A game in its entirety is defined by rules. You can immerse the player as much as you want, you can change the presentation to be as beautiful as heaven, but those things do not change the rules of the game. You can change animation sequences, you can beef up the AI, you can age fictional characters, but these do not change the rules of the game.

All of the things mentioned help the story the game happens to contain. A good story though is worthless with terrible gameplay. Good gameplay is fun, good story, bad story, or no story.

Quote:

Original post by Shpoonj I've read a lot of posts about what's wrong with the games we play and how the next generation will bring a few changes but not as many as we're hopinh for. To be honest, I don't think all the possibilities have been discussed. Does anyone remember when Lionhead and BBB were working on growing trees and aging villagers for Fable? Wouldn't it be amazing if a game was filled with towns and people who looked different each time you played? Where is a combat system that doesn't use the same three animations for attacks over and over? Why hasn't someone done the math and calculated which angles compliment each other and made a system where you can choose what angle you attack from and had impromptu animations each time? Where's the AI that knows how to learn from you? You know, the one we've been promised countless times. The one that watches your best moves, and implements them flawlessly. Maybe the technology isn't at this point yet, but I have a feeling it will be soon. If you won't make these games, I will, but it's giving me a headache. Let me know how you feel.

I think the reasons these things haven't yet been implemented in games are ones other than what might be immediately obvious.

The algorithms and techniques are already there. We know how to make amazing AIs for your FPS combatants. We have known how for quite a while. However, how long might it take to implement such complex logistics? How much money might it take to pay professional programmers (probably requiring PhDs, not just run-of-the-mill programmers) to work all that extra time? The answer to both is frankly probably too much.

But even more importantly, if you were the leading publisher, and were able to get millions of sales by secretly skimping on the AI but lying early on about it to build hype, would you? I can't answer that question for you obviously, but unfortunately 'Yes.' is the answer given by today's major publishers.

Its the little things that aren't noticed. Its also the somewhat lesser known games that bring in strange innovations and ideas.

For example, for full control of your "character" in combat, I think Robot Alchemic Drive (RAD) probably went the furthest towards total control. Every limb and joint of the robot was literally mapped onto a ps2 controller. The R1/R2 and L!/L2 buttons made the robot either take a step forward or backwards with the left and/or right foot. So, walking involves alternating between pressing R1 and L1. The arms were mapped onto the two analog sticks and you swung your punches with them, which means you had somewhat of a control as to where to punch, like those arcade boxing games. Then the torso rotation was mapped to the digital pad and the 4 buttons did other stuff, like weapon selection, special attacks, etc. Just reading the manual for the controls was a daunting task as natural as the mapping seems. I can imagine actual execution would be a completely different story. So, if you want full, control, give that game a shot, though I doubt it will be extremely responsive, but will sort of bring forth the fact that full control of your character equals to somewhat of a learning curve and physical dexterity.

Good AI has always been around the corner, but it has always stayed that way. You come up with something smarter and people just want more. 30 years ago, people were in awe of the seeming intelligence of ELIZA, which in current day terms was a simple chatbot. By about 20 years ago, ELIZA just wasn't smart anymore. So, its not that AI hasn't gotten better, its just people are demanding more. I thought Virtual Fighter 4 had a pretty damn interesting AI system just because you could really "train" your own AI character from scratch. But like most good things in AI, training takes time. Most of the times, by the time the AI has gotten smart enough, you've finished the game. So, there are ups and downs to learning AI's, since time is usually involved.

Watching games evolve is kind of like watching a tree grow. If you stare at it all day, it always looks as if nothing is happening, but go away and come back in 5 years and you wouldn't even know what you're looking at.

Okay, here's my brief take on some of these ideas...

Quote:

Original post by Shpoonj Does anyone remember when Lionhead and BBB were working on growing trees and aging villagers for Fable? Wouldn't it be amazing if a game was filled with towns and people who looked different each time you played?

Well, I agree it would be cool, but I'm not sure about earth-shattering. I mean, having people that look different is actually achievable with today's technology (such as that used in Sims 2). But the quesion is; how much more will that add to the game?

However, I am fascinated with all forms of algorithmic art, and I think this will lead to improvements in gameplay, which might be what you are talking about here. Algorithmic art is more customisable, and more customisabiliy means more flexibility to providing what the player wants, and in incorporating interactivity. Plus it would be good if a mediocre artist such as myself could get the computer to do some of my artwork for me. [smile]

Oh, and don't forget there's also a game design reason to have most of the characters look the same. If the player enters a new village and everyone is wearing the same brown peasant clothes, except for that one woman over there wearing an orange suit of chain mail, the player knows which character is the important one.

