[The Final Dakar]

I find we like to talk about how games work and don't work more lately, and the News thread isn't the most ideal place to do it. I found a fantastic article on Gamasutra about design and thought I'd post it in full here to kick things off.

Ed Fries Argues For The Artistic Necessity Of Constraint

At the first day of the Gamasutra-attended Montreal International Game Summit, entrepreneur, former Microsoft exec and hobbyist Atari 2600 developer Ed Fries argued that imposing artistic constraints may be the way forward for an industry currently producing far too much me-too product.

Fries opened up by saying, "If we want to make the video game business an art form, if we want to make art, it follows that we have to figure out something about beauty. I had this close encounter with beauty recently."

He had that encounter while working on Halo 2600, an adaptation of the popular shooter as a game for the Atari 2600. Unveiled at this year's Classic Gaming Expo, it's fully functional and has even had a limited production run on cartridges.

Joking, Fries asked, "How many of the audence are programmers? Oh good, we've got a lot. I'm sorry, to the rest of you."

Originally, Fries read Nick Montfort and Ian Bogost's Racing the Beam, a book on the Atari 2600. It got Fries "really excited" to try programming for the system, as he'd coded games for the Atari 400/800 computer system in high school and college.

"On working on [Halo 2600], I had a weird encounter with beauty and it made me want to talk about it with other people," said Fries. This encounter was with the elegance required in programming to the system's incredibly limited capabilities.

Released originally in 1977, the Atari 2600 was created to replicate Pong and Combat, and is incredibly constrained. "Part of my talk is to talk about constraint, and if you're a programmer and you want to work on a machine with some constraints, this is a great one to work on," said Fries of the system.

Fries related these constraints to art history -- particularly Greek vases, another of his interests. Vases created during "the pinnacle of Greek culture" are beautiful in their simplicity. Though later artisans created more complex techniques which allowed for more color, Fries showed two examples, earlier and later, and asked the audience "I have to ask you which is more beautiful -- this, or this?"

His clear choice was the vase created earlier, but with the tighter constraint of highly refined monochrome art.

Four-Color Complexity

The Atari 2600, meanwhile, forces the entire program to fit in 4K of ROM, can only display four colors at a time, and can only execute 76 instructions per line it draws on screen -- meaning all updates have to happen in a short span of time.

Fries took the audience through the process by which he used trickery in code to reduce instructions and create simple, elegant instructions. He called his original impulses the "dumb way" of doing things; the evolved one the "Atari way." "Look how beautiful that code is ... it has that nice, i don't know, purity to it," he said.

Even with these constraints, Halo 2600 is complex. "You fight multiple bad guys, you go into the land of the giants where everything is big, you go through 64 levels, and you fight the boss."

The boss is much larger than Master Chief, and was barely possible given the system's limitations. "If we hadn't saved cycles here, and if we hadn't saved cycles here, this whole scene would not have been possible. I could not have done the boss the way I wanted to -- if I didn't have this beautiful code."

Said Fries, of working on the 2600, "I have to do beautiful work or I can't do work at all."

Originally Posted by Artificial Constraints

He looked at one of Bach's fugues -- a highly constrained form of composition. "Why would he put himself in such a constrained environment? It got me thinking about this idea of constraint and art. Why do artists in other forms put constraints on themselves? The last 30 years of gaming have been about taking away constraints. As we saw from those Greek vases, progress is a funny thing, but progress and art don't necessarily go hand-in-hand."

Fries showed the audience a slide of a paper dragon -- and then revealed the complex figure had been created by folding a single sheet. "Something really weird happened there," he said. "I told you it was a paper dragon and you were like 'Oh, okay,' and then i told you the constraint, and it became more beautiful in some way."

Said Fries, "In other artforms, people are putting these artificial constraints on themselves. The origami guys are working with one sheet of paper, poets are using meter and rhyme..."

In painting, he said, after hundreds of years of refinement, "the artists got to the point where they could do whatever they wanted, and paint whatever they wanted. What happened? In a way, art got really boring. When everybody can paint reality, everybody could paint the same thing."

In art, realistic still lifes gave way to forms like impressionism, post-impressionism, cubism, abstract expressionism. And in games, he says, today developers are all creating the same thing -- hyperrealistic shooters that all refer to one another in form and content.

