When looks are no longer enough

[TheChefO]

And you think it's going to be cheaper to develop noticably improved physics or AI?

I sure do. Ask a developer here what the budget is for art asset creation/engine creation/licensing. Then ask what the budget is for ai. Let me know if I'm wrong.

 

[Fox5]

While I would love better AI in games, I can't see this being a big factor for most gamers. The purpose of enemy "AI" in games is to provide some small resistance to the gamer, regardless of how injured our outgunned the enemy might be, and then die in a cool-looking way. How can you improve on that? Should your enemies call for help? No sensible enemy, when confronted by some ultimate killing machine armed with chainguns and rocket lanuchers, would stand a chance unless he cooperated with allies to surround you. If you tried to attack an American military base, do you think you would just fight soldiers one-on-one? Of course not. You'd face a fight against dozens of soldiers, all highly-trained, and all communicating with each other. A scenario like this would make a terrible game because you wouldn't stand a chance. Even if this sounds like a cool idea to you, think about it this way: would anybody who isn't a hardcore gamer want to play a game like this? Most gamers would get their asses kicked within 3 minutes and then play something else.

How about better AI for NPCs? This was supposed to be one of the big selling points for Oblivion. Geeks like myself quivered with anticipation at the possibilities of Radiant AI. We were told stories of how all NPC actions were dictated by their needs and desires. NPCs would even steal from each other if they were hungry and poor. Then Oblivion was finally released and none of the stories were true. NPCs were just as stupid and lifeless as they've always been. People in Oblivion will stare at walls for 5 hours and then wander aimlessly through the streets. Most of them don't eat at all, let alone steal from each other in order to buy food. I can kick all the food off an orc's table and throw a carafe at his head, and he'll pretend like I'm not even there. Lo and behold, Oblivion still sold millions and got nothing but 9's and 10's from gaming publications. Looking at Oblivion, would you expect game developers to say "our game needs next-gen AI"?

Unfortunately, AI has always just been a buzzword that marketing people throw around. The bottom line is that most gamers just don't care, and likely wouldn't want good AI anyway. For every geek like us, there are 50 dumb teenagers who only want prettier explosions and lots of blood.

It's not like that stuff couldn't have been programmed in though.

For the most part, the problem with AI is that the AI isn't capable of much. It can move and it can shoot and is gunned down within a split second, not much chance to present AI there, especially when it's not expecting you and can't even set up a trap (which would likely require even more interactivity with the environment). We need more convincing games overall, AI is just one part of that, though if you wanted to talk AI that's limited in its range of behavior, Unreal Tournament has some decent AI, as well as fighting games like Virtua Fighter and Soul Calibur 3. Still, just adding in some really basic behaviors would give the impression of much improved AI, when it's really not, current AI just can't do much. Add in a bit of randomness and you have what appears to be an AI far beyond anything seen before when in reality it's just another art asset. There are many old games which appear to have better AI than current games, just because they tried more.

 

[RancidLunchmeat]

I sure do. Ask a developer here what the budget is for art asset creation/engine creation/licensing. Then ask what the budget is for ai. Let me know if I'm wrong.

Uhh.. What?

How does that make any sense? The art asset creation/engine creation/licensing costs aren't going to decrease.

The entire point is that we are supposedly now reaching the point where those costs are resulting in a product that is good enough, so increasing their expenses won't result in noticable increases that will relate to increased sales.

The budget for AI is less? Ok, that's wonderful. But everybody readily admits that AI should be improved. However, it's not that simple. Developers can't simply go out and license AI tech like they can for a graphics engine. So they have to spend the money without knowing what exactly it is they are going to get in return.

Additionally, they must continue to spend their current level on art asset creation/engine creation/licensing.

Unless you are taking the position that graphics are now too good and money should be saved by making games less attractive and that money should go into researching new ways to improve AI and hoping it provides a noticable difference to the average consumer?

 

[pc999]

I do. The gun model is nothing interesting, and the reflection in the visor is only beginning to get to the level that it should.

