"Can you imagine what games we would have if [id's] John Carmack decided he wanted to create a believable character as opposed to a believable gun?"
- Deus Ex and Epic Mickey creator Warren Spector expresses his frustrations with the industry's focus on violence and high-end visuals.
Spector says that he's been "actively trying to shame" developers like Carmack and Epic Games' Tim Sweeney, as he believes they spend too much time working on rendering techniques or finding new ways to simulate violence, when they could focus their efforts on recreating basic human interactions and relationships.
He notes that it's understandable that the industry has a tendency toward violent game mechanics, as those actions are easy to map to the press of a button. While creating something like a virtual conversation is certainly more complex, Spector thinks it's a goal that's at least worth pursuing.
"We focus a little bit too much on violence, but we all know how to do it. It's easy. And a lot of players seem to like it. It isn't all we can do and it certainly isn't all we should do."
There's enough room for both without trying to "shame" his fellow industry colleagues. That's like a hardcore vegan trying to shame someone who eats meat into stopping...it's going about it the wrong way & there's a compromise for both.
While I agree with the fundamental point he's trying to make, citing a specific example like Carmack is kind of dumb. We all know Carmack is a super engineer and rocket scientist, I actually don't think he could do a tremendous job making believable characters. Rendering and animating them I guess, but he probably puts as much time into that as he does the rest of the entire engine.
Somebody needs to work technology angle. AFAIK most developers are just using 3rd party engines and not going all out on technology, and they can do that because a few folks do and sell their engines and show their work.
On visual tech: sure, continued improvement there is a good thing. That increased fidelity in how things are represented helps to create a satisfying play experience.
Equal (or more) improvement in character tech would also be a good thing, and for a similar reason: increased fidelity in how *people* are represented helps to create a more satisfying play experience.
You may not be able to spend the same level of effort on both of those in one game. But the industry generally is not a zero-sum environment -- having great visual tech on one game is not a good justification for saying that there can't be other games that choose to emphasize character tech.
On failure: The original System Shock is generally perceived to have not done well commercially. But the game design concept of building a gameworld and character abilities to allow multiple modes of solving game challenges was so strong that this "failed" game has inspired highly-regarded successors such as Deus Ex and BioShock.
Trying out new ideas is never going to guarantee commercial success. Sometimes even a well-executed concept won't catch on right away. But always playing it safe doesn't insure success, either, and is bad for any industry's long-term health. Fear of failure should not prevent the occasional prudently-managed effort to test new design ideas.
Easy to say when it's not my money at stake, of course. :) But still true, I think.
"having great visual tech on one game is not a good justification for saying that there can't be other games that choose to emphasize character tech."
I didn't say that; in fact I said the exact opposite; that those individuals (as well as the folks at Unity) working on it permits others to do other things.
John Carmack is an engineer. We don't need to shame him like we're stuffing him in a locker. There's a hundreds of thousands of people here who can be doing this and, guess what, none of them have to give a rat's rear about float-int cast shortcuts or stacks or heaps or shader languages.
John Carmack was way ahead of his time. It's like if Zeppelin put of their first album in 1955. Carmack or not, the 1st person shooter would have happened, it just would have taken longer.
I don't see why there isn't room for both characters and violence. I just think the general gamer is desensitized to mass slaughter. I'd love to see a trend where the violence is more realistic or at least has some impact. Do we really see anything like Aerith's death in games any more?
Maybe this due to my arcade roots, but am I the only one that doesn't really care much about the characters, aside from what they are getting up to that makes me need to shoot them? If I want depth, I'll read a book or watch a movie. I don't want to sit there with a controller in my hand watching a skinned mesh deliver dialogue.
I think maybe you're missing the point. He's not advocating making the same violent games but with a nice friendly story. He's advocating making a game where the character depth is integral to the gameplay.
It's comments like this from Warren along with the one he made where he said "games should be more like worlds, than movie sets" or something to that effect, and he's right.
