Support NeoGAF

[amigastar]

I'm waiting for GTA VI in that regard, Should be awesome.

 

[Ozzy Onya A2Z]

Seconded OP, and thirdeddededd.

 

[Kataploom]

I wish they stop pursuing visual fidelity and start focusing more on art direction. Specially newer generations don't give a damn about visual fidelity.

 

[Danjin44]

I wish they stop pursuing visual fidelity and start focusing more on art direction. Specially newer generations don't give a damn about visual fidelity.

It’s reason why I’m fan of how FROM and Vanillaware makes games how they consistently deliver good games.

Once you make a game with high tech graphics then next game needs look better and the next game and next game, meanwhile the cost will go up the focus shifts to graphics rather than the actual game itself.

This is why most Japanese devs stopped chasing this pointless graphic race and in my opinion they are better for it.

 

[Kataploom]

It’s reason why I’m fan of how FROM and Vanillaware makes games how they consistently deliver good games.

Once you make a game with high tech graphics then next game needs look better and the next game and next game, meanwhile the cost will go up the focus shifts the graphics rather than actual game itself.

This is why most Japanese devs stopped chasing this pointless graphic race and in my opinion they are better for it.

Also don't you think this constant "bigger, better, etc." puts people in stress by expecting that all the time and leaving them in constant disapointment, ultimately making them not enojoying gaming if there's no WOW! NOVELTY! factor involved? I think it's sad but it's something I notice a lot, specially here on GAF.

 

[GymWolf]

Hopefully gta6 is gonna have something to say about simulation fidelity.

 

[HL3.exe]

I’m aware of that, my point was the Hulk game had it as a core concept so it made sense to focus on it for that game. I just personally don’t care about stuff like that unless it seems really necessary for the game. Like, some might roll their eyes if they walk into an Elden Ring NPC and the NPC doesn’t move at all, but it means nothing to me personally. I’d rather the team focus on the combat, level design, art design, etc., to me that’s what matters. I’m immersed because of the amazing world they created, not because of how much of the world I can “interact” with. Like, if all the crates, barrels, etc. in the Soulsborne/ER games were just indestructible as opposed to breaking if you hit/roll onto them, it wouldn’t matter to me at all.

And honestly, rag doll physics on corpses/KO’d enemies sometimes break my immersion. I’d rather the game have them vanish/“disintegrate“ or something rather than having them move if if I happen to walk over them as I find it a visual distraction.

If you’re referring to stuff even outside of what I mentioned, then could you lay out all the possibilities?

To be clear, I’m not saying you can’t want this, just understand this might not be a desire for some gamers. We all have different desires/priorities when it comes to gaming.

Different strokes I guess. I like games to be unpredictable and generating emergent outcome through their physics and ai. A dynamic environment were you get that 'anything can happen here' feeling because of the flexibility of their data-driven systems, is more immersive to me then easily readable game-logic that is more predictable.

Also, heavily disagree on the canned animation > ragdoll part.

These videos alone generate millions of views, because the impact and reactivity of the synthesized animation systems makes so NPCs can grab on to things and blend more correctly into the environment, creating that 'infinite possibility space of emergent outcomes' feeling. These systems are not at all perfect, but that's the beauty of it. It can only get better from here if more data-driven systems design gets more approachable.

 

[FewRope]

Maybe devs just don't know how to program/code for shit like this anymore.

Agile methodology truly fucked how coding works, Idk about game development but in web dev it destroyed any sense of making a true polish product

 

[Filben]

Also environnemental destruction. It's really baffling how little happens to the environment in shooters like Gears, The Last of Us, RE Remakes. It doesn't need to be over the top Control style, but even what FEAR manages is rarely seen these day. Totally static.

This doesn't apply to shooters only. When playing the FFXVI demo I was unpleasantly surprised that slashing at even the most unstable wooden barrel doesn't smash it or leave a mark or dent in it. Except for enemies, nothing reacts to your attacks.

 

[GigaBowser]

moooooost young devs these days are graphicimical artists and nort real game designers I'm sorry my friends

 

[Josemayuste]

Totally agree with OP, recently, I've come to realize that I enjoy the games with more of an emphasis on physics or improved interaction or mechanics over graphics, TOTK, PREY or Outer Wilds comes to mind

 

[Hypereides]

I echo your sentiment. I want to see more advancement in the tech surrounding the game. Visual enhancements have become dull and almost minute (art direction is where its at anyway). If someone can manage to implement a LLM AI (AGI) into a game to generate dynamic NPC actions and dialogue (speech) on the fly then I'll be somewhat intrigued.