Now, if those bullet scars on his armor were a result of an actual firefight, and they would increase/decrease according to damage/health limits in the game, then we'd be moving in the right direction.

As it is, it's my understanding that those battle scars on his armor are simply static. Which is hardly impressive.

So yes, I do think we "need" a better Master Chief than that one. As much as we "need" to play games at all, that is.

Ok so assuming that you can indeed notice those things while gaming how many people you think that will notice? Enought for the investiment?

And you think it's going to be cheaper to develop noticably improved physics or AI?

If not, then essentially what you are saying is that video games are dead because there's no significant room for graphical enhancement and physics/AI are going to result in even greater development costs or result in no noticable improvement.

And if so.. I'd like to know how you come to that conclusion. Developers already know how to make pretty games. Making them pretty is essentially a function of time/money/effort. But they don't know how to make noticably improved AI or interactive physics for realisitic environments.

So I doubt that going in either of those directions would prove to be less costly. Especially less costly for a noticable improvement.

Crysis already present a very interactive environment, there is midleware that can do and cost less than.

Is starting to appear AI midleware too like www.spirops.com , games like F.E.A.R. also have good reviewns/reputation because AI and RS is also marketing fear like AI.

These show that the interest is growing.

It is the first time those things start to be really wanted and showing IMO.

 

[RancidLunchmeat]

Ok so assuming that you can indeed notice those things while gaming how many people you think that will notice? Enought for the investiment?

Oh sure.. don't you? Can't you imagine Master Chief walking up to a lake or river (or a mirror for that matter) and seeing two things.. 1) realistic battle scars depending on his health/damage level and 2) the Alice in Wonderland reflection of Master Chief looking in the mirror reflected on his visor?

I certainly think other people would notice those things, and I think if they were accomplished they'd certainly add a 'wow!' factor that would be worth the investment.

Crysis already present a very interactive environment, there is midleware that can do and cost less than.

Sorry, what? There is middleware that can do [what?] and cost less than [what?]?

Are you saying there's middleware that can do more and cost less than Crysis? If so, why are you promoting Crysis instead of these cheaper, better, unnamed sources? I'd also like to be directed towards any tech demos if you got 'em!

Is starting to appear AI midleware too like www.spirops.com , games like F.E.A.R. also have good reviewns/reputation because AI and RS is also marketing fear like AI.

F.E.A.R. had good AI? I think that's being rather disengenious to this topic. F.E.A.R. was primarily marketed on it's graphics and it's 'spooky' environment (Also graphics). The fact that the AI was slightly more intelligent than your average FPS was a bonus. But it certainly didn't replace the investment or the ROI from their graphics.

If anything, I'd say F.E.A.R. is a prime example of a game that does have better than average AI and instead of being widely praised for that as it's primary benefit, it played second fiddle to the graphics engine and art direction and environment.

 

[pc999]

Oh sure.. don't you? Can't you imagine Master Chief walking up to a lake or river (or a mirror for that matter) and seeing two things.. 1) realistic battle scars depending on his health/damage level and 2) the Alice in Wonderland reflection of Master Chief looking in the mirror reflected on his visor?

I certainly think other people would notice those things, and I think if they were accomplished they'd certainly add a 'wow!' factor that would be worth the investment.

Some things are only noticiable in very slow pace and very smal scenes in close ups so if there is a game that way people would notice, but if this is possible in large scale game like Halo wouldnt that be possible in a low scale, plus are you sure that many of the things arent a art issue isntead of tech one (even if faked).

Anyway we are entering in a opinion discussition.

BTW didnt PDZ aleady have real time degradation of armor?

Sorry, what? There is middleware that can do [what?] and cost less than [what?]?

Are you saying there's middleware that can do more and cost less than Crysis? If so, why are you promoting Crysis instead of these cheaper, better, unnamed sources? I'd also like to be directed towards any tech demos if you got 'em!