What I think Warren is trying to get at, is, if games are not built as complete-as-possible worlds, or aren't the intelligent systems they could be, interest in gaming will continue to be lost and the industry will continue to shrink as it is now.
The public's interest in videogames is waning as a result.
If you want to make believable people instead of believable guns you're going to have to hire people with arts/humanities degrees to compliment those with science/engineering degrees. Asking a coder to create a believable human is about as effective as asking a philosopher to create effective code.
It won't happen as long as publishers control the games being made. You just can't sell "great AI" in an E3 trailer about shit being blown up to a dubstep soundtrack.
I would disagree. A game that provides realistic characters and is character-driven can sell well, as long as it finds the "50 Shades of..." audience. It's not inherent in publishers to want "shit being blown up", but it is inherent for them to want sales. Publishers want sales and will market to sales, ignoring whatever doesn't sell, including blowing shit up if that somehow were to stop selling.
I disagree that he should be trying to get people who are the best at what they do to try and do something completely unrelated. Let those who are best at pushing their area of expertise push their area of expertise, otherwise that area isn't going to be pushed as quickly, and they most likely won't add much, if anything, to another area.
Not everyone is capable of being a Renaissance Man, in fact most people aren't, and I don't agree with pushing people towards it who didn't arrive at it on their own.
I agree with most of the statements, but lets not forget that "new rendering techniques" can also help to make characters more believable and natural.
But my biggest concern is what Warren stated - that games in general are plagued by violent gameplay. It's like some cliche, when you create basic plot of the game and try to decide what the gameplay will be, and when you can't come up with anything interesting, you go with "cover based shooter". Take a look at "Spec Ops. The line", you could make a game with meaningful choices and interesting, dark plot, but in the end it still plays like a bland shooter, although I think it could have been made in entirely other genre, for example as an "adventure" game without any gunplay at all.
After reading Masters of Doom, I just can't imagine John Carmack worrying very much about Spector's entreaties. I love Carmack's work and respect him immensely. Yes, he's smarter than most of us. But he's a specialist, and he loves (or used to) what he specialized in. Action games drove the need for high quality visuals, and so action games were the reasonable choice for Carmack to make.
I'm sure that there are other talented programmers and software architects who would love to put some serious time into developing more sophisticated models for character personalities, virtual societies, and lots of other complex systems that we deal with in real life every day.
I think what we need are more discussions and more resources to be made available to people with different ideas and approaches. Not millions of dollars, either. Just enough to get small teams working. But whether there are any investors looking to go that way, I don't know. Otherwise, it's up to the indies to try, and start new dynasties like Carmack and Sweeney did twenty years ago.
There's a reason why character development has lagged. It's a cluster of hard problems.
For example: One limitation with chat bots is the hokey sound of speech synthesis. Unless you are working on a sci fi story where synthetic speech fits the world, the chatbots are going to be limited by voice actors and audio bandwidth limitations. Perhaps audio storage space can be improved by streaming it off of the cloud, but that would cause lag/lip sync issues. Given the large amount of voice acting that would need to be recorded, you would have to hire cheap talent. Perhaps a scheme could be worked out to off source the voice work to fans? But that could be iffy. Or perhaps voice work could be outsourced to developing countries (like call centers). Even with the above solutions, the chatbots are going to be limited to full sentences and limited emotional range.
i think we need/ could use both types of developers... i know alot of ppl that dont care about chars or story, but i personally like those things in a game. so why not focus on what u do best? theres an audiance for both types of games imao.
IMO Mass Effect 3 is a game that got characters right showing that violence and good characters are not at odds. There are a lot of talented developers out there, but Sweeny and Carmack get a lot of attention because visuals market well. Plus, when it comes to human characters I identify more with more realistic looking characters.
I think it's rather audacious to ask people to do the kind of work you think they should. Everyone is entitled to their own artistic freedom. You make the games/art you want, I'll make the games I want and who ever wants to play them can play them.
