Tiny Subversions

I played a great casual game this weekend. Diner Dash is a GameLab game. (They're a great company; in fact, I've mentioned them before on this very blog.) Diner Dash defines addictive. The game's aesthetic perfectly captures the feeling of running a diner and having to juggle a million tasks at once. This is achieved through a pattern recognition dynamic relying on a mechanic of queueing (shout-out to MDA!). Anyway, you can download a free one-hour demo of the game. It's worth trying.

I just discovered that there is a video of Mark Healy's GDC presentation of Rag Doll Kung Fu which he gave at the Experimental Gameplay Workshop. This was by far my second favorite game at GDC (after Spore, of course): you MUST check it out, especially if you like cheesy kung fu movies. His demonstration shows more than I can explain in words, and is way better than any of the official trailers for the game. This game is simultaneously innovative and incredibly stylish. See for yourself.

The most obvious technological prerequisite for video games is the existence of computer technology; that is, a machine able to perform a deterministic simulation. While video games may contain an element of randomness, they are almost all deterministic simulations of one sort or another. Before the advent of the computer, creating a simulation game anything like a modern video game would have been next to impossible[1]. This is not to say such feats have never been done; one telling example is Ace of Aces, a three-dimensional simulation of a World War I dogfight, which takes place entirely between two players in small handbooks. Each page of the book is a screen representing a three-dimensional view from the player's cockpit. The game plays out like a massively complex branching story: if the player turns left, she flips the book to page 185. To turn right, flip to page 33. Each page updates what the player sees.

And yet while this game is a flight simulator that does not use a computer, it is of critical importance to note that the game was developed in 1980, well after the advent of the computer. The game borrows critical concepts from the computer, without which it could not exist. Not only does the action take place on a paper 'screen,' but the game uses "a complex formal system to represent the speed, maneuverability, visibility, weapons fire, and other aspects of two-plane air combat" (Salen 425). The type of formal system used to create the simulation of this game is a finite state machine, devised in 1936 by Alan Turing, whose idea gave birth to the modern field of computer science. Ace of Aces takes place entirely within a book: in terms of physical materials, the game certainly could have existed before the development of the computer. Yet the concepts it uses are so bound to the abstract concepts that govern the computer, and there is no way that this game would have existed before World War II, when the use of formal systems to model aviation was pioneered. What the example of Ace of Aces shows is that the idea of a simulation game, and therefore the basic structure of the video game, is inextricably bound to computer technology.

In Rules of Play, Katie Salen and Eric Zimmerman's excellent primer on games theory, they cover an important topic that has been a sticking point for me for a long time. That sticking point is immersion, the idea that a player should forget they're playing a game and instead believe that they are inside the reality of the game itself. But Salen/Zimmerman make the point far better than I can; so I give you an excerpt from their book, titled The Immersive Fallacy. Read it. Know it. Love it. Save it to your hard drive. Forward it to your friends. And then buy Rules of Play if you haven't already.

Immersion is much hairier than we like to make it out to be. For instance, while horror games certainly benefit from immersion, a game like Eternal Darkness breaks all the "classic" rules of immersive play by utterly destroying the fourth wall at critical junctures in play (thanks to Jeremiah for pointing this out to me). In the case of ED, you have a sanity meter. As you encounter more unspeakable horrors, your sanity level goes down. As the sanity level goes down, reality begins to fall apart. At first, you hear crying children. Then the walls start to bleed. Then you hallucinate rooms that aren't there; battles that didn't happen. Finally the game breaks the fourth wall: the screen will turn blue and say "VIDEO" in the upper right corner for 5 seconds before plunging you back into the game. Or the screen compresses to a white dot and goes black, and for a few tense moments, you wonder if the power went out. This is creepy creepy creepy and while it breaks immersive rules, it is simultaneously immersive in that the game is about reality falling apart. If the game's reality begins to encroach on the player's reality--well I'll be damned if that doesn't meet the standards of "immersion" while totally shirking the classical rules by embracing the "gameness" of the game.

