Gamasutra: The Art & Business of Making Gamesspacer
View All     RSS
October 25, 2014
arrowPress Releases
October 25, 2014
PR Newswire
View All

If you enjoy reading this site, you might also want to check out these UBM Tech sites:

GDC: How Kojima Defied The 'Impossible' Throughout  Metal Gear 's History
GDC: How Kojima Defied The 'Impossible' Throughout Metal Gear's History
March 26, 2009 | By Leigh Alexander

March 26, 2009 | By Leigh Alexander
More: Console/PC, GDC

In a charmingly illustrated GDC presentation, legendary Metal Gear franchise creator and director Hideo Kojima shared his personal methodology for overcoming obstacles like time and tech constraints to achieve a vision.

"I've been in the industry for more than 20 years now, but it's the first time I've ever come to GDC," said Kojima, addressing a packed ballroom of attendees. In past years, he said, he'd been busy preparing for E3 -- "Also, the fact that E3 lost its punch over the years," he said, "and these guys called me and said if I come to this GDC, they'll give me an award."

Last night, Kojima was given the Lifetime Achievement honor at the Game Developer Choice Awards, and during his keynote he offered his own definition of "revolutionary creation."

"I think it's something that no one has done before," he said. "And I also think it's making the impossible possible."

Confronting The Impossible

Kojima illustrated what he called "barriers of impossibility" via an animated visual aid that showed Metal Gear Solid hero Solid Snake confronting steep walls in his path (briefly joined by Mario, who showed Snake up by leaping over walls the aging soldier could not).

The point? Because Mario could jump the wall and Snake could not, passing the barrier appeared to be an impossibility for Snake. "But I think that the impossible is really an assumption, and I believe making the impossible possible demands preconceived notions be discarded."

Cute animations demonstrated other ways Snake could pass the obstacle -- pole vaulting, flying over with a balloon, or breaking it with his gun. And the multiple ways that Snake on-screen was considering traversing his challenge held parallel to later Metal Gear Solid's multiple gameplay paths.

Displaying his talent for nesting metaphors, Kojima labeled the floor on which Snake stood in the animation as the foundation -- game hardware and technology. The wall represented the obstacle in the way of the design goal. Snake can stand on the "box" provided by software tech to get a little closer, and from there, a ladder representing game design takes him the rest of the way.

Starting With Metal Gear: How Tech Limitations Birthed The Stealth Game

Kojima's first game, Metal Gear in 1985, was on the MSX hardware, which was "quite popular" back then, he said. "Actually, I joined this industry... and I was stationed in a division that worked on this platform, the MSX2. And I was given a mission from the very first time in the company to create a combat game for the MSX2."

At the time, says Kojima, Rambo: First Blood was a big hit, and thus became an influence on the spirit of combat games. "So my bosses thought that why don't we bring this [kind] of combat game to a popular platform like the MSX2?"

In 1986, games were entirely 2D, and Kojima felt that to create a combat game, one needed a background, a player, several enemies and the bullets flying between them.

But the MSX2 had a limitation: "You could only place 32 sprites at a time on one screen. However, horizontally, if you display eight sprites, the ninth sprite won't be displayed. That was a restriction back then." He demonstrated Nemesis (Gradius), wherein the ship and its enemies would outright disappear if the sprites all lined up horizontally.

"I'm sure you're feeling better... that you were not creating games back then, because we had so many restrictions," he said. "This is an actual product, it's not a bug... This was the hardware restriction back then!"

With the MSX2, one could display four colors in a line by layering sprites on top of one another -- but through this method, there were even fewer sprites that could be put on screen. And this is the environment wherein Kojima had to create a combat game for the MSX2: one in which multiple bullets couldn't even exist on the screen at once if there were multiple characters on screen as well.

At the time, said Kojima, he thought it was an "impossible mission." Confronted with the need to change his vision, Kojima says he came up with the idea of a "combat game without fighting... of course this game will not sell."

So the next idea was a "combat game about escaping -- just running around. You can't fight, you're just running. And I thought this was totally uncool. I needed to think of a better idea."

Thus the idea of a "hiding" game was born. "This could be revolutionary, I thought," he said. "However, I thought maybe it will still not sell, because this is not really heroic. At the time hero games were popular, so I had to add another idea."

Thus, the idea of a game where infiltration was the primary mechanic was born.

"At the same time, I needed to grow more tension for the players -- and that's what I did," he said, deciding on the idea that the player had to evade enemy detection while sneaking through military bases. "Finally, I thought, an industry first: The stealth game genre was born."