Personally I really like pure game-play. I don't care about character development at all. I like games like Geometry wars, Unreal Tournament, Magic The Gathering (CCG), and Street Fighter. If I want a character driven experience I'll pick up another book.
Triple "A" titles are full of violence (I agree), but I think this is more of a money issue involving publishers. Violence and sex are extremely easy to sell to young men; in my opinion the main demographic for Triple A titles.
Warren, Epic and Carmack etc have their strengths (tech and violence as you said) Your strength is characters. People can only play to their strengths.
Collaborate. Design characters with depth that people will fall in love with; and marry that with world class tech and simulated violence. Everyone wins.
Also, I want to play a game with fun, rewarding and gratifying action and violence - because a game that simulates basic human interactions and relationships WOULD BORE ME TO TEARS.
Have you played any of those? No, because there simply are no games that that would make those interactions play and feel natural. Action and violence seems the "right" way to go only because we have many games in the market that makes the violence feel right, and no games that make other interactions feel acceptable as part of the core gameplay.
"a game that simulates basic human interactions and relationships WOULD BORE ME TO TEARS."
I think this is the wrong way to look at it. We don't need games where you sit around at the dinner table talking about how annoying your boss can be. But what about a video game version of Crime and Punishment? A game where, instead of just reading about the psychological dissonance of a man who has committed a murder, the entire audiovisual package was designed to make you *experience* that mental state? I think *that's* the kind of character development we should be looking to create.
what about a video game version of Crime and Punishment?
Im not sure that AAA RPG is the best platform for highbrow lit adaptations. IF might be a better economic match.
AAA games are expensive. Thats why they are targeted at wide demographics like summer blockbuster films. Another film comparison: prestige films are generally made in hopes to get an academy award marketing boost. They are also made to lure key talent to a studio: "if you do this blockbuster we will do your pet project."
Like most of you i agree with his words to a point. There used to be games that engulfed violence along with story and character development (true crime, GTAIV, etc; and theres one that just came out now that im finding is very violent buut not as violent as say some of the other games out now and thats Sleeping Dogs. There is background to his character and there is purpose and it keeps you playing. I want to see by the end of the month how this game has done sales wise to see if the style of the game (story+violence) would be successful and help WS's point
Gregory Weir's article on this topic, "Why So Few Violent Games?", is one of my favorite pieces of games writing ever.
"In a sense, it’s a self-fulfilling prophecy. We know how to make games about the human heart, but struggle to deal with fists, swords, and guns. And the gamers that grew up exploring Mario’s frustrated relationship with Princess Peach before moving on to the transhuman sci-fi romance of 'Cortana' as adults are reluctant to buy games full of blood and muscles. With those uninterested gamers, big AAA game companies will never invest in, say, a video game where you play a US special-ops soldier fighting terrorism. Combat games are destined to remain the domain of small indie developers."
Maybe instead of "actively trying to shame" others, Warren should try and do it himself. Mass Effect, Heavy Rain, Last of US and other gamers are trying to do this. But what he wants takes a lot of processing power and specialized tools.
...But the thing is that Mass Effect and Last of Us are about shooting badies with big guns and punching them in the faces. All the other interactions are there to break up stuff between combat sequences.
How about a game that has no combat in its core gameplay and consists of something that remotely resembles real human activity, or such problem solving skills that people would apply in real world circumstances.
I think the things Warren Spector has accomplished in gaming, both directly and indirectly, pretty well insulate him from any charge of trolling.
To the question of whether a game without representations of physical weapons would be boring: "no guns" does not imply "no conflict." Plausible actors in an environment where resources and/or rewards are perceived as finite will naturally produce conflict since not everyone's goals can be satisfied. Depending on the setting, that might be economic conflict (a trading game), or emotional conflict (a story game), or even intellectual conflict (a puzzle game). Those aren't to everyone's taste, but then not everyone's into shooters. So why should constant destruction be the default setting?