Yesterday, during a talk given by a faculty candidate for the Interactive Media and Game Development major, I was drawn into a mini-debate. We were discussing some of the social and cultural ramifications of the way we model the human body in video games (by breaking it down to its components and turning them into abstract polys). Paolo, one of the grad students, said that because we spend more processing time and polys modelling the human face than any other part of the body, this means that the way we model the body has some great humanistic tendency to it. Needless to say, I was pretty pissed off at this implication.

Yes, we spend more resources to model the human face in a 3D game. But take the example of an FPS. While we have these very life-like NPCs in the case of Half-Life 2, the point is that the player is still a floating hand. When I brought this up, Paolo said that you can look in a mirror in HL2 and see yourself. But anyone who understands how an FPS is built from the ground up knows that mirrors are just a cheap trick. Try to do some decent collision detection between yourself and a foreign body in an FPS--let's say bullets whizzing by and grazing your body instead of hitting dead-on. It's nearly impossible because in very real way, the player is just a floating hand with a gun--a spectatorial consciousness whose existence is simply as a force to act upon and manipulate the environment. And while the NPCs are modelled very nicely, what is the point when they exist to be manipulated by the player? At the very least, this certainly does not point to any humanistic values.

So my IQP (thesis) needs finalization. I have lots of interesting chapters, but there's no real coherency to the paper. Here's my reason for writing the paper, although I'm just making this up now:

In essence, the problem that I have seen in the games industry, from the many conferences I've been to and developers I've interacted with, is that game developers take too narrow a view of what they are doing. We are concerned with first-order effects of our games, such as sales, how much fun the player is having, violence we may or may not be directly inspiring, and so on. What about second-order effects? Our games may be promoting an instrumentalist view of reality, a view that is detrimental to rational human progress. Our games may be causing our players to develop a hero complex, a notion that works against ideas of teamwork and promotes mindless individualism. (Note: I approve of individualism, but not in its deprecated forms. I may come back to this later.) While I don't think that video games cause school shootings, I do think that there is something wrong with presenting "kill everything" as a valid solution (or an invalid but glorified solution) to life's problems. Maybe as game developers, we see these problems in our peripheral vision. But our brains go, ack, huge problem! and ignore the issues as either inevitable side effects or as not problems in the first place.

So I've pointed out what I see as the problem that my thesis is addressing. But what does my thesis actually offer? Sadly, I have few answers that game developers can use. But what I really want to see happen as a result of my throwing these ideas out into the world is for game developers to stop ignoring these issues. Even if the issues can't be solved, at least consider them during the process of game development. I'm not saying stop making Doom clones. Keep on making Doom clones, but think about these issues. I believe that if we do this, we may naturally, over time, come up with more interesting baby steps towards solving these gigantic problems.

And that's it. My thesis throws out some ideas, and what I really want is for game developers to think about these. I know for a fact that I will bring up these issues at any company I work for. And I want anyone who reads my thesis, or reads this blog, to consider these issues as well.

So the dilemma is this: how does this make my thesis coherent? It's fragmented almost by design. Any thoughts would be appreciated. My advisor really wants me to pull my writing together.

First, a bit of exciting news for adventure game fans: the demo for Psychonauts is available! The game goes on sale April 20th. Tim Schafer has never let me down--Grim Fandango and Full Throttle are two of my favorite games of all time. This is his first game since Fandango, so I am very excited to see what he's been working on in the intervening 6 or so years.

Tim has moved to the action-adventure model for this new game. I think it's the right direction to go. It's no secret that pure point-and-click adventure games don't sell like they did in 1994. Some blame this on lack of appropriate marketing, blaming the industry for killing the genre. But I think the industry is not promoting these kinds of games for a reason: people don't like point-and-click adventures very much. For one thing, these games lack a strong feedback loop. Typically you spend 20 minutes clicking on random things before figuring out the puzzle and being "rewarded" with plot advancement. In the best cases, like the first three Monkey Island games, this can be incredibly amusing since everything you click on contains a hilarious joke. But in most cases, this is incredibly tedious. Almost always, no feedback = gaming disaster.

This is why I thought Full Throttle worked so well. A lot of people hate the fact that it was so easy, but that's exactly why I liked it. The puzzles were easy enough that you only had to click on things for 2 minutes before figuring out the puzzle and getting a movie. The feedback loop was tighter.

