Shawn Layden's advice on making games faster and cheaper | GI Sprint
As part of GI Sprint, The former head of PlayStation Worldwide Studios on ways to bring down the cost of games
It was back in 2020 when Shawn Layden began sounding the alarm around rising game development costs.
Speaking as part of Gamelab Live, Layden said the current rising cost of AAA game development was become unsustainable, and that when the industry moves to the PS5 and Xbox Series X generation (which was due to take place later that year), the industry might struggle to grow.
Now, four years later, we have caught up with Layden as part of our new GI Sprint series of video, podcasts and articles, which is all around lowering the cost of game development.
"Sadly, it does my heart no good that I think I was right," Layden tells us. "And it wasn't any great prediction. It was watching trend lines of over 25 years of gaming. The numbers only go in one direction. Games don't get any cheaper, they don't get shorter, they get more complex, they become more costly.
"The large blockbusters, when people are swinging for the fences, they're coming in at $150 to $250 million bracket, and that is a huge burden on the game development business model, on the publishers for carrying that, and [that's led] to some of the contraction in the marketplace that we've seen."
Over the next three weeks, we'll be exploring numerous ways that games developers can make games quicker, and therefore at a lower cost. And as part of our opening podcast, Layden shares his views on what studios might want to consider. You can watch the podcast below, download it here, or find it on your podcasting platform of your choice.
Make your games shorter
One thing developers should ask themselves is whether bigger really is better?
"We live in a world where only 32% of gamers actually finish the game, so we're making a lot of game that 68% of the people aren't seeing," Layden explains. "So should we continue to build games that are unlikely for most of the people to even see the end of it?
"It costs to build to the end. You can tighten that up. If you can make your games on a shorter timeline that will reduce your cost. It'll get you to market faster, you can please your customers sooner rather than always telling them to wait 45 years for your next opus. I think we just have to reexamine how we present ourselves and our games to the gaming public."
"We live in a world where only 32% of gamers actually finish the game, so we're making a lot of game that 68% of the people aren't seeing"
Layden believes length is not necessarily the deal breaker that it used to be.
"It was such a big deal in the early days of gaming. In the PlayStation 1, 2 and 3 generations, [length] was like your top review poin. We kept judging games by you know how much gameplay you get for your dollar. And maybe that was a decent metric back in the times when the average gamer was late teens/early 20s, which means they're time rich and money poor, so having to sit down for that long of a sesh to get through some huge RPG seemed reasonable. I just think now the average age of gamers is approaching early 30s, you got the flip, they're more money rich and time poor. You have to really kind of strap on some free time if you're going to sit down with Red Dead Redemption 2 and get through that.
"We need to understand what our customers are looking for. Do they want a high impact, high enjoyment piece of gaming, which may not include large sections of the game where you're going on a quest to find this blue rock to bring it to the red Troll and he'll open a door for you. It's just kind of like burning time. It's called grinding for a reason."
Stop chasing photorealism
So much investment is going into areas such as ray-tracing, but are we spending too much time, effort and money on visual improvements that the majority of our players aren't noticing?
"We've made a lot about the visual quality of games, the graphic quality, the resolution, the near photorealism that so many games seem to chase," Layden says. "And our fans thought that was a was a was a noble journey, and we saw the difference between graphics on PlayStation 1 where Lara Croft is 800 polygons and, if you squint, kind-of looks like a person. And now we get to the highly realized modelling. But did it improve the gameplay? Did it improve the story?
"I don't believe you can get across the uncanny valley, I think that will always be just five steps ahead. So instead of chasing that, let's go back to exciting game design. I love a good anime. I love highly realized animated characters. They are exciting, because they can tell a different kind of story."
"We're in the realm of graphical differences that only dogs can hear"
Layden says that we're approaching a point where the improvements we're seeing on the screen are not even being noticed by most players.
"The console war began as a missile race. With each side trying to push the edge of tech. People talked about teraflops all the time without really understanding what it meant. You had all these different kind of metrics that people were throwing around," he says.
"But we've got to the point now where you have advanced ray-tracing and most of the platforms can do 60 frames per second, some can do 120, which your eye can't register anyway. I think we're at the edge of that universe now, we're in the realm of differences that only dogs can hear. And maybe that's not where the emphasis should be anymore. So let's go back to… what can I do that would be amusing, entertaining and interactive so that someone would want to spend their money and time, and enjoy themselves in a way that means they get value for money, and we can continue to pay at least living wages or better to the people who make them."