For that matter, games don't have to be about people competing for scarce resources. There's plenty of design space left for games that are about people cooperating to create new resources.
Multiplayer Minecraft is a limited but reasonable example that there are plenty of people who enjoy cooperating to build things. This could be taken further; I'd like to see someone make something like the Big Challenge game I've described (in a Gamasutra blog post).
The idea here is to create a multiplayer world with raw resources and numerous physical and chemical processes simulated, then give the players some technically challenging goal whose achievement requires their cooperation. The example I like is to plop players into pre-Industrial Earth and tell them to send three characters to the Moon and return them safely to the Earth.
Because there are costs to processes, there would be some conflict to create the most efficient processes (similar to the concept behind SpaceChem). But players would still need to work together to apply those processes to the top-level goal.
This is an off-the-cuff example to make the point that there are plenty of possibilities for popular games -- with varying levels of conflict -- that aren't focused on exciting killing. Mario, The Sims and Tetris have all shipped more units than CoD, so it's not impossible that there's a good market yet to be tapped for games that center on well-developed characters rather than well-developed weapons.
Computer games started a market for high end graphic cards for PC. Perhaps there needs to be something similar for game characters.
Take for example IBM's Watson AI that did the Jeopardy demonstration. It needed big hardware platform just to handle one limited type of question in one controlled conversational setting. For a 1000 NPC game world simulation, one might hope for a hardware platform 1000x to 100,000x more powerful than this. DOes this mean cloud based games? Or does this mean we need a path through videochat AR LARP gamesand more dynamic engaging MMORPG worlds (to take advantage of human intelligence) before we get to NPC AI personality based games.
"We don't need strong AI to make character-based or interaction-based games fun."
I had fun playing pong. I had fun playing lunar lander in basic. Does that mean I'm having less fun, just as much fun, or more fun playing state of the art AAA games? IDK.
My point is that some people really appreciate graphics and created a positive feedback loop between hardware and software enterprise related to powerful PC game graphics. ideally something like that could work for game NPC characterization.
My parallel suggestion was to bypass the AI approach and plug in more humans into the game characterization void.
People excel at things they find interesting. Sure, someone might be able to do a good job at something they're not passionate about, but that's generally not how amazing things are made.
It's a cop-out to say, "I'm not good at X, but he could be, and so he should advance what I'm interested in." If you find something interesting and want it advanced, get good at it and do it yourself. Rally support by inspiring people and by setting an example.
And Spector's always right.
Somebody needs to work technology angle. AFAIK most developers are just using 3rd party engines and not going all out on technology, and they can do that because a few folks do and sell their engines and show their work.
Equal (or more) improvement in character tech would also be a good thing, and for a similar reason: increased fidelity in how *people* are represented helps to create a more satisfying play experience.
You may not be able to spend the same level of effort on both of those in one game. But the industry generally is not a zero-sum environment -- having great visual tech on one game is not a good justification for saying that there can't be other games that choose to emphasize character tech.
On failure: The original System Shock is generally perceived to have not done well commercially. But the game design concept of building a gameworld and character abilities to allow multiple modes of solving game challenges was so strong that this "failed" game has inspired highly-regarded successors such as Deus Ex and BioShock.
Trying out new ideas is never going to guarantee commercial success. Sometimes even a well-executed concept won't catch on right away. But always playing it safe doesn't insure success, either, and is bad for any industry's long-term health. Fear of failure should not prevent the occasional prudently-managed effort to test new design ideas.
Easy to say when it's not my money at stake, of course. :) But still true, I think.
I didn't say that; in fact I said the exact opposite; that those individuals (as well as the folks at Unity) working on it permits others to do other things.
John Carmack is an engineer. We don't need to shame him like we're stuffing him in a locker. There's a hundreds of thousands of people here who can be doing this and, guess what, none of them have to give a rat's rear about float-int cast shortcuts or stacks or heaps or shader languages.