All this points to the general brokenness of the classic adventure game. I am really excited to see what Tim Schafer will bring to the now-boring action-platformer.

The following material is a section of my IQP (read: thesis on technology and society) that makes some sense without the context of the rest of the paper. No offense meant to Will Wright or any of the excellent folks at Maxis--I think they produce some of the best, most innovative games ever made. But this fragment is postulating a criticism of The Sims 2 that I have never seen, one that I think is important for people to be aware of.

Andrew Feenberg's wonderful Critical Theory of Technology contains a passage that resonates strongly with a recent trend in video games. According to Feenberg, in a machine society, we encounter the "new narcissism," which is an "intensified pursuit of personal pleasure by individuals who have less identity than ever before." Via "the collapse of public [and family] life" the "individual becomes a discontented spectator on his or her own life, engaged in strategies of manipulation and control directed toward the self and others alike […] the subject [is] now in the position of programmed device, now in the position of programmer." (98) This passage resonates strongly with The Sims, and more precisely its sequel, The Sims 2, both a part of the famous series of people simulators or virtual dollhouses created by game development company Maxis. The concept central to the game design of The Sims 2 is that of human interaction modeled as a technologically deterministic system. This can be traced back to the common design assumption that actions taken by the player should have immediate, obvious consequences. Will Wright, the Lead Designer for the original game, elaborates that "you can roughly look at all of [Maxis'] Sim games and divide them into one of two categories: the economic ones and the biological ones. And, in general, the economic ones have always [sold] better." Wright's game SimEarth involves players managing an evolving planet. The game used a very scientifically accurate model of the Earth, yet players would often find themselves confused. "The biological systems [like SimEarth] tend to be very soft, squishy" systems that respond in a complex, high-order manner to a player's actions. In the economic games like The Sims, players find it easy to "assign credit to [their] failures" and develop a strategy for winning the game. (Rouse)

A Sim is one of the simulated people that populates the game, and is the avatar through which the player interacts with the game itself. Players can control a whole family of Sims, or to some extent a whole neighborhood of Sims. By and large, a Sim is defined by his or her actions. While a Sim has internal thoughts which are represented by thought bubbles that appear above his or her head in a cartoon style, and while a Sim has default predispositions and a "personality" (neat/sloppy, active/lazy, nice/grouchy, etc.), the strategy of the Sims comes down to action. This is a video game, after all.

The internal emotional life of a Sim is almost nonexistent in terms of the game model. Though Sims do have "memories," these are represented as wholly separate from the main game and it is difficult to see the impact of these on gameplay except in very slight ways. Interpersonal emotional life is reduced to the Sims pushing each other's buttons, in a literal sense. The Sims interact with objects and people in the same way: the player clicks the mouse on an object or person, and a ring of buttons appears. In the case of an object, the choices may be "eat," "drink," "use," or some similar action whose context depends on the object itself. In the case of another Sim, buttons include actions such as "play," "chat," "irritate," "flirt," "hug," "entertain," and so on. That Sim may or may not refuse to participate in your action. So Bob's interaction with Ming is walking up to Ming and in effect pushing the "chat" button.

The other important characteristic of a Sim's actions is that the actions are queued up. If the player says, "eat lunch, trim the weeds, study cooking, go to bed," the four actions are indicated in a queue at the top of the screen in the order in which they were assigned, and the Sim proceeds to carry out the actions. It is possible to plan a Sim's entire day like this, although the difficulty of dealing with spontaneous events makes this an unadvisable game strategy.

A Sim's intrapersonal emotional state is represented through the use of a number of status bars. These bars represent concepts such as "hunger," "bladder," "energy," "fun," and so on. If a Sim isn't fed, his or her "hunger" meter goes down. By the end of a busy day, a Sim's "energy" meter is low, and can only be replenished by sleeping or drinking coffee.

