Graphical Improvements: Why aren't realistic physics given as much importance as others? (Next Gen Wishlists)

[HussiZooT]

I'm jumping straight to the point - Realistic physics in games make or break it for me in terms of immersion.

Now immersion isn't the same as enjoyability of a game - aka Mario, but when a game is trying to have a realistic world, realistic design, etc. physics that are outdated just don't set right with me.

Cardboard boxes gliding down the floor, shooting them only makes them move away from you without causing any actual damage to the box, or it just magically vanishes after taking too many shots should not be acceptable in the next generation of gaming,

Also, realistic physics shouldn't only relate to things breaking properly, but also that whatever has been broken, stays broken.

God of War 2018, such a graphical masterpiece, has physical objects that disappear literally within seconds of you breaking them. Why is this acceptable?

I don't get why more people aren't asking for realistic physics over 4K or HDR or 120FPS games. Why aren't physics given more importance in gaming?

EDIT 1 - Control previews are out and it's exactly what I'm talking about.

 

[Aokiji]

lol do you realize the cost of keeping such physics? you think they dont care to add them?

 

[Rodjer]

God of War 2018, such a graphical masterpiece, has physical objects that disappear literally within seconds of you breaking them. Why is this acceptable?

I don't get why more people aren't asking for realistic physics over 4K or HDR or 120FPS games. Why aren't physics given more importance in gaming?

Click to shrink...

Becuase physics is really taxing on hardware and PS4/Xbox One CPUs are 6 years old ass hardware.

 

[-Tetsuo-]

That kind of tech has insane upkeep and absolutely is not worth it

 

[ckareset]

Because realistic physics aren't fun to play. Just like people asking for realistic AI. It's more possible then people think. But that doesn't make it fun

 

[nsilvias]

the same reason better ai isn't a priority, its not fun and the resources aren't worth it.

 

[Lashley]

Costs.

 

[Absoludacrous]

They're complicated to implement, difficult to bug test, and very resource heavy on CPUs. It's a bit different than asking for HDR.

 

[Teeth]

Because what you are asking for doesn't scale well. The computation cost vastly outstrips what can be done with other visual flourishes that an audience would generally care for.

Also, physics objects with proper collision and interference creates gameplay issues for player and enemy characters. They can be "fun" in a sandboxy sort of way, but they also allow people to completely break things. They also can make player navigation and mobility really messy.

They are also extremely difficult to QA unless you don't care about your game looking/acting like Goat Simulator.

 

[Sparks]

Because the amount of resources a proper physic system uses is just way over budget for what actual players notice. It's better to invest more time in the character physics, animations and environments than it is to have a fancy cardboard box breaking. Also, with the higher fidelity of characters this generation, keeping them on the floor is insanely expensive. We already have to lower Environmental art significantly if there are more then 2 characters on screen at one time.

Plus, having things rely on physics is super inconsistent, hard to test and allows for too many variables that can affect things negatively. Better to use that time testing more critical bugs.

There are exceptions to this, but usually they are budgeting that stuff as part of the initial discussions for what is important with their game or designing stuff around that.

(We all desperately want it on the development side as well)

Because what you are asking for doesn't scale well. The computation cost vastly outstrips what can be done with other visual flourishes that an audience would generally care for.

Also, physics objects with proper collision and interference creates gameplay issues for player and enemy characters. They can be "fun" in a sandboxy sort of way, but they also allow people to completely break things. They also can make player navigation and mobility really messy.

They are also extremely difficult to QA unless you don't care about your game looking/acting like Goat Simulator.

Click to shrink...

Yea, basically this.

 

[Stopdoor]

Do you really need to have the forum spell this out entirely for you? Seems like you haven't even tried to think about why they might not do this.

 

[low-G]

lol do you realize the cost of keeping such physics? you think they dont care to add them?

Click to shrink...

I want to know why game physics aren't usually as good / elaborate as those of Myth (the 1997 strategy game). That game had physics that affected gameplay. Instead most gameplay is trivially deterministic these days.

 

[Nooblet]

Having clutter that physically react and have persistence has exponential increase in computational power requirement. You start out with 3 objects but break them, now you have 10 objects to keep track of...you break them all in half now you have 20 objects to keep track of. The computation requirement for calculating that won't be double, it'd be exponential instead. And for most players there's no difference between objects that can be shattered to 50 pieces and objects that can be shattered to 5 pieces. Not to mention how this would affect the player's animation...which will have its own computation cost if it were to take all those clutter into account during animation.

Because realistic physics aren't fun to play. Just like people asking for realistic AI. It's more possible then people think. But that doesn't make it fun

Click to shrink...

Realistic destruction has led to the creation of the most unique shooter this gen (Siege), so it is indeed fun to play and makes the game/map infinitely deep & replayable....when done right.

