Game Economics, Graphics and AI ...

[DavidGraham]

Speaking to The New York Times, several game developers and some industry figures spoke out about how the gaming industry's AAA studios couldn't reasonably handle the stress of creating cutting-edge graphics

particularly in light of major waves of layoffs throughout the past two years, and several high-fidelity AAA games underperforming in the market

As former Square Enix executive Jacob Navok noted to The New York Times, "It's very clear that high-fidelity visuals are only moving the needle for a vocal class of gamers in their 40s and 50s. But what does my 7-year old son play? Minecraft. Roblox. Fortnite."

When millions are happy to play old games with outdated graphics — including Roblox (2006), Minecraft (2009) and Fortnite (2017) — it creates challenges for studios that make blockbuster single-player titles. The industry’s audience has slightly shrunk for the first time in decades. Studios are rapidly closing and sweeping layoffs have affected more than 20,000 employees in the past two years, including more than 2,500 Microsoft workers.

Optimizing cinematic games for a narrow group of consumers who have spent hundreds of dollars on a console or computer may no longer make financial sense. Studios are increasingly prioritizing games with basic graphics that can be played on the smartphones already in everyone’s pocket.

Industry figures joke about how a cartoony game like Luigi’s Mansion 3 on the Nintendo Switch considerably outsells gorgeous cinematic narratives on the PlayStation 5 like Final Fantasy VII Rebirth.

We have a market that has been in growth mode for decades,” Ball said. “Now we are in a mature market where instead of making bets on growth, companies need to try and steal shares from each other

Rami continues, "How can we as an industry make shorter games with worse graphics made with people who are paid well to work less? If we can, there might be short-term hope. Otherwise, I think the slow strangulation of the games industry is ongoing."

Reitman sees a future where most of the heavy costs associated with cutting-edge graphics are handled by artificial intelligence. He said that manufacturers were working on creating A.I. chips for consoles that would facilitate those changes, and that some game studios were already using smart algorithms to improve graphics further than anything previously seen.

Some independent developers are less convinced. “The idea that there will be content from A.I. before we figure out how it works and where it will source data from is really hard,” said Rami Ismail, a game developer in the Netherlands.

Video Games Can’t Afford to Look This Good

The gaming industry spent billions pursuing the idea that customers wanted realistic graphics. Did executives misread the market?

www.nytimes.com

 

[see colon]

Is this an article, or just a slideshow of quotes? It's paywalled so I can't read it myself.

 

[arandomguy]

If you look at the points brought up it shows the problem with current incentive model.

The demographic that wants those high fidelity AAA SP games is willing to spend money on their hardware and believes the content side should deliver more to them based on their hardware spend. The GaaS demographic spends more on the content side and believes the content side should deliver more to them based on that content spend.

You can see the issue once you compare the above two and why the content side is being incentivized to the latter group.

This is also going to be why there is going to be increasing onus on the hardware side to drive graphics forward.

 

[trinibwoy]

Gamers in their 40’s and 50’s grew up with what’s now considered extremely basic graphics. The same thing will happen when Fortnite and Minecraft kids grow up. So unless you want to completely ignore the seasoned gamer demographic you will need to push the envelope.

This isn’t just a graphics problem either. Are they also saying that animation, physics, voice acting, writing etc are too expensive to make?

I get that graphics tech requires investment but that should be amortized over many years and many games. You’re not reinventing AO or TAA in every new game. I think the industry has a major efficiency problem because I don’t understand why every new game is so expensive to make when it’s mostly using well established tech.

Content is another cost but it’s the same question there - are people really designing doors and trees and fire hydrants from scratch every time or just downloading assets from a library curated over many years?

 

[Dominik D]

I think the industry has a major efficiency problem because I don’t understand why every new game is so expensive to make when it’s mostly using well established tech.

Because it doesn't reuse well established tech. If your production cycle is 2-3 years (and it's often longer for AAA games), by the time you're done the de facto standard tech stack is massively different from what you started with. Either you rolled with changes as you went - which resulted in waste and more money spent - or you didn't. But you still have to catch up for your next game (which will be more expensive since you have to adapt to newer technology).