So the original idea for Metal Gear's seminal game mechanics actually came from Kojima's thinking his way around the obstacles presented by the MSX2.

He showed a simple animation illustrating how the early idea for Metal Gear was a sneaking-oriented puzzle game one screen at a time (the MSX2 couldn't scroll).

"I told the player, when you get spotted you're going to get a lot of enemies chasing after you, and I wanted players to learn that this was the rule," he said. "And that was Metal Gear, which was released in Japan and Europe and unfortunately wasn't released in America."

He called the NES version of Metal Gear a "crap game -- because I didn't participate," he laughed.

The enemy's range-of-sight mechanic, plus the AI that causes their algorithm to change when the player gets spotted was the "ladder of game design" that helped Kojima to his revised mission of developing a stealth game on the MSX2. "I think I completed my mission," he said.

Metal Gear 2: Deepening The Stealth Game -- On The Same Hardware

After the game's success, when tasked with developing a sequel, Kojima and team set the next "mission" to create a "deeper" stealth game on the next gaming platform.

But the next platform didn't come. With no hardware advantage and a higher set of goals, the obstacle (still represented by a new wall in front of the animated Snake) is even steeper.

In creating Metal Gear 2: Solid Snake, then, depth needed to be added on the design side, since there was no boost from the tech. One way Kojima evolved the game design was to transform the enemy's sightlines from a single straight beam into a more human-like fan.

The screen-by-screen, puzzle-style layout was also evolved, by adding the radar map to allow players the opportunity to know where enemies off-screen were.

"The player... thinks about the surroundings... and I thought that was quite a good idea," Kojima said.

Also added in Metal Gear 2 was an "evasion" phase, an expansion on the alert mode that creates a cautious behavior for the enemies after Snake evades the alert.

"I added hearing to the enemies... so not only the vision you have to be careful, but the noises you make you have to be careful as well. With the evasion phase, I thought I created more rhythm to the hide and seek."

The addition of radar and sound sensitivity created the experience of vision and hearing, giving the player the feeling of infiltrating an "area" rather than simply a puzzle screen, and the three alert modes created varying tension levels. "Mission completed," Kojima said.

Metal Gear 2 was released only in Japan in 1990. "There was a game called Snake's Revenge on NES... that was a little crap game because I didn't do that game either," Kojima joked.

Metal Gear Solid: Going 3D With PSone

When it was time for a sequel, it was again time to create a new mission -- and it was to "create a 3D stealth game for the MSX2."

"The wall of impossibility this time was really high -- probably it's quite impossible, isn't it? Well, come to think of it, creating a 3D stealth game for the MSX2 was totally impossible -- as I explained, on the MSX2 it was only possible to display 2D and you can't create a 3D stealth game. So there are really some things that are really impossible."

But four years later in 1994, the PlayStation and Saturn arrived. "This was an incident for us," he said, "because at that time, only supercomputers allowed realtime 3D polygons. But PlayStation brought this to your homes."

So using Kojima's illustrated metaphor, in this case, the hardware "foundation" on which Snake stands elevates him closer to surpassing obstacles in the way of creating a 3D stealth game.

In Metal Gear Solid, the addition of 3D created additional hiding spaces when the player is pursued by enemies. When Snake crawls into a duct, for example, the view switches to first person. "This was really 3D," Kojima said.

And there was voice-over for the first time. "We did recordings for various languages," said Kojima, showing examples of conversations between Snake and the Colonel in English, Japanese, German, Italian, Spanish and French.

The addition of 3D, dynamic player-controlled perspective, and realtime cut scenes and voice work ("I guess these cutscenes are not popular now," he joked) helped create a new layer of depth for stealth gaming -- the completion of the mission objective.

And after Metal Gear Solid was a worldwide hit, it was time to raise the bar once more and establish a new mission: to create a still more-realistic stealth game on PSone.

Metal Gear Solid 2: Realism Is More Than Visuals

"A realistic looking stealth game is a real high wall of impossibility," Kojima said. "And we thought this could be quite impossible, but there was an incident that occurred. In the year 2000, PlayStation 2 came along... and this machine was really capable [of creating] a lot of things in detailed graphics. Also, the medium changed to DVD and we could put in a lot more. So why don't we use this technology to rise Snake even higher?"