But will gamers really be ok with shorter games that aren't pushing the graphical side?
"Well you don't say the quiet part out loud [Laughter]. Again, let's stop looking at the game as merely a collection of time-based activities with a certain visual acuity to it. Let's create an activity that would be fun, and these are the rules, these are the characters… and not be begin to deconstruct a game simply based on whether its frame rate held 60 every time. You don't deconstruct movies like that."
Get the machines to do more of the work
On the subject of AI, Layden believes there is a need for programmers to start building tools that will enable some automation in the development process.
"The way we make games has essentially remained unchanged in 40 years," he says. "When the game gets more complex, or there's more heavy lifting to be done, or we need more art assets… typically we've just thrown people at it. Or we hire some people in Malaysia or Eastern Europe and throw them at it. Labour has always been the default response for increased workload and scope of game. People in the interactive entertainment business, certainly on the coding and programming side, are some of the smartest minds out there in computer science. And we need to make the machine do more of the work.
"We need to get more heavy lifting out of the technology, build the toolsets and build the engines that can help. I look at what happened with the guys at Hello Games with No Man's Sky. A game with ultimately infinite scope, but it's essentially done by less than 10 people because they spent a lot of time building the pipeline, the toolset, which allowed them to create over and over, making the machine do most of the procedural heavy lifting. We need to get more of that into gaming."
"We need to make the machine do more of the work."
Some of the AI tools will help with that, but Layden doesn't believe AI will be building games for us.
"AI will be an assisting technology," he says. "Of course, you've got some large business consulting group claiming that by 2030 50% of games will all be written by AI… that is not going to happen. AI only sees in one direction, which is backwards. It puts stuff together to make you think you're seeing forward, but you're really not, you're just seeing a rehash of backward. AI is kind of like the really eager intern that you can say: 'hey give me nine pages on something' and they're like 'sure boss' and they crank it out. But you do have to fact check. AI hallucinates and goes off the rails.
"It's a great tool for taking a bunch of knowledge into a space and summarizing it in four paragraphs. It's really good at that. I think we'll see it doing more first draft work. Some of the video AI kits that are out there means you can mock up a scene pretty fast to get an idea of whether something looks interesting. So in that ideation phase it can help speed things up. But I don't see it writing games anytime soon."
Be disciplined and murder your darlings.
Finally, Layden says be strict with deadlines, and that we can all learn a thing or two from how sports games are made.
"One thing we see in studios is they will hold on to a thought or an idea or a wish, and you have to constantly interrogate what it is you think you're doing," he says. "Speed up proof of concept, speed up proof of tech, so you can make the hard call and say 'that's not working'. The people who continue to try to make the thing work over an extended period of time… they're thinking if we just do a little bit more we can get there, but the rest of your game is sometimes just waiting for that mechanic to be established and it's going to slow you down. You've got to have a disciplined idea of what you want to make and how you going to make it. Hold your deadlines tight. Don't be slippy slippy on your alpha or beta Target.
"You've got to have a disciplined idea of what you want to make and how you going to make it. Hold your deadlines tight. Don't be slippy slippy on your alpha or beta Target"
"Look to teams that do it right. Any team that builds an annual sports game… that's a wonder to behold. They do a new game in nine months every year. At my old shop, where they do MLB: The Show in San Diego… Major League Baseball is not changing opening day. It's April 1st every freaking year. You can say 'well, they know what they're doing. It's the same 32 teams, it's the same stadia, it's the same players where they just have to move them around.' Yeah, because they've identified what their variables are year to year, then they have a list of things they can add. They take care of all the maintenance of moving the players around, updating stadia, getting new face scans… and then here are some new things to add. But there's a date when you go 'that's it, feature locked, we're done'.
"You want to be able to have [time for] people to be inspired and come up with new ideas and really have a flash moment where they've solved the problems of the universe. But you've got to have a strong hand in production management. It's: 'That feature is great, but we've missed the window to put it in. If we do, it'll break all these other things. So hold on to that, we'll come back to it."
Ultimately, if we can get back to making games in three years, the thought of dropping an idea or a feature becomes more palatable.
"If you can turn games around in two to three years rather than five to six years… it's easier to set aside an idea because you go: 'I'll get back to it in two years'. In the current model, if you don't get this idea in now, there may never be a chance to get it in."