 

[HussiZooT]

I totally get that it's taxing on systems and it's not easy to implement with so much fidelity, but then... is this where gaming development just gives up?

Gaming as a whole keeps advancing all the time so for how long do we have to wait to actually get more realistic physics in gaming?

Microsoft went with the cloud for Crackdown 3 to balance the load, why can't similar things be attempted for future games?

I'm not asking for heavy physics details in current gen games, the console hardware is 6 years old now. But for the next gen... isn't it a viable request to actually try and improve physics?

 

[Thatguy]

People have been asking for better physics and AI for generations now. We'll get a little more, but most of the focus will just be even better fidelity, draw distance, and lighting effects as always.

 

[lazygecko]

Just Cause 3 went for increased physics instead of making a greater leap in visual fidelity. Technical issues aside, I'd say that was worth it.

 

[Bhonar]

c'mon this is obvious, it's not exactly easy and resource effective

 

[foxuzamaki]

Play BOTW and you too will learn the difference between realism and reality

 

[s y]

I really enjoy games with nice physics. Like in Watchdogs 2, everything besides light poles behave pretty realistically and stick around for a long time. Makes driving and crashing more fun.

 

[mxbison]

I always love destructible objects and environments. I'm very surprised to read people in this thread saying it's not fun and nobody cares.

 

[delete12345]

I totally get that it's taxing on systems and it's not easy to implement with so much fidelity, but then... is this where gaming development just gives up?

Gaming as a whole keeps advancing all the time so for how long do we have to wait to actually get more realistic physics in gaming?

Microsoft went with the cloud for Crackdown 3 to balance the load, why can't similar things be attempted for future games?

I'm not asking for heavy physics details in current gen games, the console hardware is 6 years old now. But for the next gen... isn't it a viable request to actually try and improve physics?

Click to shrink...

How realistic can you get? Ever heard of diminishing returns, where the more realistic the games are getting, the less differences players would notice.

Game physics on the other hand, is a different topic. If it's fun, it's fun. If it breaks immersion, then it's not fun. There's a fine line drawn between those two.

Graphics however is so blurred you can't tell the difference from an old graphics to new graphics that are 1 to 3 years apart on PC. For console generations, could be.

 

[Deleted member 40102]

Personally I never seen realistic physics adds excitment to the game as much as graphics I would rather have graphics with 60fps over anything for single player game and 240fps for multiplayer. Unless of course physics related to gameplay then its another story.

 

[Nooblet]

I totally get that it's taxing on systems and it's not easy to implement with so much fidelity, but then... is this where gaming development just gives up?

Gaming as a whole keeps advancing all the time so for how long do we have to wait to actually get more realistic physics in gaming?

Microsoft went with the cloud for Crackdown 3 to balance the load, why can't similar things be attempted for future games?

I'm not asking for heavy physics details in current gen games, the console hardware is 6 years old now. But for the next gen... isn't it a viable request to actually try and improve physics?

Click to shrink...

Just Cause 4 came out 3 months ago, decided to focus on destruction AND performance. Game got a lot of flack for looking worse than JC3 despite improvements elsewhere. Basically people always want improvements and only improvements, it can't be improvement in one area at the expense of going worse than the previous game in some areas...cause then that'd be a negative.

Additionally unless the game is designed around it to make use of it (Just Cause, Siege), this kind of physics is pointless in practical terms and an added headache that just adds to development time and affects the game's performance....just so people can get their 2 hour fill of breaking objects before ignoring it and getting on with usual game. It'd be fun to break everything for a while, but after a certain point you'll stop doing it unless it's integral to the gameplay itself like those two games I mentioned.

 

[Teeth]

Microsoft went with the cloud for Crackdown 3 to balance the load, why can't similar things be attempted for future games?

I'm not asking for heavy physics details in current gen games, the console hardware is 6 years old now. But for the next gen... isn't it a viable request to actually try and improve physics?

Click to shrink...

Crackdown 3 is a perfect example of why it isn't happening: because it's a massive design problem beyond being a technical and artistic nightmare.
Notice how multiple teams took a crack at Crackdown 3 and eventually it got to the point of stripping things way back to tiny environments just for the multiplayer. It's not because the teams involved were incompetent. That's a design nightmare that is extremely difficult to make work.

Red Faction Guerilla is an excellent example of what you want, but that game was designed from the ground up for it. You can't just add that type of destruction to Uncharted and have it just work out.

You say you want it in next gen, and undoubtedly, there will be games that have what you ask for (i wouldn't be surprised if the next Elder Scrolls uses more physics jank and I also the next Battlefield games will probably expand on destruction some more....though they seem to go back and forth with the destructible fidelity). But these are games that have that type of interaction in their DNA. Taking the template of Last of Us and "just add physics" isn't really in the cards (though undoubtedly, LOU2 will have some neat flourishes).