Sorry for the mess, but take novodex (PPU), future DX physics, Havok FX those are cheap and can do at least whatever a PS3 can and certanly much more than Crysis show and that game already have very nice interaction with the environment.

F.E.A.R. had good AI? I think that's being rather disengenious to this topic. F.E.A.R. was primarily marketed on it's graphics and it's 'spooky' environment (Also graphics). The fact that the AI was slightly more intelligent than your average FPS was a bonus. But it certainly didn't replace the investment or the ROI from their graphics.

If anything, I'd say F.E.A.R. is a prime example of a game that does have better than average AI and instead of being widely praised for that as it's primary benefit, it played second fiddle to the graphics engine and art direction and environment.

That as been my primary reaon to like it and gfx are more a bonus.

I am not a Nintendo ****** saying that a game is only good if the gfxs are bad, I love good gfx but if I cant have all then I think that from now gfx dont need to improve anymore in order to get better AI etc...

 

[fearsomepirate]

I certainly think other people would notice those things, and I think if they were accomplished they'd certainly add a 'wow!' factor that would be worth the investment.

Gosh, I thought $60 games and $400+ consoles were expensive, but now that I know I'll be able to occasionally look at my reflection and see different chinks in my armor than I did 2 hours ago at the last time I looked, it's all worth it!

(P.S. Timesplitters: Future Perfect already did the Hall of Mirrors thing, and it didn't look like hundreds of dollars of awesome.)

 

[MfA]

Ive said it before, but personally I just want more and better scripting.

Whether it be for situational tactics or for branching dialog, a scripted expert system will be outperforming other forms of AI for many years to come. Yes, it's a metric ton of work and doesn't offer very much replay value ... but it still comes over a hell of a lot more realistic.

(Gothic vs Oblivion.)

 

[mckmas8808]

Gosh, I thought $60 games and $400+ consoles were expensive, but now that I know I'll be able to occasionally look at my reflection and see different chinks in my armor than I did 2 hours ago at the last time I looked, it's all worth it!

(P.S. Timesplitters: Future Perfect already did the Hall of Mirrors thing, and it didn't look like hundreds of dollars of awesome.)

Don't be silly. He did mean it like that.

 

[PARANOiA]

I guess time will tell. There are some of us who will pay premium prices for top looking games. Sony took that gamble. Nintendo, on the other hand, bet that people will not care if their machine is better value and doesn't look quite as good. There's a place for both, but who will make the big bucks this gen?

Let's rendezvous in 2009

 

[scooby_dooby]

But how do you know graphics are diminishing? Remember when people here thought that the FFXIII video was totally CGI? Now that we now most of it was real, what's so dimishing about it?

It just seems natural to me. As you keep increasing the level of power, the visual difference between each upgrade becomes less and less. I don't think anyone can argue returns aren't diminishing to some degree, the argument is really about how much. Personally I think they're taking a huge jump, so while they may be diminishing I don't think it's very signifigant.

btw: Halo 3 may very well do what you guys are discussing, the developers have sai they are showing the damage that master chief takes. I don't know if this is prior to his arriving at earth, or is accumulated during gameplay, but it's possible it's the latter.

 

[mckmas8808]

It just seems natural to me. As you keep increasing the level of power, the visual difference between each upgrade becomes less and less. I don't think anyone can argue returns aren't diminishing to some degree, the argument is really about how much. Personally I think they're taking a huge jump, so while they may be diminishing I don't think it's very signifigant.

btw: Halo 3 may very well do what you guys are discussing, the developers have sai they are showing the damage that master chief takes. I don't know if this is prior to his arriving at earth, or is accumulated during gameplay, but it's possible it's the latter.

So what about what I said about FFXIII? I mean people today still think that some of the real-time shots are pre-rendered. Can you imagine games on the 360 in 2009?

 

[assen]

It just seems natural to me. ... I don't think anyone can argue...

Excuse me, but this is no way to argue.