In addition to status bars, the Sim's behavior is dictated by "wants." Every morning, the Sim wakes up with a new series of "wants," which represent short-term or long-term goals. The goals can be simple desires such as "eat a bowl of cereal." However, most wants are based on career, relationships, self-improvement, or consumerism; respectively, a Sim might want to get a promotion, to tell a joke to her friend George, to improve her creative skills, or to buy a table costing at least $1,000. If you achieve a "want" in the course of a day, your Sim gets "reward points" which can be saved up to spend on fantastic rewards, like a pair of glasses that make the socially awkward cool, or a money tree, or a fountain of youth. Relationships are measured in a similar way. There are status bars for long-term and short-term relations; if you have a spat with your best friend, short-term will read low but long-term will still read high.

Thus is the automated life of a Sim. Happiness is typically achieved either through Feenberg's "manipulation and control directed toward the self and others" or, more subtly, through the Sim's mastery of his or her environment--the house in which the Sim lives.

While the ramifications of the automation and inherent consumerism of the Sims has been a topic of discussion on the Internet, possibly the most important aspect of the game remains the most overlooked: the control of time. Perhaps this feature has been glossed over simply because it was inherited from earlier Maxis games like SimCity. Perhaps the feature is treated as such a natural part of the game that it only registers subconsciously; or it is possible that the "human" features of the game necessarily steal the spotlight in the context of any critical discussion of The Sims 2.

Whatever the reason for its low profile, player control of time in The Sims 2 is critical to the flow and structure of the game. The mechanic for time control is simple: in the lower left-hand corner of the screen are four VCR-style buttons. These buttons pause the game, allow for a normal playing speed, a fast speed, and a faster speed. The buttons serve two functions. First, the pause button is used to allow the player to queue up more actions for the Sim to complete without wasting any time in between. The Sim wakes up in the morning, the player pauses the game, and orders: use the toilet, take a shower, change your outfit, eat breakfast, read the newspaper, go to work. Then the game is set to normal speed, and the player watches the Sim carry out these pre-programmed actions. However, watching the Sim's morning routine can be boring, and here is where we meet the second functionality of the time buttons: skipping over boring actions. As soon as the Sim starts the morning actions, the player can hit the button for fastest speed, and now the morning routine is compressed from what might be 5 minutes of the player's time to about 10 seconds.

When combined with the queuing of actions, pausing and time compression will lead to the player developing strategies for attaining happiness entirely unlike those a person in real life would undertake. One such strategy is building strong relationships in a day. Bob invites his acquaintance Ming over in the morning. The game is paused, and the player queues up: "chat with Ming, tell joke to Ming, admire Ming, chat with Ming, chat with Ming, play with Ming, tell joke to Ming, serve lunch, chat with Ming, play with Ming, use the toilet," and so on. Once a sufficient number of actions is queued, the player can then hit the fastest time compression button, and in the course of one minute of play, Ming has become the player's friend.

Strangely, the game automatically pauses time whenever the player wishes to purchase a product for a Sim. If the Sim indicates that she wants an exercise machine costing more than $500, the player clicks on the "buy" icon. In a scene reminiscent of the movie Fight Club, time freezes and all items in the house, upon a examination, have a price tag attached to them, indicating what model they are and how much they can be sold for. Additionally, a window appears that emulates the behavior of online furniture, clothing, and media catalogues—a kind of Amazon.com for the Sims. In the world of The Sims, purchasing a commodity is painless process with no cost to the efficiency of a person's day. As soon as the desire manifests, the need can be fulfilled (provided the Sim has enough capital).

The simulated world in which the Sims live is precisely the postmodern lifestyle of "disposability, novelty, and […] instant obsolescence" postulated by Harvey, taken to a logical, inevitable extreme. The game gives the player control over time to more efficiently allocate consciousness of actions. The game collapses space by allowing any and all friends of a Sim to show up at the house exactly 5 minutes from when they are invited (provided that the invitee is not engaged in wage slavery, which engenders an irritated refusal from the invitee).

Works Cited
Feenberg, Andrew. Critical Theory of Technology. New York, NY: Oxford University Press, 1991.
Rouse III, Richard. Game Design: Theory and Practice. Wordware Publishing, 2001. 19 Jan 2005 <http://library.books24x7.com/>.

by Darius Kazemi
Copyright 2005

So I made a scary decision. I have decided that grad school is not for me, and that what I really, really want to do is go into the games industry full-time. My reasoning?