It's time for a breakthrough, he argued -- and imposing constraint could be the key. "As a way to go forward, as a way to avoid the sameness that is happening to our games. I think that in some ways some teams are starting to do this, even if they aren't thinking about this the way I am thinking about this."

Important Opportunities Ahead

Three recent examples Fries ponted to are Kirby: Epic Yarn, in which the entire world is built of yarn and fabric, Minecraft, which is made entirely of blocks, and MadWorld, which only used black, white, and red as colors for its graphics.

"Maybe we've had it wrong. Maybe we've been in this rush to get rid of these constraints that we thought were so limiting to our progress going forward in the gaming business. Maybe there's something to be said for constraint," he concluded.

"If you get rid of the constraints it's kind of boring." He also noted that, as platforms improve, "We have to put artificial constraints on our work because we're losing the real ones." And he advised developers to pick a constraint that gives "a unique look in the marketplace."

There are three key opportunities which Fries suggested constraints can enhance when making games:

1. "They can create an environment in which things can be done really well." This goes back to his experience creating code for the 2600 -- his final code, after many revisions, was much, much more efficient than what he'd programmed at the outset. "Once you have the constraint on the code you can create code that really matters," said Fries.

2. "This creativity you don't get otherwise. The constraint forces you to make decisions you would not have otherwise made," he said.

3. "It can leave space in your work, leave room for interpretation in your work," he said. Poetry leaves much unsaid, and the same can be said for other constrained forms, he argued. If someone were to make a game with only two colors, he said, "then maybe we could make something as beautiful as the Greeks made 2500 years ago."
"Constraint is happening in some ways already. People are doing these 24 hour game jams -- that's a great example of putting an artificial constraint on themselves," said Fries. A student told him recently that his best idea had been born in a game jam project, he noted.

An audience member asked if the ultimate constraint for games is interactivity. "I think that it's true that making games is harder than any other medium really, because the player has control. But I think we're getting good at it too. My thesis is that we're getting to the point where artists got to, that it started to get sort of blah -- because everything started to look the same. I think maybe constraint can lead us away from the blahness."

I'll add my thoughts when I have time later today.

[The Final Dakar]

Alright, finally had time to put some thoughts down.


Personally, I think the concept of constraint is both overrated and dishonestly used here. One of the examples the guy uses is paintings getting too realistic, because people got to good at it. I don't think that was the entire problem though. The problem was eventually photography came along, and started doing what painting was doing, necessitating for painting to evolve.

If anything his Halo 2600 reinforces my idea: it's not possible unless you've become really good at coding. But does it make a good game in and of itself? Is Halo 2600 (which I haven't had a chance to see) good because its a fun game, or good because of what it has accomplished? Would it be relevant or worthwhile without a modern comparison? Would it stand-up in 80s?

To me, it seems, he measures too much of the game by how it was achieved rather than content (be it design, music, gameplay, etc.). His measure can and is valid, but in a more backwards looking perspective (i.e. when we talk about FPS of the past, GTA III, Sonic the Hedgehog's speed on Genesis).

To cut to the real chase, however, I don't see constraints (limitations in hardware he alludes to) as leading to better games. If this were the case, wouldn't the Nintendo Wii, XBLA, and handhelds be at the forefront of creative gaming? I believe I can dismiss any argument for the Wii as coming down to Miyamoto's genius and them removing a constraint by adding motion control. Handhelds? The most creative game I can think of on DS, Scribblenauts, could easily go to consoles if it weren't for their constraint in lacking a stylus. That leaves XBLA. There's potential there, but I wonder how much of it is nostalgia (see: Shadow Complex, Trails HD). In the end, isn't that why Halo: 2600 is relevant? Nostalgia for the tool?

But let's go back to modern games. What about BioShock? Is that not a perfect counter-example? Is there anything about that game that would be improved by moving it to worse hardware?

Honestly, the industry just needs regular constraint. A move to removing the kitchen-sink attitude that comes to game features and size. The shallowness of the industries artistic ability doesn't lie in how its made but by who controls whats being made.

My 2¢

[mattyb]

Originally Posted by The Final Dakar 

Alright, finally had time to put some thoughts down.


Personally, I think the concept of constraint is both overrated and dishonestly used here. One of the examples the guy uses is paintings getting too realistic, because people got to good at it. I don't think that was the entire problem though. The problem was eventually photography came along, and started doing what painting was doing, necessitating for painting to evolve.