But the technology didn't do as much of the work as far as deepening the game as was expected. "Realistic-looking" wasn't good enough or achievable enough. "So I changed the mission, not just to make the game graphics-wise better looking, but to create an immersive stealth game, I thought was the key. Of course, this was still mission impossible."

But developing on the PS2 allowed for shadows enemies could cast (and spot), weather that could create atmosphere, and character expressions that could deepen emotion. It was also the first use of motion capture. "I believe that some of you didn't like these long cutscenes," said Kojima -- "You're supposed to laugh," he reproached when the audience was silent.

Further interactivity in the environment -- "not just how it looks" -- was more important to realism than visuals, so Metal Gear Solid 2 allowed players to hide in lockers, apply locational damage to enemies, and more.

Metal Gear Solid 3: Surpassing Its Predecessor Through Software Technology

After Metal Gear Solid 2 broke Guinness records, it was of course time for a new sequel. "So we set a mission -- let's create a stealth game that surpasses the previous one on the next gaming platform." But of course, PlayStation 2 had a surprisingly long life cycle, so without advancements in the hardware, the challenge to surpass Metal Gear Solid 2 was steeper.

"We need to change the software technology, and also the game design," said Kojima. "We need to look back at what we've done to determine how we can advance using nothing but software and game design."

One way was to take the game's setting further along its gradual evolution from closed, artificial environments into natural, open environments -- hence the complex outdoor locations of Metal Gear Solid 3, rich with grass to hide in, logs to crawl through and multiple ways to camouflage.

"We recreated our own 3D engine so we could go into the natural environment," said Kojima. This time, software technology was a bigger support than hardware. Metal Gear Solid 3 added more than environments, though -- the camouflage system, the food scavenging system, the cure system and the CQC combat system ("you're in the jungle, so you can't really fire your rifles all of the time") helped create a significant design evolution.

And Kojima says he tried to make Metal Gear Solid 3's cut scenes a little bit shorter thanks to feedback on MGS2 -- "You're supposed to laugh," he again urged, "I'm talking about my cut scenes!"

Metal Gear Solid 4: 'The Amazing Power Of The Monster Gaming Platform'

"Even though the third game was a hit, I always said I wanted to end Metal Gear Solid as a trilogy. But the world said we needed a sequel," Kojima continued. "So I've come up with a plan -- if I make an ultimate stealth game, then I don't have to work on Metal Gear anymore."

In 2005 Kojima heard rumors of a "monster machine" coming out soon. "I heard that it could do anything -- you don't even have to use the game design," he joked. "You can do anything you want. I thought, maybe then I could create an ultimate stealth game."

Facetiously, he declared: "So the mission was, use the rumored 'amazing power' of the monster gaming platform to create the ultimate stealth game. I thought this will be my final mission."

The audience erupted into applause when Kojima's animations of Snake falling from a high step suggested that the PS3 didn't quite live up to the vaunted, mythical pre-buzz.

"Of course, our expectations and dreams were so high. And of course we can't create an 'ultimate stealth game' that we imagine. So we changed the mission to more reality -- use the actual power of the PlayStation 3 to create the ultimate stealth game." So the goal became "a new infiltration experience."

Metal Gear Solid 4 -- the story of a soldier's final mission using advanced technology -- created the experience of a war zone. "You had an option to demolish both sides, or actually just watch which side is winning. Or, you could ally with the favored side."

"The environment is always changing, so I added an element where the user could select what to do and change his or her situation." The octo-camo disguise system was also a design innovation, since hiding in a battlefield was much harder.

"It's a monster machine, so the cut scenes are monstrous as well," Kojima joked, showing footage from MGS4. "You're supposed to laugh even more," he told the chuckling audience ruefully.

Making The Impossible Possible

To recap, "making the impossible possible" in game production comes down to confronting walls of impossibility with the combined support of hardware design, software capability and designer-driven game design ideas.

Kojima's illustration, presenting Snake with one big wall still left to climb showed, however, that he hasn't yet achieved what's labeled as the "ultimate stealth game."

"But you can see we have climbed quite a few steps up these stairs. And what I want to say which is important... remember the first mission I was given, really impossible to create? If I had given up, saying the wall is too high, I think there wouldn't be any Metal Gear series.

"Or there wouldn't be any Splinter Cell series," he joked, and again: "You're supposed to laugh here." The audience erupted into applause.

"It took this whole 20 years... and we are where we are today. And there is a next trend which is popular right now."

The Next Metal Gear: Integrating Western Principles

Kojma said he believes Western design principles that favor advancements in software technology over either the onward march of hardware or innovations in game design, is this new trend.