If you started working on a game in late 2020 using UE4, you faced impossible choice mid 2022: continue on UE4 or switch to UE5? Your studio will switch at some point whatever you decide to do right now (as in: in 2022) or not, so your established tech isn't really established. But even without significant events like the introduction of UE5 was, your narrative, localization, graphics, level pipelines will be ducktaped together for your first game. And your second game - you're expected to have more stuff in it so even if you pay some of the existing debt, you're now saddled with some more debt. The same goes for any subsequent game. You're never working on a game using "established tech".

On top of that if you're not committed to a single genre, you've got even taller hill to climb. And if you are a turn-key developer for various publishers, you're fucked even more. Even if you had tech for game A for publisher K, you typically can't use it for game B for publisher M. So every. single. game. you. make... has to be written from scratch.

 

[trinibwoy]

Because it doesn't reuse well established tech. If your production cycle is 2-3 years (and it's often longer for AAA games), by the time you're done the de facto standard tech stack is massively different from what you started with. Either you rolled with changes as you went - which resulted in waste and more money spent - or you didn't. But you still have to catch up for your next game (which will be more expensive since you have to adapt to newer technology).

If you started working on a game in late 2020 using UE4, you faced impossible choice mid 2022: continue on UE4 or switch to UE5? Your studio will switch at some point whatever you decide to do right now (as in: in 2022) or not, so your established tech isn't really established. But even without significant events like the introduction of UE5 was, your narrative, localization, graphics, level pipelines will be ducktaped together for your first game. And your second game - you're expected to have more stuff in it so even if you pay some of the existing debt, you're now saddled with some more debt. The same goes for any subsequent game. You're never working on a game using "established tech".

On top of that if you're not committed to a single genre, you've got even taller hill to climb. And if you are a turn-key developer for various publishers, you're fucked even more. Even if you had tech for game A for publisher K, you typically can't use it for game B for publisher M. So every. single. game. you. make... has to be written from scratch.

Isn’t this stuff pipelined though? In an ideal world you would have your engine team, artists, writers etc working on a somewhat independent cadence with the engine guys generally ahead of everyone else. If the problem is that the engine team is still tweaking stuff for day zero patches that highlights the problem right there. Tech is forever changing sure but why are game developers forever scrambling on the execution side of things?

Engines can evolve continuously but feature sets and tools should be locked down and stable for the artists and content developers that use them for a given title. Why is that so hard? Is it overreach? Poor estimates? Scope creep?

 

[QPlayer]

Developers should draw a line and finish their game with the technology they started on time and more cost-effectively. Why do they think that their game will not be beautiful if it is based on an earlier technology? Graphics aren't just Tflops and feature sets...

Use aesthetics, art, style!

 

[Dominik D]

Isn’t this stuff pipelined though? In an ideal world you would have your engine team, artists, writers etc working on a somewhat independent cadence with the engine guys generally ahead of everyone else. If the problem is that the engine team is still tweaking stuff for day zero patches that highlights the problem right there. Tech is forever changing sure but why are game developers forever scrambling on the execution side of things?

Ideally it would be pipelined but due to how financing works in gamedev, very few studios can actually afford to do that. I gave a slightly longer explanation here:

Mark Darrah (Former Bioware) on why AAA games cost so much

This game was made by a single person a less than a month apparently. So I believe that gaming is just too inflated. It's a load of bought assets and a super basic gameplay loop. It proves nothing about nothing. All this shows is the entry-level creation, what's possible by indies, is a much...

forum.beyond3d.com

Engines can evolve continuously but feature sets and tools should be locked down and stable for the artists and content developers that use them for a given title. Why is that so hard? Is it overreach? Poor estimates? Scope creep?

It's not hard, it's simply unrealistic. What do you do with your content people while programmers "finalize" tech for the game? What do you do with programmers while artists produce assets using locked down tech? The handoff point is not only unclear, it differs between systems. Games aren't built from siloed pieces and (staff) specialization for all domains (design, art, code, audio) is a must given how complex things are.