The original plan was to go to grad school, get an MS in EE as a safety net, and then jump into games. But I have realized that the whole concept of a safety net is based on the assumption that there is a good chance I won't succeed. And it's horrible to make a huge life decision based on underestimating myself. I know I'll be wildly successful in this industry. I have no doubts. So I should just jump in. Even if the grad school thing worked out in the end, I do not want to look back on my major life decision and say, "That was the safe, cowardly way to do it."

Getting a job in the video game industry isn't hard for me. I already have one offer, and if I applied for more shitty jobs, I'd probably get those, too. Getting the right job is the problem. Fortunately, I have contacts coming out the ears, and lots of people willing to help me--in fact, lots of people already have helped me. I have one standing offer as a code monkey. But I'm not the best coder in the universe. I'm a good programmer, a good designer, even a decent audio guy, but my skills that really shine are in communication, organization, and networking. And I know way more people in this industry than a someone my age with my limited experience rightly should.

So... here we go!

I'd like to point out that the hypothetical Civ-like game with good modding tools for educators mentioned in Darren's comment is under development at Muzzy Lane.

Muzzy Lane’s inaugural game, represents an innovative new direction for PC strategy games. Using accurate historical data, Making History™ puts you in the role of an actual head of state, leading your nation through real historical challenges. Everything flows from your decisions. For every action you take, the game’s sophisticated AI and dynamic world models challenge you with unique, historically valid consequences. In-game Advisors (based on actual historic figures) give you intelligent advice on military, diplomatic, domestic, and economic options. You negotiate with other players, prepare your plans, make the choices, and take action.

Eric Zimmerman and Nick Fortugno, both from the venerable GameLab, just published a really cool article in Gamasutra. The article is about the development of educational games, i.e. games for learning. Among the great things they say:

Everyone - both developers and educators - forgets this one: making games is really hard. Even creating a wholly derivative game (a blow-by-blow clone of Bejeweled, or You Don't Know Jack, or Tomb Raider) is incredibly difficult to do well. When you add to this the ambition of creating an innovative game with new kinds of content and gameplay, as well as a game that actually tries to teach something meaningful to players, the problem is multiplied by orders of magnitude.

Amen to that. Making an educational game that does an amazing job of teaching students should, by all rights, be way harder than creating the next blockbuster AAA title. Which ties into the next point:

This is why we are skeptical of many educators' claims that given access to the latest game engines, they will be able to create top-notch educational games and succeed where everyone else has failed. It's simply not going to happen. Tools by their nature limit as least as much as they liberate, and creating innovative games on any scale usually means coding them from scratch. That's not to discourage educators from getting into game development. But all sides that want to get involved need to recognize the challenges and demands of making games.

Amen to that, as well. If I were an educator looking to make games for learning, I wouldn't even start with a relatively simple 3D engine like Torque. What I would do instead is concentrate on 2D and really delve into the game design and the pedagogical aspects of what's going on. Once I felt very comfortable with 2D, maybe after making three complete games, only then would I venture into 3D.

What do you guys think? I think that what Eric says definitely holds true for educators working on their own. What about educators working in collaboration with professional game developers? Part of me thinks that educators don't understand enough about games, and game developers don't understand enough about education, and you'd need some kind of breakthrough weekend to sort it all out before development even begins.

First off, I think that Darren offered up a nice defense of what his actual position concerning The Sims is. He brings up the idea of The Sims as sort of a pointer or gateway to richer gameplay styles, I think kind of going in the direction of Craig's idea of player-content orientation. Though I think that Darren has the single-player and not the multiplayer experience in mind. I'm waiting to see Darren post something more substantial about The Sims 2 on his blog--I know for a fact that he's been working it over in his mind for months now.

Solipsistnation mentioned something really interesting about multiple win conditions, particularly win conditions that involve a "don't violate your ethics" clause. I think that for a lot of live action (or even tabletop) roleplayers, this is a solved problem. We have a GM who can think critically about ethical situations, who can improvise challenges to the player's ethics, and who can understand something about who the player is in real life and incorporate that into the game choices.