Technological breakthroughs have lead to immense jumps in visuals without a corresponding jump in content creation effort before (eg. bilinear filtering, skeletal animations), there's nothing to prove this won't be the case anymore. A decent pseudo-GI system ("dynamic ambient occlusion"? "real-time radiosity"?) will do miracles for visual quality, from the technology side only, without involving art.

 

[TheChefO]

Uhh.. What?

How does that make any sense? The art asset creation/engine creation/licensing costs aren't going to decrease.

The entire point is that we are supposedly now reaching the point where those costs are resulting in a product that is good enough, so increasing their expenses won't result in noticable increases that will relate to increased sales.

The budget for AI is less? Ok, that's wonderful. But everybody readily admits that AI should be improved. However, it's not that simple. Developers can't simply go out and license AI tech like they can for a graphics engine. So they have to spend the money without knowing what exactly it is they are going to get in return.

Additionally, they must continue to spend their current level on art asset creation/engine creation/licensing.

Unless you are taking the position that graphics are now too good and money should be saved by making games less attractive and that money should go into researching new ways to improve AI and hoping it provides a noticable difference to the average consumer?

You're putting words in my mouth - I don't know why you're being so argumentative when my position has been stated clearly in this thread.

Graphics will improve - I want them to
AI will improve - I want that too
Physics will improve - guess what, I'm not against that either and in fact ... hmm ... yup I want that too.

The POINT is when working with anything the more time you put into something the better it can be. But everything has a point of diminishing returns. It is simply a matter of what people are willing to pay. For the budgets of today in games, there is a limit to what publishers will pay for games. In the current scheme I would rather the budget for the last 1 percent that was going to go into the effort for visor reflection being rendered at 720 native and scaled to fit while keeping a steady 30fps framerate, went into reducing that res to and acceptable 320 for reflection and focused that time on improved AI. While I realize this is not how the budget works and the graphics/programers would typically not be the ones coding the AI, the concept is still the same for increasing the relative effort on AI in comparison to graphics in future budgets in squeezing something in that would not be noticed 99% of the time on something that is noticed 99% of the time.

Edit - example:
Your garbage guy probably comes on a regular basis and handles your trash for you. Coud he do a better job? Probably yes. He could come once a day instead of twice a week. When he arrived he could empty the trash and clean your garbage can(s) for you. Potentially clean up the misc trash that might have fallen onto the ground in front of your house. But how much time is it going to take him vs what you as a citizen are willing to pay for those services?

 

[RancidLunchmeat]

You're putting words in my mouth -

I haven't. I have taken your unclear comments and very clearly asked if I was correctly assessing your position. Therefore, giving you an opportunity to either respond accordingly or correct the record.

The POINT is when working with anything the more time you put into something the better it can be.

Categorically untrue.

But everything has a point of diminishing returns.

Another statement that isn't quite true, and more to the point of this thread is whether or not visuals have reached that point. Some of us are obviously saying no, while you appear to take the opposite position.

Actually, you are trying to take a position that isn't really one of the choices. You want visuals to improve but want other improvements as well.

Great. We all want everything to be better. Re-read the thread title. This is specifically a focus on the utility of continuing to improve visuals, whether such improvement is noticable or worth the investment.

If your answer is that the visuals need to increase as well as improvements in AI, physics, and every other aspect of the game then you, myself, and Mck are in agreement. As opposed to others who believe it is pointless to continue to attempt to increase the visuals of the game because that time/effort/resource should be spent on improving the other aspects at the expense of continuing to improve visual quality.

While I realize this is not how the budget works and the graphics/programers would typically not be the ones coding the AI, the concept is still the same for increasing the relative effort on AI in comparison to graphics in future budgets in squeezing something in that would not be noticed 99% of the time on something that is noticed 99% of the time.

How do you know improvement in AI would be noticed 99% of the time? Just because the AI is 'smarter' doesn't mean it will appear any different or better to the consumer than a 'dumber' AI that is just heavily scripted. On the other hand, every single consumer will be able to notice the reflection in the helmet when they look in a mirror/lake/whatever.