I don't see why there isn't room for both characters and violence. I just think the general gamer is desensitized to mass slaughter. I'd love to see a trend where the violence is more realistic or at least has some impact. Do we really see anything like Aerith's death in games any more?
What I think Warren is trying to get at, is, if games are not built as complete-as-possible worlds, or aren't the intelligent systems they could be, interest in gaming will continue to be lost and the industry will continue to shrink as it is now.
The public's interest in videogames is waning as a result.
End philosophism now!
I disagree that he should be trying to get people who are the best at what they do to try and do something completely unrelated. Let those who are best at pushing their area of expertise push their area of expertise, otherwise that area isn't going to be pushed as quickly, and they most likely won't add much, if anything, to another area.
Not everyone is capable of being a Renaissance Man, in fact most people aren't, and I don't agree with pushing people towards it who didn't arrive at it on their own.
But my biggest concern is what Warren stated - that games in general are plagued by violent gameplay. It's like some cliche, when you create basic plot of the game and try to decide what the gameplay will be, and when you can't come up with anything interesting, you go with "cover based shooter". Take a look at "Spec Ops. The line", you could make a game with meaningful choices and interesting, dark plot, but in the end it still plays like a bland shooter, although I think it could have been made in entirely other genre, for example as an "adventure" game without any gunplay at all.
I'm sure that there are other talented programmers and software architects who would love to put some serious time into developing more sophisticated models for character personalities, virtual societies, and lots of other complex systems that we deal with in real life every day.
I think what we need are more discussions and more resources to be made available to people with different ideas and approaches. Not millions of dollars, either. Just enough to get small teams working. But whether there are any investors looking to go that way, I don't know. Otherwise, it's up to the indies to try, and start new dynasties like Carmack and Sweeney did twenty years ago.
For example: One limitation with chat bots is the hokey sound of speech synthesis. Unless you are working on a sci fi story where synthetic speech fits the world, the chatbots are going to be limited by voice actors and audio bandwidth limitations. Perhaps audio storage space can be improved by streaming it off of the cloud, but that would cause lag/lip sync issues. Given the large amount of voice acting that would need to be recorded, you would have to hire cheap talent. Perhaps a scheme could be worked out to off source the voice work to fans? But that could be iffy. Or perhaps voice work could be outsourced to developing countries (like call centers). Even with the above solutions, the chatbots are going to be limited to full sentences and limited emotional range.
Personally I really like pure game-play. I don't care about character development at all. I like games like Geometry wars, Unreal Tournament, Magic The Gathering (CCG), and Street Fighter. If I want a character driven experience I'll pick up another book.
Triple "A" titles are full of violence (I agree), but I think this is more of a money issue involving publishers. Violence and sex are extremely easy to sell to young men; in my opinion the main demographic for Triple A titles.
Collaborate. Design characters with depth that people will fall in love with; and marry that with world class tech and simulated violence. Everyone wins.
Also, I want to play a game with fun, rewarding and gratifying action and violence - because a game that simulates basic human interactions and relationships WOULD BORE ME TO TEARS.
I think this is the wrong way to look at it. We don't need games where you sit around at the dinner table talking about how annoying your boss can be. But what about a video game version of Crime and Punishment? A game where, instead of just reading about the psychological dissonance of a man who has committed a murder, the entire audiovisual package was designed to make you *experience* that mental state? I think *that's* the kind of character development we should be looking to create.
Im not sure that AAA RPG is the best platform for highbrow lit adaptations. IF might be a better economic match.
AAA games are expensive. Thats why they are targeted at wide demographics like summer blockbuster films. Another film comparison: prestige films are generally made in hopes to get an academy award marketing boost. They are also made to lure key talent to a studio: "if you do this blockbuster we will do your pet project."