Some of it is just money. But it's all a give and take. If you want realistic animations like say Spiderman, a lot of that has to prescriptively done, because adding unpredictable colliders will often make your game look like Human Fall Flat.

 

[HussiZooT]

Also, just to clarify, I'm not saying that every game needs to have realistic physics, but games striving for it, should really really do. Visceral shooters that are more confined like Rainbow Six: Siege would benefit from more detailed environmental destruction and bombastic shooters such as Battlefield could do with larger scale destruction.

While I like how physics have been implemented in both the games, there's always room for improvement, especially with BFV. Even a game like Hitman 2016 has a great physics system. Try going guns blazing in one of your runs and you'll see how reactive the environment is. It's quite astonishing that a game revolving around stealth has such good physics system. But yes, since the hardware of current gen consoles is quite old, the framerate dips quite a bit.

More games that revolve around destructive combat could benefit heavily from detailed physics.

But I get what you all are saying. The returns for such implementation is quite less compared to the hardwork the teams and systems have to put in.

So I guess I'll just have to wait until PS6 to maybe expect something in line of what I'd love to see from games.

 

[MP!]

Eats up too much processing power

 

[gabdeg]

I have high hopes for better physics next-gen and I think we'll get them. Current console CPUs were just too weak for significantly better physics over last gen. The rumored Zen cores however should have plenty of processing power for more elaborate physics simulation if fully utilized.
The GPU bump won't be that big this time around, at least compared to the mid-gen consoles, so I hope devs really make use of that additional CPU power to show off what can be done with next-gen hardware.

 

[Duxxy3]

If GTAIV taught me anything, it's that realistic physics can make a game completely unfun.

 

[HellforLeather]

Its difficult and there is almost zero marketing value in it.

 

[Fitts]

Why would anyone want games to be less fun? On the technical side, 60fps should be made priority since that actually impacts gamefeel. Otherwise, focus on fun mechanics. And stylize your visuals for Christ's sake.

 

[Aokiji]

Microsoft went with the cloud for Crackdown 3 to balance the load, why can't similar things be attempted for future games?

Click to shrink...

and how did that turn out?

 

[808s & Villainy]

God of War 2018, such a graphical masterpiece, has physical objects that disappear literally within seconds of you breaking them. Why is this acceptable?

Click to shrink...

Is there some gameplay purpose to keeping them? Just do you can marvel at the pretty, highly CPU taxing, uselessness of the object in pieces? Things like this are done because keeping them serves little purpose and uses CPU/RAM for no reason

 

[Vulcano's Assistant]

Simulators, like Gran Turismo try to. As of current game design only a handful of genres really benefit their gameplay from realistic physics. A lot of other examples are more or less an extension of graphics, but a much more taxing one.

 

[Mani]

Baked lighting was giving the best result visually past gens and was more performance efficient but it meant most geometry had to be static, hence no destruction.

 

[laxu]

Next gen consoles should be able to handle them better as their CPUs aren't trash tier.

There's still not much sense to keep breakable stuff around.

I think physics based systems are far better when used for actual gameplay purposes. They add both consistency with the real world and also unpredictability. BotW for example makes great use of them and so does RDR2.

 

[Crossing Eden]

Performance costs.

Just Cause 4 came out 3 months ago, decided to focus on destruction AND performance. Game got a lot of flack for looking worse than JC3 despite improvements elsewhere. Basically people always want improvements and only improvements, it can't be improvement in one area at the expense of going worse than the previous game in some areas...cause then that'd be a negative.

Click to shrink...

^^^^

 

[Quad Lasers]

I don't get why more people aren't asking for realistic physics over 4K or HDR or 120FPS games. Why aren't physics given more importance in gaming?

Click to shrink...

You don't get why people want well performing games versus useless visual fluff that have no bearing on games that have no mechanical focus on destruction?

 

[ShapeDePapa]

The interactivity of the world/physics in BOTW should be the minimum to expect in games.

 

[Vulcano's Assistant]

Why would anyone want games to be less fun? On the technical side, 60fps should be made priority since that actually impacts gamefeel. Otherwise, focus on fun mechanics. And stylize your visuals for Christ's sake.

Click to shrink...

OP is talking about visual effects that wouldn't change the way the game plays. You know glass breaking and whatnot. They could be stylized as a Disney movie and still benefit from physics if the performance was there (which I think most agree it isn't there so it is not a priority worth tackling). Let's not freak out because s/he is using the R word. It is superficial stuff, but it's not going to change the games.

 

[Crossing Eden]

The interactivity of the world/physics in BOTW should be the minimum to expect in games.

Click to shrink...

No it shouldn't, because a lot of games aren't designed like BOTW nor do they have the same gameplay loop or focus. I know absurd OPs tend to lead to absurd posts but jesus christ.