If anything his Halo 2600 reinforces my idea: it's not possible unless you've become really good at coding. But does it make a good game in and of itself? Is Halo 2600 (which I haven't had a chance to see) good because its a fun game, or good because of what it has accomplished? Would it be relevant or worthwhile without a modern comparison? Would it stand-up in 80s?

To me, it seems, he measures too much of the game by how it was achieved rather than content (be it design, music, gameplay, etc.). His measure can and is valid, but in a more backwards looking perspective (i.e. when we talk about FPS of the past, GTA III, Sonic the Hedgehog's speed on Genesis).

To cut to the real chase, however, I don't see constraints (limitations in hardware he alludes to) as leading to better games. If this were the case, wouldn't the Nintendo Wii, XBLA, and handhelds be at the forefront of creative gaming? I believe I can dismiss any argument for the Wii as coming down to Miyamoto's genius and them removing a constraint by adding motion control. Handhelds? The most creative game I can think of on DS, Scribblenauts, could easily go to consoles if it weren't for their constraint in lacking a stylus. That leaves XBLA. There's potential there, but I wonder how much of it is nostalgia (see: Shadow Complex, Trails HD). In the end, isn't that why Halo: 2600 is relevant? Nostalgia for the tool?

But let's go back to modern games. What about BioShock? Is that not a perfect counter-example? Is there anything about that game that would be improved by moving it to worse hardware?

Honestly, the industry just needs regular constraint. A move to removing the kitchen-sink attitude that comes to game features and size. The shallowness of the industries artistic ability doesn't lie in how its made but by who controls whats being made.

My 2¢

Ed Fries is a 'system' guy, not a commercial or marketing type. He's in games, I'm in databases and we strive for games/applications to be efficient, not greedy for resources. He was limited by the hardware that he was programming on (and yes I think that it was done for a typical hacker reason 'itch that scratch' or lets see if I can solve that problem) and not for any business requirement. Nobody who is programming for the Xbox or PS3 is only going to program to use 50% of the CPU or 50% of the GPU. No, they're going to exploit it as much as possible. Just like in my field where we have gigs of ram for Java, gigs of ram for the database, gigs for cobol and tuxedo and the app server. Business hardware is now relatively cheap, and if we paid for it then we should use it. I presume that development managers in the game industry are thinking the same.

The game jam stuff is basically brainstorming, and the constraint is nothing new to any field : its time. Same constraint that we all run into day in and day out.

I can't see how you can sell the 'we must restrict ourselves artistically' to arseholes like Bobby Kotick when they've just had 12 months of dream sales. But then I know nothing about the gaming industry.

[The Final Dakar]

It's strange, that he's a system guy trying to make an artistic argument, but then again, that could explain why I find so many flaws in his argument.

Either way, I obviously found this way more interesting than everyone else on here, maybe because I studied art and art history in college.

[The Final Dakar]

The Tip of the Sphere

Here is the paper design I wrote for for the Halo 3 Sniper Rifle:
Sniper Rifle
Role: Long-range instant-kill sniper rifle, but reloading makes it hard to use

• Two zoom levels (2x – 7x)
• Reloading unzooms
• Magazine of four quick shots, with slow reload
• Does headshots, even through shields
• kills any biped in one shot (even Players)
• [anim] special death animation for headshots
• kill shot accelerates units
• Does massive damage
• kills a Player in two body shots
• kills small bipeds in one shot
• Over-penetrates flesh, glass and soft materials
  
• ~0.5s delay before un-zooming to reload after the last shot of a magazine, so you can see the result of your shot

The bold point seems so simple yet does so much.

[The Final Dakar]

Massive dissertation on the newly opened Grantland.com
Tom Bissell reviews L. A. Noire, the latest hit game from Rockstar Games - Grantland

Video games can do a lot of things other storytelling mediums cannot. Its penance, however, is to have to deal with things foreign to other storytelling mediums, one of which is a uniquely damaging form of audience disruption. Just about every storytelling game employs various masking systems that attempt to anticipate internally disruptive player behavior.

while playing through the rest of L.A. Noire the following question was never far from my mind: How big of a problem is it that players can effectively screw up video-game stories? It is a question that is never far from my mind when I am playing any game whose fiction works in tandem with my decisions to create something thematically unified and dramatically satisfying. So, how big of a problem is it? One answer to this question is: There is no answer to this question. Another answer is: Strong interactive fiction will compel players to behave in ways roughly analogous to how the interactive fiction's author intends them to behave. Another answer is: The whole purpose of interactive fiction is to encourage this type of crisis. Another answer is: This is precisely why the video-game medium is incompatible with authored forms of storytelling. In the past few years, I have thought about this question a lot — maybe more than any other question, in fact.