"In a sense, it’s a self-fulfilling prophecy. We know how to make games about the human heart, but struggle to deal with fists, swords, and guns. And the gamers that grew up exploring Mario’s frustrated relationship with Princess Peach before moving on to the transhuman sci-fi romance of 'Cortana' as adults are reluctant to buy games full of blood and muscles. With those uninterested gamers, big AAA game companies will never invest in, say, a video game where you play a US special-ops soldier fighting terrorism. Combat games are destined to remain the domain of small indie developers."
http://ludusnovus.net/2011/08/15/why-so-few-violent-games/
He should stop trolling and do more doing.
How about a game that has no combat in its core gameplay and consists of something that remotely resembles real human activity, or such problem solving skills that people would apply in real world circumstances.
To the question of whether a game without representations of physical weapons would be boring: "no guns" does not imply "no conflict." Plausible actors in an environment where resources and/or rewards are perceived as finite will naturally produce conflict since not everyone's goals can be satisfied. Depending on the setting, that might be economic conflict (a trading game), or emotional conflict (a story game), or even intellectual conflict (a puzzle game). Those aren't to everyone's taste, but then not everyone's into shooters. So why should constant destruction be the default setting?
For that matter, games don't have to be about people competing for scarce resources. There's plenty of design space left for games that are about people cooperating to create new resources.
Multiplayer Minecraft is a limited but reasonable example that there are plenty of people who enjoy cooperating to build things. This could be taken further; I'd like to see someone make something like the Big Challenge game I've described (in a Gamasutra blog post).
The idea here is to create a multiplayer world with raw resources and numerous physical and chemical processes simulated, then give the players some technically challenging goal whose achievement requires their cooperation. The example I like is to plop players into pre-Industrial Earth and tell them to send three characters to the Moon and return them safely to the Earth.
Because there are costs to processes, there would be some conflict to create the most efficient processes (similar to the concept behind SpaceChem). But players would still need to work together to apply those processes to the top-level goal.
This is an off-the-cuff example to make the point that there are plenty of possibilities for popular games -- with varying levels of conflict -- that aren't focused on exciting killing. Mario, The Sims and Tetris have all shipped more units than CoD, so it's not impossible that there's a good market yet to be tapped for games that center on well-developed characters rather than well-developed weapons.
Take for example IBM's Watson AI that did the Jeopardy demonstration. It needed big hardware platform just to handle one limited type of question in one controlled conversational setting. For a 1000 NPC game world simulation, one might hope for a hardware platform 1000x to 100,000x more powerful than this. DOes this mean cloud based games? Or does this mean we need a path through videochat AR LARP gamesand more dynamic engaging MMORPG worlds (to take advantage of human intelligence) before we get to NPC AI personality based games.
http://www.ibmsystemsmag.com/ibmi/trends/whatsnew/It%E2%80%99s-Technical,-Dear-W
atson/
Watson is made up of a cluster of 90 IBM Power 750 servers (plus additional I/O, network and cluster controller nodes)
Watson has 10 racks, nine for the servers and one for the associated I/O, network and cluster controller nodes
The Power 750 server was not specially designed for Watson. It’s commercially available and used by thousands of companies today.
The Power 750 uses a 3.5 GHz POWER7 eight-core processor, with four threads per core
Watson has a total of 2,880 POWER7 cores
We don't need strong AI to make character-based or interaction-based games fun.
I had fun playing pong. I had fun playing lunar lander in basic. Does that mean I'm having less fun, just as much fun, or more fun playing state of the art AAA games? IDK.
My point is that some people really appreciate graphics and created a positive feedback loop between hardware and software enterprise related to powerful PC game graphics. ideally something like that could work for game NPC characterization.
My parallel suggestion was to bypass the AI approach and plug in more humans into the game characterization void.
It's a cop-out to say, "I'm not good at X, but he could be, and so he should advance what I'm interested in." If you find something interesting and want it advanced, get good at it and do it yourself. Rally support by inspiring people and by setting an example.