AAA used to be "cutting edge" and pushing the forefront, yet they've been either resting on their laurels or sleeping at the wheel on this regard.

Back on the early 2000s videogames were still a hobby for nerds. Someone would come out on stage, show you a demo of Half-Life 2 physics and everyone would piss and shit their pants in place.

Nowadays gaming has become way more mainstream, and my guess is that the average videogame consumer doesn't care or even know about AI, physics, that cool fire propagation in FarCry 2, etc. Most care about stories and graphics, which is why we still play the same games as 10 years ago, just looking a bit better.

Because sure, game companies could invest on better AI, better physics and whatnot. But that doesn't sell. Cool screenshots and bombastic trailers do, so I don't see any AAA investing in this kind of stuff unless it's the main hook of the game, like with Zelda TOTK, which really isn't that impressive when you consider we were seeing this stuff in Garrys Mod many years ago.

I'm inclined to agree and you have to look no further than this thread to spot those people.

 

[Madserb2023]

Would love to see it. You can have both though. In 2006, 2007 you did with games like Crysis. I can see why you might not want fully destructible environments as level design would be difficult but we don't even have what you had with Half life 2 these days.

 

[Mr.Phoenix]

i think devs dont do that because most gamers have a short attention span and have no patience

Not true. This is the new lazy cop out we use as to why things are the way they are now. But this is just not true. We say this every single year. And then boom, a game like BOTW, Elden Ring, BG3, Hogwarts.... etc comes out and breaks records.

Gamers seem to have shorter attention spans in games because the devs are not doing enough to keep the gamers engaged. Too much hand-holding which only ends up dumbing games down and making them nothing more than task simulators. What literally makes the bulk of most gameplay in games these days is going from map marker to map marker and completing fetch quests with the sprinkle of killing these spawned enemies here.

Just look at that, who is going to want to be doing that for 40+ hours game in game out? The ubisoftification of games.

Indeed there is. Game development has become as easy and accessible as it's ever been, yet developers seem to lack the talent or passion as those from years ago.

Or maybe they still have it, but the suits won't let them use that talent. Who knows.

There is a reason why the games that have the most elaborate physics systems... look the way they do. As it stands, what is simply happening is that the bulk of development budgets are going to graphic and art assets. Because those things are easier to sell and market. But there are budgets. And those budgets don't just mean money they also mean time. If devs are spending 80% of their monetary and time budget on graphics... the simulation aspects of the game will suffer.

Our only hope is that we can get to "can't really do much better" graphics ASAP so we now get back to the point where the only real way for developers to differentiate themselves is by focusing on innovative game design instead.

Think about it, BOTW series on the Switch, is pretty much up there with the best that platform can look like visually. When your visual ceiling is so low on any platform, then the devs have the time to focus on all the other stuff. I can promise you though, if BOTW was looking like HFW? And they still wanted to do all the simulation stuff they did? That game would probably still not have been released yet.

Another good easy example is the recently released Tekken 8. That is the best-looking tekken game ever made (obviously), but compared to tekken 4/5/6/7... hell even tekken tag on the PS2. It has the worst environmental physics of the series. Nary any environmental response to player movement, when in the older games you had better water physics, and environmental deformation like sand and snow...etc.

 

[BbMajor7th]

For me, it's innovation that's really missing. Big publishers should be greenlighting off-the-wall stuff every year - small teams on modest budgets pushing ballsy new ideas that drive the industry forward. Massive tentpoles are so expensive that innovation becomes prohibitively risky; new business strategies that focus on developing new ideas - uncovering the next Among Us, Minecraft or Undertale - should exist alongside the giant AAA games that take years to develop. Plugs gaps in release windows and plant seeds for future gaming trends.

It's the shortsightedness of leadership and the obsession with a YOY growth model that is killing the industry.

 

[Plague Doctor]

OP I am with you on the simulation fidelity front. Also when it comes to just the visuals, art direction and vibe > visual fidelity any day of the week. Seeing minute facial expressions and swaying blades of grass means fuck all if the world you created is basic and dull. Infact, I'll take lower polygon count if the world and characters are interesting and unique.

 

[Inuteu]

all I want is swords that really cut

 

[KXVXII9X]

Back on the early 2000s videogames were still a hobby for nerds. Someone would come out on stage, show you a demo of Half-Life 2 physics and everyone would piss and shit their pants in place.

Nowadays gaming has become way more mainstream, and my guess is that the average videogame consumer doesn't care or even know about AI, physics, that cool fire propagation in FarCry 2, etc. Most care about stories and graphics, which is why we still play the same games as 10 years ago, just looking a bit better.