Those who are immune to the pleasures of video-game storytelling argue that games have far more in common with music and visual art than film and literature. According to this view, games are primarily rule sets or interactive systems, and it is in these arenas where the true art of video games resides. This is undoubtedly true, but must it be an excluding truth? No one, after all, ever gathered around a campfire to hear a rule set. It may be that our existing storytelling models are, in many ways, ill-suited to video games. Perhaps the video game medium is not a storytelling medium at all but an experiential medium in which storytelling possibilities are allowed to occur.

. Now, driving from one end of Los Angeles to the other is not exactly storytelling. Neither is walking Phelps up a hill, across a warehouse, down the sidewalk, or into a travel agency. You do a lot of bullshit in video games generally, but in open-world games you do an extraordinary amount. This creates a fascinating tension that I believe is unique to video games, and it is a tension that open-world games particularly inflame. Most video games are defined by arbitrary hindrances, such as restricting which buildings you can enter, and magical allowances, such as regenerating health. One might think this would foreclose any expectation of narrative realism on the part of the player. For many players, though, it is the opposite.

helps' method of investigation is like a game of Bad Cop, Terrible Cop. His loose-cannon tendencies began to worry me, and soon I was basing my interrogation decisions not on the actors' occasionally hammy nonverbal performance cues but on whether I could trust Phelps not to lose his shit. Thus I wrongly accused numerous shifty-eyed suspects of lying when I should have merely doubted them, blew countless leads by stupidly believing lies, and alienated dozens of friendly witnesses when I doubted statements I should have believed. If you regard the interrogation aspect of L.A. Noire as a game, you are going to have a terrible experience, because, as a game, it is not at all enjoyable.

...

During interrogations, you are not supposed to control Cole Phelps. You can only guide him, as one might guide a slightly crazy boxer.

The dialogue in a game like L.A. Noire is probably incapable of genuine emotional consistency, and its branching possibilities guarantee that some transitions will be combative. L.A. Noire's dialogue is also burdened with having to send actionable signals to the player. Scarcely any exchange in L.A. Noire does not, in some way, contain actionable material. Freer, more naturalistic dialogue has no obvious place in a game in which virtually everything that is said must give players important motivational and directional cues.

In what is probably cosmic justice, the elements intended to prevent the game from being boring are easily the game's most boring elements. Take the "street crime" side missions, which typically end with a cut scene that shows Phelps shaking his head while the body of the guy he just shot is lifted into the back of the coroner's hearse. When Cole Phelps has shot down his third perp of the day, this bookend of putative sobriety achieves an inadvertent hilarity.

Equally hilarious are L.A. Noire's fist fights, the gameplay element most impressively devoid of interest. After the game's seventh or eighth fist-fight, I ... popped into my Xbox 360 a game about which I had been curious for some time, 50 Cent: Blood on the Sand, which some regard as a great twinkling light of video-game preposterousness ...The object of the game is to kill everything and earn money ... which makes 50 Cent: Blood on the Sand a rare game whose authored narrative is not at all disconnected from its player-generated narrative ... I played the game through in two days, after which I wondered if the single most damning thing about video games is the fact that one could argue, legitimately, that 50 Cent: Blood on the Sand is a better game than L.A. Noire.

When I returned to L.A. Noire, I made the decision to stop thinking of it as a video game. Very quickly, something happened. In my notes I jotted down "L.A. Noire: the gay teenager of video games — i.e., 'It gets better,'" but this was a flippant way to process a real attitudinal readjustment on my part. When I stopped thinking about him as someone with whom I was supposed to feel any kinship, Cole Phelps became a deeply compelling character. You begin the game strongly assuming that Phelps has a Dark Secret, and he does. But that is not the most interesting thing about him. The most interesting thing about Cole Phelps is that he is an asshole who might also be insane. ... Once you accept that Phelps is not your avatar but a guided missile whose damage you are constantly trying to mitigate, L.A. Noire gives you an experience unlike any other game I have played.