"There's a tendency to [use] this kind of design method... in American or European development studios, I kind of realized recently," he said. "So not just creating rules, but they kind of overcome the hurdles or the wall by technology itself.... which is really a high technology, actually."

"And in the future of Kojima productions and how we design games, I want to do it this way. I want to add both Western way of designing and the Japanese way," he said.

"There's a hardware advancement and also a software technology advancement... and I believe there's a next Metal Gear series if we do this. That's my image, actually, of the next Metal Gear."

Closing with some final thoughts, Kojima advised: "Before giving up and saying 'I can't do it,' let's identify what the impossible barrier is. Because I think it's only just what you're thinking... and actually it is possible."

"90 percent of what is considered impossible is in fact possible. The other 10 percent will become possible with the passage of time and technology. So in other words, if you look in the future, think nothing is impossible," he said.

"And I want to make great games with everyone, so let's overcome this barrier of impossibility and create great games in the future."

Related Jobs

Giant Sparrow
Giant Sparrow — Playa Vista, California, United States

Lead Artist
Digital Extremes
Digital Extremes — LONDON, Ontario, Canada

Digital Extremes
Digital Extremes — London, Ontario, Canada

Generalist Programmers
Petroglyph Games
Petroglyph Games — Las Vegas, Nevada, United States

Unity Engineer


profile image
The MG games are all great! My only criticism of the entire series is MGS4 had too many cut-scenes compared to gameplay. A lot of cut-scenes isn't bad in my opinion, but when you make a trade off of more cut-scenes less gameplay then it becomes a problem. If the balance is retained then the number of cut-scenes doesn't really matter. I actually like a lot of cut-scenes, but not at the expense of more gameplay.

MGS4 was great, but there really wasn't that much gameplay when compared to other MGS games. I mean act 5 was just 1 stealth screen, a boss, then button mashing. Act 4 (awesome nostalgia revisiting those places from MGS1) was like 10 stealth screens long with 2 boss fights. Act 3 was an on the rails shooter part with only 3-4 stealth screens/following someone down some streets then a boss. Act 1 and 2 had a good amount of stealth screens and overall gameplay, but once act 3 came the gameplay time was cut down greatly.

Michael Rivera
profile image
It was interesting seeing how Metal Gear became a stealth game, but do wonder if Kojima will ever overcome the "impossible" and start telling his narratives through gameplay rather than tediously long cut scenes.

Bob McIntyre
profile image
B N, you just said that there wasn't much gameplay, but then you listed the huge variety of gameplay types in MGS4.

MGS4 has a ton of different gameplay types. It starts with a little bit of the classic MGS sneaking, followed by the new "battlefield sneaking," followed by a team battle in an indoor level. Then it moves on to bigger two-sided battles in an outdoor setting. It has a low-speed rail shooter segment with some player movement, basically a "zombie" combat game. Chapter three opens with a "track and protect" mission. Then a high speed rail shooter segment. Then a stronger stealth section (stronger because you can't just KO your enemy) in chapter four, followed by a defensive shooter and a mech-fighter section. Then a short, high-tension stealth section, similar to the Ocelot unit battle(s) from MGS3. And interwoven in there are a variety of boss fights.

MGS4 doesn't dwell on any one type of gameplay for very long, but there is a huge variety in there!

profile image

Oh yea, it definitely has different game types. The bosses are also all well done. I was talking more about time taken for gameplay, not quality of gameplay. The quality of the gameplay is really high, but the time of actual gameplay is way less than the cut-scene watching time. Lots of cut-scenes as I said are fine with me, but not at the expense of gameplay time. For instance take a Final Fantasy game, they have lots of cut-scenes, but you may play for 5 hours before you actually see one. In MGS4 it was like 10 minutes of gameplay then a 30 minute cut-scene. That trend continued the entire game so it ended up being more cut-scenes than gameplay, 3 to 1.

Dave Endresak
profile image
I'm not a fan of the Metal Gear series but I liked some of the observations Kojima-san made in this presentation.

I'd also like to add my personal preference for cut scenes in games because I play games for story (just like any other media experience). I vastly prefer cut scenes (as long as they are well presented and develop the story, events, settings, and characters, of course) over so-called "gameplay" that is nothing but repetitous combat (or jumping, or other repetitive endeavors). Without a story and characters I can empathize with, I have no reason to play (except perhaps to waste time, but I really don't have time to waste and there's always other excellent stories to experience and study).