Because sure, game companies could invest on better AI, better physics and whatnot. But that doesn't sell. Cool screenshots and bombastic trailers do, so I don't see any AAA investing in this kind of stuff unless it's the main hook of the game, like with Zelda TOTK, which really isn't that impressive when you consider we were seeing this stuff in Garrys Mod many years ago.

I was always more of a casual gamer, and I remember being kind indifferent when people would gush about the NVidia PhysX and all of this tech. I got my first PlayStation (Non Nintendo console ever) and got introduced to a lot of games pushing these boundaries. It wasn't until after the PS4 generation where I really took notice of some of the physics and other forms of interactivity introduced in the last generation were being sidelined. The jump from playing GTAV to WatchDogs was very noticeable with how less interactive WatchDogs was.

A casual may not know or care about these things, but I think they will notice and appreciate them over time.

 

[KXVXII9X]

I definitely find physics have more potential to actually provide interesting gameplay mechanics. I also find it a way more interesting aspect than improved visuals.

But that said, RDR2 is one of the most boring “games” I’ve ever played. And I’m sure if more devs started to go the “physics-route”, it’d be more in line with that than anything else since everyone seem to want to go with “realism” all the time. In that case, I’d take the visual route any day of the week. At least the majority of games that lack heavy simulation are still fun to play.

I think Mario and Sonic are a good, overlooked example of using momentum physics to enhance gameplay while not trying to be realistic. It is a big part of what makes those games feel great to play. I definitely agree that the games should be fun and not go overboard with realism all the time. Zelda BotW and TotK also have great physics that don't bog down the gameplay. Also, my profile picture is another example of physics providing interesting gameplay mechanics lol.

 

[Men_in_Boxes]

Games with bleeding edge graphics are universally terrible. Designed to sell to the casual gamer as a cheap (expensive) marketing trick.

 

[zeroluck]

If you look at the below is there any real difference to adding ray/path tracing?

Again I ask, how does this enhance a game if the NPC Ai is not smart enough to navigate the game world as an example. We had such examples when Cyberpunk launched and there's tons of videos to show the brain dead AI. No point looking photorealistic if the AI is something from the early 2000s.
That breaks the immersion of the game world.

This is like saying LCD looks just as good as OLED at displaying a lot of content. Therefore OLED is useless.

 

[GymWolf]

Different strokes I guess. I like games to be unpredictable and generating emergent outcome through their physics and ai. A dynamic environment were you get that 'anything can happen here' feeling because of the flexibility of their data-driven systems, is more immersive to me then easily readable game-logic that is more predictable.

Also, heavily disagree on the canned animation > ragdoll part.

These videos alone generate millions of views, because the impact and reactivity of the synthesized animation systems makes so NPCs can grab on to things and blend more correctly into the environment, creating that 'infinite possibility space of emergent outcomes' feeling. These systems are not at all perfect, but that's the beauty of it. It can only get better from here if more data-driven systems design gets more approachable.

Canned animations, and especially death precanned animations without ragdoll are dogshit in 2024, i fully agree.

Thank god we have like what? 1 or 2 big devs that still use canned death animations like sucker punch and maybe kojima?

We kill thousands of enemies on every game, visual variety for every kill is paramount, this is not the 8bit era anymore, even a cartoony, childish game like the last 2 zelda or palworld adopted ragdoll because it's fun to send enemies flying "realistically"

The problem is that not everyone is as good as rockstar or nd when it comes to weighty ragdoll that looks great and doesn't glitch (ok it glitch a bit in rdr2 but euphoria is more complex than just ragdoll so they get a pass).

 

[diffusionx]

I echo your sentiment. I want to see more advancement in the tech surrounding the game. Visual enhancements have become dull and almost minute (art direction is where its at anyway). If someone can manage to implement a LLM AI (AGI) into a game to generate dynamic NPC actions and dialogue (speech) on the fly then I'll be somewhat intrigued.

I think devs are going to be very hesitant to do this, not for tech but market reasons, the first time someone manipulates the AI into going full Tay then they will be screwed. They don't want anything they haven't coded in themselves into their games.

 

[HL3.exe]

The problem is that not everyone is as good as rockstar or nd when it comes to weighty ragdoll that looks great and doesn't glitch (ok it glitch a bit in rdr2 but euphoria is more complex than just ragdoll so they get a pass).

True, but it's not about being good. It's about physics middleware being on the out. These things were relatively still affordable at the time as budgets were still manageable. This isn't the case anymore, that's why most op'd for a out of the box solution like Phyx which isnt great. Havok and such are excellent but really expensive. R* at the time when they were building their new tech base (around 2005/2006, built off of Angle's engine) uses BULLET Open-sources physics library combined with Natural motion Euphoria. Euphoria was originally being indented for cutscene, but Sam Houser wanted it at runtime (realtime gameplay). They hired the Euphoria team (which would be nuts these days) to incorporate the tech into it's engine core.