Quote:

Where is a combat system that doesn't use the same three animations for attacks over and over? Why hasn't someone done the math and calculated which angles compliment each other and made a system where you can choose what angle you attack from and had impromptu animations each time?

I'm sure that people are working on this right now, and will be in games within the next couple of years, if not already. But I'm also not sure what this will add, unless it's coupled with a realistic physics engine.

Quote:

Where's the AI that knows how to learn from you? You know, the one we've been promised countless times. The one that watches your best moves, and implements them flawlessly.

Not sure if this would be fun though. An AI opponent that is unbeatable? With AI, either it is a really hard problem to make a tough opponent, or it's riddiculously easy and just results in an unentertaining experience. Imaging a fighting game like Street Fighter where the AI opponent always blocks your every move; it's easy to implement, but just no fun.

Learning AI (such that it is) has been around at least since Killer Instinct. I remember it being abundantly clear that the enemy fighter was getting better at blocking particular sequences of moves. If you throw him a high-high-low combo four times in a row, he'd figure it out and block it. If you immediately performed a high-high-high combo, he'd block the first two and then block low while you socked him in the face.

If that isn't a learning AI, you need to explain to me just what you mean.

We've all read the articles about how raphics are getting too much attention and game design is going down the crapper. I think the next generation of games will have a handful of great games, a few dozen good games, a bunch of popular yet highly derivative games (Madden 2007! Now with nasal hair!), and a horde of crap titles.

I've found games at addictinggames.com that are way better than many of the big-budget games that get released.

I think David Wong was right about the impending video game crash, but I think it'll be commercialism that kills the industry, not intellectual stagnation.

These ideas are scrapped because they're not economically feasible, not because nobody is cool enough to go through with it. If great ideas were magically turned into finished projects, I'd be living on Mars with my twenty foxy clone girls right this minute.

The fact of the matter is that making a huge, detailed and innovative video game now requires an amount of effort and testing and funding and expertise comparable to a flight of the freaking space shuttle. I can't do that from my basement. Neither can you.

I think that the field of video game design benefits from innovation that takes place in the form of academic discussion on these boards just as much as when a crappy game with a really neat aging system gets released for the XBox. The simple fact that innovation != fabulous cash prizes doesn't mean it isn't happening.

I encourage you to make these games. Do it. I know funding is a problem, but if everyone on these forums contributes two cents, I'm sure you'll be okay.

Quote:

Where is a combat system that doesn't use the same three animations for attacks over and over? Why hasn't someone done the math and calculated which angles compliment each other and made a system where you can choose what angle you attack from and had impromptu animations each time?

That sounds sort of like reverse kinetics. I remember hearing about a game being developed along those lines a few years ago. The idea is that you can put the character's hands or feet wherever you want, and the game figures out a reasonable position for the rest of the body. They were going to try and use it to make a gymnastic fighting game, sort of like Prince of Persia.

There were a lot of problems, though. It was unpredictable, it was taking forever, and they couldn't figure out a reasonable control scheme. Eventually they ended up using the reverse kinetics as a development tool to generate static animations.

The problem isn't a lack of technology, it's that a totally unpredictable fighting game with thousands and thousands of moves would be a nightmare to create and no fun to play. Less is more.

Quote:

Original post by trickster721 The problem isn't a lack of technology, it's that a totally unpredictable fighting game with thousands and thousands of moves would be a nightmare to create and no fun to play. Less is more.

I've seen a bit on inverse kinematics done in robot control theory. I don't think you would want to imlpement this as thousands of moves, but more as it's done in robotics; a robot welder knows where it wants to position the welding arc, it just has to figure out how to move the joints in the arm to get there.

As long as the "real future of games" still has room for the likes of Bangai-O and Viewtiful Joe then I'll be content. If it doesn't then I'll be spending far more of my entertainment funds elsewhere.

The only thing I see here as truly innovating is the On the spur animation.

This is one of those things that could be alot of front-end work, but in the long run ends up being a boon. Especially if you want to run a game that has a lot of free-form stuff, such as combat...

Though you do see something similar to that. That's the "Rag-doll physics" in games like Unreal Tournament. I think this one could work well..

This however is one of those things where it's usefulness depends on the needs of the game...is it easier to write a code to do animations, or to make 1000 animations for each character?

In my opinion, the only future for gaming lies in virtually eliminating performance penalties. Where a traditional game would have you fail a task - die, for instance - a "new" game would have you narrowly escape, then plot an alternative, perhaps lengthier trajectory for you to accomplish the goal, or even a lesser goal. Every complete play session would end in success/victory, but at varying costs and of varying reward.

What can I say? I hate reloading.