E3 - next generation games?

last week was E3 (the electronic entertainment exposition) in LA, and though I did not go, quite a few people did. therev was footage from some of the next gen consoles (good mix, skip to the last 20 minutes of this PS3 video) as well as Nintendo's Game Boy Micro on demo. however, after all the coverage I read, I feel somewhat as does this ars writer that E3 is "a great big bag of cottony fluff."

of everything I saw, the only thing that really excited me was some footage from the PS3 video linked above that appears to be in game footage from a fps where the enemy shoots a gargoyle statue that goes to pieces as it's shot at, and then the player shoots out a glass bridge, causing his opponents on it to fall to the floor. that sort of interactive environment is something that, to my knowledge, has not been previously achieved to that degree. however, everything else is just pretty, and that's about it.

don't get me wrong. the level of detail and realism that we've achieved with real time graphics is truly amazing. and it's entirely possible that if graphics aren't the most important thing, I should be looking more at the GDC. however, I can't help but wonder if there's a better use of our creative hardware designing abilities. during a recent conversation with a friend back at UCF, he mentioned that one of the reasons comics such as Dilbert are so successful is that they only have a limited degree of realism. they are abstract enough to allow the anyone to identify with the character, but realistic enough to allow the reader to maintain psychic distancing. also, Masahiro Mori's uncanny valley comes to mind. furthermore, I attended a talk by Ed Niecikowski in which he intimated that in movies, and often now in video games, we block out the real world in order to create an artificial and somewhat realistic version of the real world (although he was speaking more in reference to sound design). is there a point at which we can have too much realism? I think the obvious answer is "yes," but the difficult part is determining what that point that is.

however, a big part of the reason that games focus so much on graphics now is that that's the hardware's strong point: pushing lots of matrix manipulations and the specialized math to compute ray tracing, particles, etc. however, there are other processor-intense calculations that need to be performed by lots of games, such as physics engines (for which a physics processor unit has been proposed, but may not catch on). why don't hardware companies develop more devices like this? could the problem be technical, or that not enough games would use the hardware to generate a large enough market?

or, does it have to do with capitalism? game companies are market-driven, and the market is consumer-driven, and apparently consumers want pretty graphics. it doesn't matter that the fps they are playing is essentially a reincarnation of Doom (or, to go back farther, Wolfenstein) with a prettier interface. gamers seem to get all caught up in the look of a game and seem to care less about its play value, and even less about its originality. yes, technically, you'll be fighting different enemies in different areas and the "plot" will be different, but the game play is essentially the same; very rarely does anything innovative come along, and when it does (e.g., GTA), it only strays slightly from the beaten commercial path.

however, we should step back for a moment and consider the fact that the market is in fact consumer-driven, which means that the majority of consumers want to play reincarnations of the same games over and over. the majority, but not the entirety. I, though not an avid gamer, would love to see something more plot- and character-based. don't write a script for what happens in the game, but create characters with personalities that react to the player's action, ally themselves with the player, conspire against him, interact with each other and not just with the player, and have plans and goals of their own. this is just one example of the sort of game I'm talking about. admittedly, such a system would be quite computationally expensive. so is pushing thousands of millions of floating point and matrix operations through a GPU, and we have developed specialized hardware to do it. is the difficulty that it's a problem of such complexity that we can't write anything like an efficient set of algorithms to solve it, or is the difficulty in marketability? people generally play games to win (I'm pretty sure that's a part of the definition of a game), and I'm not sure that anyone would accept a game that couldn't be definitively "won" as such.

in much the same way that art films not palatable to the general movie-going public spawned, I think an indie game culture may evolve for such non-mainstream games (it appears that such a culture is already beginning).

there's a great deal of human nature to be explored in the way of why we play games, why we choose to play the games we do, why some people prefer some games over others, what people want out of games, and what a game is exactly. however, without the impetus of novel games from the industry, I think this exploration may be a long time coming.