 

[trinibwoy]

Ideally it would be pipelined but due to how financing works in gamedev, very few studios can actually afford to do that. I gave a slightly longer explanation here:

Mark Darrah (Former Bioware) on why AAA games cost so much

This game was made by a single person a less than a month apparently. So I believe that gaming is just too inflated. It's a load of bought assets and a super basic gameplay loop. It proves nothing about nothing. All this shows is the entry-level creation, what's possible by indies, is a much...

forum.beyond3d.com

It's not hard, it's simply unrealistic. What do you do with your content people while programmers "finalize" tech for the game? What do you do with programmers while artists produce assets using locked down tech? The handoff point is not only unclear, it differs between systems. Games aren't built from siloed pieces and (staff) specialization for all domains (design, art, code, audio) is a must given how complex things are.

Building an engine while the car is running is always a recipe for disaster. It’s up to the industry to acknowledge that and adapt. I understand the reality of the current situation but everyone involved including publishers and financiers need to wake up.

The problem with idle hands is a tough one. Hiring and firing isn’t a solution like you said. Only thing I can think of is to pipeline projects. However if money is waiting to see the outcome for project N before financing N+1 the whole system is configured for failure and nothing will change.

 

[Shifty Geezer]

Hiring and firing isn’t a solution like you said.

Isn't this how it had been for a good decade or two though? A studio of 20-30 people works on the start, then scales up to >hundred for filling in the content, then out the door, thanks very much everyone, good luck with your next project wherever you go, and that core 20-30 people have a month's holiday and then get to work on the next thing?

 

[Dominik D]

Building an engine while the car is running is always a recipe for disaster.

This is how pretty much every game has been made for the past 50 years. Not every game was a disaster though, far from it.

It’s up to the industry to acknowledge that and adapt.

Sure. Why don't someone just implement this obvious improvement.

Only thing I can think of is to pipeline projects.

This is extremely hard and doesn't yield high utilization anyway. You have to be extremely, extremely good at bizdev or build your studio around infinite expansion (Bioware model). The problem is: you can't really expand indefinitely so the more you pipeline, the more projects will cannibalize each other within the organization leading to delays, departures, and losses (sad consequences of Bioware model).

What I'm trying to say is this: having overlapping projects isn't something nobody thought of until this thread. Many great studios tried to implement it and most at some point failed, because very few things are actually in their hands.

financiers need to wake up

Money people are, very often, fucking clueless and benefit from the system that makes it incredibly hard for rich people to fail. Games and movies are investments in content to them. They won't wake up because they aren't dreaming. Gamers are, thinking that there are some secret cabals making games "worse". It's not a secret, it's unchecked capitalism.

Isn't this how it had been for a good decade or two though? A studio of 20-30 people works on the start, then scales up to >hundred for filling in the content, then out the door, thanks very much everyone, good luck with your next project wherever you go, and that core 20-30 people have a month's holiday and then get to work on the next thing?

Sort of. Most studio heads recognize that you can't build game without the right people (they may not know what it means to have right people, but that's a different story). So the idea isn't to scale up and then cut, but rather to build a successful game and then have this cohesive team you've nurtured move on to another project. If you resist the temptation of growing beyond a reasonable size, you could sustain this. But some studios never stop growing and at some point simply implode. This typically stems from the "fire in the brothel" phase of development where it looks like the only thing that can be done to finish a game is to drop additional people into the team.

Solution to this is potentially a Hollywood model where you have different specializations coalesce around unions with development studios signing favorable contracts with groups of people from different specializations. But this doesn't have to work if money people decide to starve the game dev out in response to (any) structural changes. It's unlikely (GD is still massively profitable) but it's something that could unfortunately happen.

 

[IQandHDR]

You also cannot alienate your target audience and thus hurting sales.
Or release games whith shitty stories.
A lot of studio are learning that the hard way atm.

So gaming cost and sales are conncted.

 

[see colon]

Or release games whith shitty stories.