I also love his point about "real multiplayer." The implication being that we have fake multiplayer right now, because we really can't share the spotlight yet: everyone has to be the hero. He said that nobody wants to be fodder--the brilliant thing about most LARPs is that nobody has to be fodder. In some LARPs the concept of fodder doesn't exist, yet the concept of a main character doesn't exist, either. Everyone is important, but everyone realizes they're not the center of the universe. Again, this because the games are run by human minds capable of critical thought ; a computer cannot do this.

A few questions come out of this:

1) Do most players see themselves as the hero in multiplayer video games? If this is the case, then why?

My quick response is that this may have to do with the way the game is visually and conceptually framed for player. If you're the all-important "user," with a UI centered about you, of course you are the hero! What if we included interactions in multiplayer games that forced existential confrontation between the players, sort of the equivalent of looking a man in the eye before you shoot him? Craig has written some excellent stuff about player vision; I'm wondering if he's considered the effect of player vision on player ontology. If I'm not mistaken, Merleau-Ponty has done some work in the vision/ontology area, although well before video games existed.

2) Do we need to solve the "strong AI" problem in order to get noninstrumentalist games? People can do it, but computers can't. The first instinct is to make computers as good at critical thinking as people. I'm not really interested in questions of whether or not we can do this. Let's assume we can. The real question is, will this help our games get any better? Do we need to? Or should we focus more on multiplayer LARPs?

Instrumentalism, in the context in which I'm discussing it, is the tendency to view all entities as means to an end, as fundamentally neutral instruments to be played or manipulated. Technological instrumentalism is the unsettlingly common idea that technology is fundamentally neutral, and it is only "good" or "bad" in the hands of a "good" or "bad" person. This is the kind of theory that leads to absurd declarations like "nuclear weapons are neutral, they can be good or bad depending on who has them." My opinion is that a nuclear weapon is decidedly "charged" from its very conception, but that's something a little bit off-topic.

Anyway, the theme of the Indie Game Jam this year was "people interacting." Almost all of the games (at least the ones that were truly about people interacting) took an instrumentalist view of humans and human social relations. For instance, consider Chris Butcher's game about a high school girl moving up in the social pecking order. The game presents the NPC girls as objects to be manipulated: befriend a loser and dump her sorry ass in front of a cooler kid to gain the cooler kid's respect, and increase your cool in the process. Charles Bloom's game puts you in the ultimate treat-people-as-objects role: the manager of a software development project. The employees generate productivity (or the lack thereof) and you interact with them to increase the productivity. Chaim Gingold's game, Mortal Retail, had the same arcadey manage-the-influx-of-people feel to it in the same way that Dan Ogles' manage-a-bar game did.

The games that didn't treat people as objects were pretty much absurdist statments and not games. Ranjit's Waiting for Godot game is an example thereof.

Really almost all the games used the idea of people-as-resources as the central motif. But of course, that's what games are all about, right? Manipulation of resources to get what you want, the "win" condition? Maybe the problem here is that these games run on a computer. Even modern computers are still Turing machines, and the Turing machine itself is about manipulation of your current state to approach a final state. While many non-electronic games use this manipulation theme, there are also plenty of social games, such as LARPs about storytelling, that are more about constructive imagination than manipulation.

Is there any way we're ever going to get past the instrumentalist nature of games? I don't think there is any way we will truly see single-player video games about people interacting until we find a way to ditch this people-as-resources rut we're stuck in.

Proof that I am Nerd Supreme: I had a dream last night that I became friends with Greg Costikyan.

A little background: Greg has been in the industry for a gazillion years, starting with the pen and paper games of way back in the day. He has worked on a million zillion excellent titles. This weekend was Gaming Weekend at WPI, and I was running Ghostbusters, an RPG by West End Games. Greg used to work at West End, and is credited as an editor on the game. He also was a co-creator of the excellent Paranoia. I had also seen him at the Game Developers Conference recently. So all in all, I guess he's been on my mind for last two weeks.

Additionally, I think the reason why I dreamt I became his friend is simple: at GDC2004, I spoke to him briefly and told him I liked his games. He was shy or busy or something, so he said "thanks" but kind of turned away instead of engaging me in conversation. Folks, this is proof that I am not the Greatest Networker Known to Man. So in this dream, we were sitting around a coffee table, chatting about the game industry and design, and having a grand old time. Hah!