mechanics can deliver surprises. (Orbital Defense Platforms + Heavy Carriers = mobile blockade! :D )

I understand what Dave Brevik intended to do with the random maps in Diablo, I just don't think it worked out as well as he wanted. Instead of creating a wholly new world each game, with surprises around every corner, Diablo generated a uniform soup of the same experience done eight or nine different ways. This was true of early builds of Angband, too, which Brevik names as the inspiration for Diablo.

Personally, I'm not convinced that interesting world generation--rather than just noise--is a problem that has been completely solved. The best solution I've seen is in Dungeon Crawl Stone Soup, and that involves adding non-random, carefully designed "vaults" (sub-dungeons) to the pool of dungeon features that can be randomly generated. Which is great, as it's fantastic to unexpectedly discover a gnoll fortress or underground lake as you turn a corner in your random-noise roguelike cavern, but it sort of defeats the point of random map generation in the first place, since you can look up maps of the vaults on wikis and whatnot.

I'm looking at the casual games boom as an opportunity, myself. Ever since I was a little kid all the games I've come up with have been built around complexity. Having the chance to come up with something simple enough to sell for a dollar in the App Store, under the constraints of zero development costs, a very short timeframe, and a single-developer team, forces me to get away from that. And if I pull it off, maybe I can even recoup some of my costs.

I've been hard at work implementing some of my new ideas. It's anyone's guess how the final product will be received. And that excites me.
[User Picture Icon]
Date:June 17th, 2011 09:59 pm (UTC)
(Permanent Link)
Yup, the App Store was another reset, though we can see the years of burgeoning complexity there too now.

Good luck, looking forward to getting to play what you've built! :)
[User Picture Icon]
Date:June 17th, 2011 04:26 pm (UTC)
(Permanent Link)
Submit this to GammaSquad!!!


And, awesome. Awesome!
[User Picture Icon]
Date:June 17th, 2011 05:52 pm (UTC)
(Permanent Link)
Also, I can see this as a speech.
[User Picture Icon]
Date:June 17th, 2011 07:43 pm (UTC)
(Permanent Link)
This appears to be an advertising blog for UGO. While I'm sure there are folks among its readership who would enjoy the baroque screeds of a pathologically frustrated game developer, I don't think they publish this kind of thing.

And 1UP.com (which is also owned by UGO) doesn't do editorials.

Have been thinking about pitching to the Escapist, though. Just need something I can write about in a less bloggy, more journalistic tone.