To be fair, most games have shitty stories. Hell, look at the most successful recently released console - Nintendo's Switch, and their first party output. Zelda and Metriod, yeah they probably qualify as having good stories. Mario Kart/Party/Sports/Odyssey/Maker, not so much. Ringfit - nope. Switch Sports -nope. Arms - nope. Tetris 99 - nope. 1-2-Switch, nope. Labo - nope. Game Builder garage -nope. Nintendo games still sold very well without a story.

 

[IQandHDR]

To be fair, most games have shitty stories. Hell, look at the most successful recently released console - Nintendo's Switch, and their first party output. Zelda and Metriod, yeah they probably qualify as having good stories. Mario Kart/Party/Sports/Odyssey/Maker, not so much. Ringfit - nope. Switch Sports -nope. Arms - nope. Tetris 99 - nope. 1-2-Switch, nope. Labo - nope. Game Builder garage -nope. Nintendo games still sold very well without a story.

Generally the games I play have great stories, but there has been a influx of games disbanding/ignoring their target groups at the PC market and studios are now learning that is bad business (wich is weird, because common sense should not be a surprise to anyone).
People buying Nintendo I suspect do it for nostalgia, as I have never owned a Nintendo product and have zero interest in their games I am not qualified to speak about their games, other than they do not interest me, never have and most likely never will

 

[trinibwoy]

This is how pretty much every game has been made for the past 50 years. Not every game was a disaster though, far from it.

Oh yeah? How about the last 10 or last 5?

Sure. Why don't someone just implement this obvious improvement.

The other option is to accept the status quo. Is that a solution?

Money people are, very often, fucking clueless and benefit from the system that makes it incredibly hard for rich people to fail. Games and movies are investments in content to them. They won't wake up because they aren't dreaming. Gamers are, thinking that there are some secret cabals making games "worse". It's not a secret, it's unchecked capitalism.

The money people aren’t losing money in the current dysfunctional system? So who suffers when games don’t hit sales projections? If the issue is that they’re making less profit and not losing money then yeah nothing will change.

 

[Dominik D]

Oh yeah? How about the last 10 or last 5?

Last 50 includes "last 10" and "last 5", yes.

The other option is to accept the status quo. Is that a solution?

Sure, because there's one correct solution and everything else is status quo.

The money people aren’t losing money in the current dysfunctional system? So who suffers when games don’t hit sales projections?

Developers. Salaries are stagnant, and consequences of failures always result in workforce being fired to minimize cost. People investing in games (whether publishers or VCs) are also first to recuperate. So if game under-performs despite what otherwise would be considered decent sales, every last cent goes to whoever invested in the game, not to the studio or its employees.

Bizdev in the last two years was especially cut-throat. Small and medium size publishers scaled down investment in new titles. Their pipelines are typically 2 years deep so they can afford this conservative stance. The only games publishers were interested in were promising games near completion. So essentially the deal was "we'll help with your final 20% and marketing for ROI 150% or more". Very few studios can afford carrying game to alpha and then look for a publisher. But this is how market looks right now.

 

[Shifty Geezer]

People failing to self moderate and bringing up clearly OT RPSC topics will be removed from the forum.

 

[tuna]

People buying Nintendo I suspect do it for nostalgia, as I have never owned a Nintendo product and have zero interest in their games I am not qualified to speak about their games, other than they do not interest me, never have and most likely never will

Yeah the kids that buy/wish/want Nintendo products do if for nostalgia. Makes sense.

 

[tuna]

Building an engine while the car is running is always a recipe for disaster. It’s up to the industry to acknowledge that and adapt. I understand the reality of the current situation but everyone involved including publishers and financiers need to wake up.

To a certain degree you have to. Wukong would probably be a much better performing game if they were on mainline UE5 and not the old branch they forked off. The same goes for a lot of games.

Fornite is actually the future of how you develop games, always follow mainline.

 

[IQandHDR]

Yeah the kids that buy/wish/want Nintendo products do if for nostalgia. Makes sense.

Nostalgia from their parents
And I should mention that Nintendo has a sub 1% marketshare in DK, so I guess culture/geography is another big part of it.
It has been more than a decade since I last saw Nintendo unit IRL and around 2 decades since I dropped their controller in a "Yuck" moment aka touched one