Edit - example:

I don't understand your example at all. My position is simple. That we are no where close to the point where 'looks are no longer enough', because 1) we still lack things like accurate reflections (or because creating those things take so much time and money that the necessary resource drain reflects a poor grasp of how to implement those features. Which means we need developers to continue to work on those things until they are second nature and in every game without added expense) and 2) because physics and interactive environments are necessarily aspects of 'looks' and games are most certainly lacking those things currently.

Could games also use improved AI? Of course. But I don't understand those who take the position that visuals have already reached a point of diminishing returns so that budget should be 'locked' and any further funds should be spent on AI or some other aspect.

I'm under the belief that the average consumer, and the hardcore consumer, will notice and appreciate interactive environments and realistic physics.. both of which are visual enhancements far more than 'smarter' AI. (if gamers even want smarter AI to begin with... which is also an issue up for debate.)

 

[Johnny Awesome]

This article is bunk. We have a LONG way to go visually IMO.

 

[mckmas8808]

This article is bunk. We have a LONG way to go visually IMO.

I'll add to this by saying. "A looooooong way to go". One day in my lifetime I want to see GRAW's intro to be done in real time. We are not there yet, but one day we will be.

 

[TheChefO]

Guys - read the first post again. The snippet of the article does not say "Graphics are good enough now lets move on to something else". It states that while the graphic improvement shown thus far is good, other aspects of the game need to be addressed to bring them up to the same level as the enhanced graphics. I believe this to be a correct statement. I think more attention needs to be payed to other aspects of the game than just a screen shot. Marketing wants screenshots to be improved asap as it makes selling the game an easier task because joe gamer can look at the back of a box in the store and clearly see the difference. While I don't disagree with marketings perspective, I'd say that most of the better games under development at this point are showing graphics that clearly distinguish themselves from the previous generation. It would now be a good time to also address the other issues of a game that are more difficult to sell but are important to the core experience.

 

[scooby_dooby]

Excuse me, but this is no way to argue.

I'm not trying to argue. There is no objective way to measure the amount graphics have 'improved,' so it would be a pointless argument. I will just appeal to common sense on this one.

So what about what I said about FFXIII? I mean people today still think that some of the real-time shots are pre-rendered. Can you imagine games on the 360 in 2009?

Already answered:

Personally I think they're taking a huge jump, so while they may be diminishing I don't think it's very signifigant.

 

[fearsomepirate]

What yardstick do you use to measure "good enough?" I think that people are talking past each other.

Some people don't think we've reached "good enough" until there is no visual feat which cannot be accomplished. Let's call that the "technical yardstick." It's clear that this standard has not yet been met. I can't see every blade of grass in a field, I can't mar any surface I want with my weapon, all bricks in the world magically look identical, etc.

Some people think we reach "good enough" when technical improvement no longer results in increased sales justifying the cost of improving content. We'll call this the "financial yardstick." Arguably, this point has been reached or almost been reached. Some indicators would be the continued market performance of the PS2 in the USA and the current state of the PC video card and games markets.

A handful of people think we reach "good enough" when games are simply aesthetically attractive in their own right and hardware is sufficiently powerful to not serve as a serious limiting factor in the kinds of games available--i.e. new games do not accentuate the inherent ugliness or unplayability of the old ones. We'll call this the "aesthetic yardstick." Personally, I think this point was hit a long time ago in the PC space (around the DX7 era) and is pretty much there in the console area, though the current-gen consoles could all use some sprucing up in one area or another. PS2 could have had a better rasterizer, Gamecube could have had more RAM and disc storage, and Xbox could have had a better memory subsystem, and 480i isn't a terribly clean, beautiful image. But basically, to me at this point (really, ever since I played Unreal on a Rage 128), the vast majority of graphical improvements have seemed like icing on the cake to me.

So which yardstick are you using? I think the guys talking about "diminishing returns" are measuring things aesthetically or financially, while the guys saying we have soooooo far to go are using the first.