Why would anyone want games to be less fun? On the technical side, 60fps should be made priority since that actually impacts gamefeel. Otherwise, focus on fun mechanics. And stylize your visuals for Christ's sake.

Click to shrink...

Devs DOOOOO stylize their visuals. They stylize them in a way that emulates the look of films. There's a huge misconception among gamers that stylizatio ONLY=toon shaders and models that resemble caricatures. And a focus on fun mechanics instead of narrative is one of the trends of this gen.

 

[Revolsin]

Big agree with OP.

BotW's realistic(given its world) physics are a major reason that game is so well-loved. They're so intricately designed to interact with every facet of the world, it's positively fascinating to watch what people can do with it.

 

[devSin]

I want to know why game physics aren't usually as good / elaborate as those of Myth (the 1997 strategy game). That game had physics that affected gameplay. Instead most gameplay is trivially deterministic these days.

Click to shrink...

The physics model in Myth was very simplified.

You can think of it as basically a projectile system, with units all having a corresponding pre-determined bucket of parts that spawn whenever they get chunked and inherit a velocity based on the killing blow. There wasn't a real physics simulation of the game world.

It was still great fun, though.

 

[Vulcano's Assistant]

I'm still holding on for a game to use temperature as an In depth mechanic. Like those browser sand games. Let me melt metal and evaporate water.

 

[LinkStrikesBack]

The interactivity of the world/physics in BOTW should be the minimum to expect in games.

Click to shrink...

That's not a realistic expectation. BotW took many years of development and was the most expensive game Nintendo ever made, and in a way, had Nintendo's entire future riding on it.

You can't possibly expect the same level of expense as a minimum in all games.

 

[elenarie]

Also, just to clarify, I'm not saying that every game needs to have realistic physics, but games striving for it, should really really do. Visceral shooters that are more confined like Rainbow Six: Siege would benefit from more detailed environmental destruction and bombastic shooters such as Battlefield could do with larger scale destruction.

While I like how physics have been implemented in both the games, there's always room for improvement, especially with BFV.

Click to shrink...

We already have thousands of instances of networked assets whose state the servers have to track and update 60 times a second for 64 clients for the duration of a whole round. Just states, which are as close to predetermined binary representations as you can get.

Imagine having to add what effectively is infinite amount of states for each of those instances, and send that information 60 times a second to 64 clients.
(well, not every single instance of every asset, but you get my point)

Consoles currently have hard time with updating 30 times a second to 64 clients as it is. :)

 

[ArnoldJRimmer]

Because realistic physics aren't fun to play. Just like people asking for realistic AI. It's more possible then people think. But that doesn't make it fun

Click to shrink...

I disagree. A lot of good physics effects add to... well the physicality fo the world, improving immersion, and at least for me, improving the fun.

 

[ArnoldJRimmer]

We already have thousands of instances of networked assets whose state the servers have to track and update 60 times a second for 64 clients for the duration of a whole round. Just states, which are as close to predetermined binary representations as you can get.

Imagine having to add what effectively is infinite amount of states for each of those instances, and send that information 60 times a second to 64 clients.
(well, not every single instance of every asset, but you get my point)

Consoles currently have hard time with updating 30 times a second to 64 clients as it is. :)

Click to shrink...

There's probably a lot of ways you can have physics simulations not being reliant on the state of the game on the cloud and be kept local instead.

Not that staying local means there's no overhead of course.

 

[elenarie]

There's probably a lot of ways you can have physics simulations not being reliant on the state of the game on the cloud and be kept local instead.

Not that staying local means there's no overhead of course.

Click to shrink...

Never trust local clients in multiplayer games. :)

 

[ShapeDePapa]

No it shouldn't, because a lot of games aren't designed like BOTW nor do they have the same gameplay loop or focus. I know absurd OPs tend to lead to absurd posts but jesus christ.

Click to shrink...

That's not a realistic expectation. BotW took many years of development and was the most expensive game Nintendo ever made, and in a way, had Nintendo's entire future riding on it.

You can't possibly expect the same level of expense as a minimum in all games.

Click to shrink...

But I want amazing physic and interactivity everywhere. 😭😭😭

 

[Deleted member 12790]

There's probably a lot of ways you can have physics simulations not being reliant on the state of the game on the cloud and be kept local instead.

Not that staying local means there's no overhead of course.

Click to shrink...

dirty little secret about floating point operations: they aren't deterministic among different hardware. Different processors will yield different results in different situations. This makes handling physics calculation client-side a near impossibility for large scale games, as there needs to be constant error checking among all clients otherwise.

This is what makes pixel perfect point sampling so difficult to do across multiple GPU vendors. You either need to waste memory to ensure a sort of error buffer surrounding pixels to be sampled, or just write lots and lots of edge cases for various kinds of vendors.