[The Final Dakar]

Another great piece from Bissell.
Tom Bissell on the making of 'Madden NFL' - Grantland

True fact: The modern football video game pushes current-generation technology to its limit. Why are football games so "expensive," in the sense that programmers use that word? Well, first, a football game has to render all the players, all of whom have idiosyncrasies of movement and appearance that must be accurate, and, in the case of marquee players, downright meticulous. Second, the game has to render all the coaching staff and the refs. There's also the crowd to render, not to mention the crowd noise, which is keyed to surprisingly complicated crowd A.I. Let's not forget the grass on the field. Or the light. Or the broadcasting. The game's also scripting, on every play, the individual behavior of every player on the field, most of whom will be doing different things on any given play. In a basketball, hockey, or soccer game, the range of behaviors is more limited. In basketball everyone's doing roughly the same thing from an A.I. perspective. Hockey gets a little more complicated, and soccer a little more complicated yet, but football gives you 22 individual actors obeying a wide range of A.I. scripts. Not to mention the fact that every NFL team has an elaborate playbook with distinct tendencies and play styles. Meanwhile, during the plays themselves, there's tons of contact between those 22 individual actors, all of which they have to respond to. And this has to look good — seamless, even. When you start pondering the immense complications of a game like Madden — the product of more than 10 million lines of code — you begin to wonder how the game even runs without shooting fire out of your console.

One aspect of Madden that Coach is particularly fond of is how it changed the way we watch football at home — this coming from the man who popularized the telestrator, which allowed casual fans at home to understand, as a coach or player would understand, what was happening on the field. "I really knew [what Madden accomplished] when I was at Fox," Coach told me. "David Hill, the president of Fox Sports at the time, had a meeting with a bunch of us, and he said, 'What we want to do is make our game on television look like the video game.'" Hill didn't have to say which video game he was referring to. The visible lines of scrimmage floating beneath players' feet? The forward-pointing yardage arrows? This is visual language drawn directly from Madden. How professional football is played from year to year is reflected in Madden, and how Madden is experienced is reflected in how professional football is watched. That cannot really be said about any other sports-game franchise.

Several people told me that the Madden franchise is regarded by the NFL as the league's "33rd franchise." This derives from the fact that, at the end of every week, EA Tiburon, along with every NFL team's coaching staff, is sent a massive searchable database of film in which every play of every game is broken down by its situational peculiarities. If you want to see every play in which Drew Brees was facing a third-down conversion of more than 10 yards, or see every instance in which Rashard Mendenhall ran for a loss, you're in luck. This NFL "black box" is how opposing teams scout one another, and it's how Madden game-design devs identify and develop team and player tendencies.

Maybe the most interesting person I met at EA Tiburon was Donnie Moore, who's in charge of the ratings system to which every virtual Madden player is subject. When I learned that Moore was largely alone in overseeing this much-contested feature of the game, I had two visions. The first was of a stern older gentleman in a large office, a glass of water on his otherwise empty desk, wearing a bow tie and staring down all visitors with icily piercing blue eyes. The second was of entering a room dominated by a huge computer screen, and saying, as I approached, "You're ... you're a machine."

Instead, Moore is a 34-year-old EA veteran who sits in a dark cubicle on one of EA Tiburon's two American football floors. He talks loudly, quickly, and very, very entertainingly about the Madden ratings system while occasionally pausing to drink from a can of Pepsi that I strongly suspected was neither his first nor last of the day. Moore has the largest number of Twitter followers of any of the Madden devs, not surprisingly, and he calls himself "an artist with the numbers." I think there is something to be said for this. There's probably no one else in the country doing what Moore does, or at least no one doing it for the same purpose. Lots of people do what Moore does, actually, but they are one of three types: scouts trying to bring honor to themselves while enriching their organizations, analysts trying to bring honor to themselves while enriching gamblers, or fantasy-league freaks trying to bring honor to themselves while triumphing over their fellow freaks. But Moore doesn't get anything extra if he's accurate. According to him, his ego is not on the line, at least not in the same way the ego of a scout or traditional analyst would be. The easiest part of his job, he said, is assigning a player with his initial numbers. The hard part is updating them week after week. He's a first responder, not a prognosticator.