The Making Of: Grand Theft Auto IV | Features | Edge Online

Sitting down with Rockstar Games president Sam Houser in the freshly decorated demo room of the company’s headquarters in downtown New York, we ask him

web.archive.org

Sam Houser: “When we initially saw Euphoria, I was floored. I was like, ‘That’s my dream – it’s happening, it’s there, let’s do it’. But all the realists who actually have to make the stuff were like, ‘Sam, man, it’s never going to work, it’s never going to happen’. I think initially it was very much pitched as something to use for cutscenes, to have a really cool-looking action of someone falling down the stairs or whatever. But there was a bunch of guys in our crew who really looked at it and they were like, ‘I think we can actually get this running in the game, in realtime’. So it’s been incredibly collaborative, which I love, and I think the fruits of it are amazing. When you’re taking a shot at somebody and they go staggering procedurally, and they lift up their gun to try and get a shot back at you – it’s giving people unique moments like they never had before.”

Also some interesting thread from a immersive sim dev about modern day physics implementation:

 

[UltimaKilo]

Canned animations, and especially death precanned animations without ragdoll are dogshit in 2024, i fully agree.

Thank god we have like what? 1 or 2 big devs that still use canned death animations like sucker punch and maybe kojima?

We kill thousands of enemies on every game, visual variety for every kill is paramount, this is not the 8bit era anymore, even a cartoony, childish game like the last 2 zelda or palworld adopted ragdoll because it's fun to send enemies flying "realistically"

The problem is that not everyone is as good as rockstar or nd when it comes to weighty ragdoll that looks great and doesn't glitch (ok it glitch a bit in rdr2 but euphoria is more complex than just ragdoll so they get a pass).

I really dislike the ragdoll physics. There’s no weight to anything and makes the engine feel
cheap to me. VR is a different story, though. Your point on “visual variety” is very well taken.

I should be able to chop an enemies limbs off (not scripted) if I wanted to. Heck, when I played Half Life Alyx, it bothered be that I couldn’t grab the crabs with my own hands and tear them apart.

 

[nemiroff]

Yeah I'm pretty sure I'll take both, thank you.

 

[GymWolf]

I really dislike the ragdoll physics. There’s no weight to anything and makes the engine feel
cheap to me. VR is a different story, though. Your point on “visual variety” is very well taken.

I should be able to chop an enemies limbs off (not scripted) if I wanted to. Heck, when I played Half Life Alyx, it bothered be that I couldn’t grab the crabs with my own hands and tear them apart.

You probably don't like shit ragdoll made by lazy\untalented devs.

Ragdoll in stuff like tlou2 or rdr2 is weighty and realistic most of the time (rdr2 is more advanced but also glitchier)

But even with that in mind, i take the shittiest ragdoll over the best precanned animation that get old by the 20th times you see them any day of the week.

Vr games are not the only games where you kill stuff, every game where you kill stuff should have visual variety for the kills.

 

[HL3.exe]

I really dislike the ragdoll physics. There’s no weight to anything and makes the engine feel
cheap to me. VR is a different story, though. Your point on “visual variety” is very well taken.

I should be able to chop an enemies limbs off (not scripted) if I wanted to. Heck, when I played Half Life Alyx, it bothered be that I couldn’t grab the crabs with my own hands and tear them apart.

Guessing you mean 'cheap ragdoll physics'? Then I agree. But that's because of cheap implementation with off-the-shelf solutions. Their are really good looking ragdoll implementations like the ones in Rockstar titles, but they are really expensive, financially and hardware resource wise, still to this day because of a decline in single threaded CPU speeds. (Single core speed are very important for deterministic game-logic simulation)

Bending Moore's Law: Single-Core Stagnation and the Cloud | Man Group

Moore's Law is broken - that's why we built our own cloud.

www.man.com

"For single-core CPU performance, Moore's Law (which states that the number of transistors in new integrated circuits will double around every two years) isn’t quite as true as it once was. Whilst the law has kept consistent, single-core CPU performance has stagnated over recent years, and whilst it is still trending upwards, the improvement has slowed considerably. Back in the day, when Moore's law was still in full swing, CPU performance tracked transistor count, and it made perfect sense to replace hardware every two to five years. However, our own internal benchmarking shows that for real life workloads, a current compute core is only about 1.6 times faster than a comparatively ancient CPU from 9 years ago."