To each their own, of course. There's nothing wrong with anyone prefering to avoid story-driven experiences, but there's also plenty of products available for that preference. Similarly, I empathize with female-focused stories, including games, so that's why Metal Gear never really appealed to me (although I do like some other games with a rather male-focused presentation such as Half-Life). However, there's plenty of female-focused games on the worldwide market, so many that no one could possibly play them all, so it's really not a problem.

Michael Rivera
profile image

I prefer story driven games as well, but I would much rather play a game that told its story through gameplay rather than outside of it (as in most of the MGS series). I think that's why most people have a problem with his over reliance on cut scenes.

[User Banned]
profile image
This user violated Gamasutra’s Comment Guidelines and has been banned.

profile image

Just because a game has cutscenes doesn't mean it can't also push the story forward through gameplay. When a cutscene pops up of my character doing something I still consider myself "in the story" even though I'm not playing at that very moment. Half-Life and HL2 will block your path with a locked door then play a communication message over some screen in the game world, but really all you are doing during that time is watching the screen within the game world, it's the same thing almost. It actually could be the exact same thing if the cutscene was done in first person of you watching a screen. I think people get a little bit carried away with the whole tell the story from within the game world idea. To me both ways are viable and both ways can be done at the same time. I don't really prefer one over the other.

Carlos Mijares
profile image
I apologize in advance for any spelling mistakes. I'm posting from my mobile without a spellchecker. Thanks.

@ Dave

It has already been said, but if you play videogames for the story exclusively, and call a great length of time playing a game over watching its cutscenes as "repetitive," then you don't like videogames. Your time would be better spent reading good books and watching good movies, as these have been telling stories better than videogames, and for a longer time.

@ Michael

I sympathise with your opinion, but there's something I've noticed from playing great games that rely heavily on cutscenes to tell a story (MGS, Okami, Final Fantasy X) and those that avoid them as much as possible when telling a story (BioShock, Portal, Half-Life). I've noticed that I care more for the characters and story of a great videogame that uses cutscenes over one that doesn't. Not only do I care more while playing, but their characters and story leave a greater, more lasting impression on me than a story not told through cutscenes.

I theorize (probably others have to, from students to fellow designers to Roger Ebert) that this is because storytelling is not the strength of videogames, but of books and movies. Therefore, if a developer wants to ensure the Player cares for Character A, or feel sad from a key event in the story, that developer will choose movies (cutscenes) to tell it to the Player. Even the best story event of a cutscene-free game occurs when the Player has no meaningful control, such as the big twist in BioShock, or whenever the Player is temporarily gated in a cutscene-free game to pay attention to the story, either by physically trapping the Player in a room, or by making the Player walk an enormous hallway where no meaningful interactions occur. This is because, if we allow the Player complete control to advance or do whatever he wants, the developer risks the Player missing a huge turning point in the story.

Essentially, if a developer has a story he intentionally wants to tell all of his Players, he'll have to reduce the Player control to meaninglessness or outright remove, as frankly it makes little difference. For the developer that has a story to tell, but doesn't require all his Players to completly absorb it, optional story devices are put in place. The former seems to be the intention of Kojima Productions, and the latter the intention of 2K Boston, for example.

With all that said, I loved the keynote. Hopefully I can get it on video, as audio alone doesn't do it justice. It was quite inspirational.

Michael Rivera
profile image
Carlos: When I said stories should be told through gameplay I didn't necessarily mean the "first-person cut scene" method that Bioshock and HL games use. I was more talking about sequences like the opening chase scene in HL2 or how exploring Bioshock's world tells you more about the events leading up to your arrival in Rapture. I think it's easy for some players (and even developers) to forget how much detail can be added to a story though gameplay like that.

Personally, I don't have any problem with cinematics in games. They work well as a reward mechanic, and they can make a story seem more epic than first-person cut scenes. However, I don't think they should be used as the primary storytelling device, even in a narrative driven game. If a developer is relying on movies to tell their story, why didn't they just make a film/tv series in the first place (especially if they apparently think that's the "best way" to portray the narrative)? It just seems to me like they are ignoring the strengths of their own chosen medium, like if a film director that decided to tell most of their film's story by scrolling text across the screen.

Of course, I also don't buy the assertion that films, novels, etc. are superior storytelling devices to video games. There are ways to make a player feel sympathetic for a certain character using gameplay, and too often I feel like developers use cinematics simply because it is "easier".