Technical ambition, optimism, and timeframes: what we learned about The Witcher 4 from speaking to CD Projekt Red
"The next game we do will not be smaller, and it will not be worse¡ we don't want to go back."
Earlier this week, CD Projekt joint CEO Michał Nowakowski announced the next game in The Witcher franchise - otherwise known as Project Polaris and unofficially referred to by most as just The Witcher 4 - has finally entered what he described as "full-scale production - the most intensive phase of development".
It's good news for Witcher fans of course, although information remains incredibly thin when it comes to what exactly we can expect from the game itself, or even when we might hear more. All that's been publicly shared so far is that early-phase development was confirmed back in March of 2022; that the game won't "focus" on previous protagonist Geralt of Rivia; and that Geralt's voice actor, Doug Cockle, will still be "part of" the new game in some way.
But having spoken to a number of people from CD Projekt - including Nowakowski - over the past year for our long-read on what happened with Cyberpunk 2077 and how the studio turned it around, there are a few little nuggets of info we can share that might help piece together some extra context.
First up, a bit of chatter about process - albeit chatter that leads into a few other little hints at what we can expect about how things are going. One of the few other tidbits of already-public information is that CD Projekt Red has moved away from its own bespoke, internal REDengine to the more widely-adopted Unreal, which is owned by Epic Games.
"I know a lot of people are curious about it, so I'll try to explain," CDPR's vice president of technology, Charles Tremblay told me earlier this year, wryly joking that the decision has been a "very polarising question in general, for the company and for everyone". He explained that, contrary to popular belief, the REDengine wasn't the specific cause of the studio's troubles with Cyberpunk 2077's development, and nor was its severely troubled launch the key reason for the switch.
"The first thing I want to say again, to be sure, 100 percent clear, is that the whole team, myself included, are extremely proud of the engine we built for Cyberpunk. So it is not about, 'This is so bad that we need to switch' and, you know, 'Kill me now' - that is not true. That is not true, and this is not why the decision was made to switch."
Instead, it came down to the developer wanting to have multiple projects in development at once - namely the next Witcher game, the further-off Cyberpunk sequel codenamed Orion, and the still distant new IP codenamed Hadar. "The way we built stuff in the past was very one-sided, like one project at a time. We pushed the limit - but also we saw that if we wanted to have a multi-project at the same time, building in parallel, sharing technology together, it is not easy," he said.
The other factor was that the developer felt it could benefit from "having a good partnership with Epic, and working together on the technology," as Tremblay put it. "We can also help them to achieve their vision, to do open world game[s], and also they can help us too, from some technical perspectives, on some of the aspects that we would like them [to] and that we would like to not have to be focusing on too much - because in the end, we are game company, right?" In other words, it's a decision made with the intention of getting more of CDPR's developers working on the games themselves.
"So the idea was that we can push the technology, we can finally have all the technical people in the company working together on different projects, rather than super centralised into one technology that can very difficultly be shared between other projects.
"We can share expertise, share people, share knowledge. And also we wanted to be sure that we developed some of the technology correctly this time around - with our expertise, we know how to do things with the experience we had in the past, and now it's time to actually make it shareable across all the groups."
"It will be better than Cyberpunk - because for us, it [was] unacceptable"
One of the major lessons for CD Projekt after its issues with Cyberpunk 2077 was the studio's difficulty with balancing its ambition - and confidence in delivering on that ambition that bordered on "magical thinking," as joint CEO Nowakowski put it to me - with the realities of the studio's rapid growth and surprise obstacles of Covid-19. But Trembley, as with many other developers at CDPR, maintained that the team's ambition itself remains undimmed.
"Again, I will not say it's easy," he added, "but I think that we have some cool stuff going, and hopefully that will have some good showcase [of the technology]. The only thing I will say is that changing the tech for us does not change the fact that we always will be ambitious," he said. "And the next game we do will not be smaller, and it will not be worse. So it will be better, bigger, greater than The Witcher 3, it will be better than Cyberpunk - because for us, it's unacceptable [to launch that way]. We don't want to go back.
"Even if there will be some 'sweaty moments' and maybe even some bad stuff happening, still, I think that we will try everything we can to make it even more than what we achieved in the past years. So the technology should not be a blocker for us, basically."
Coming back to the next Witcher game moving up to "full-scale production", Tremblay also explained part of what that process looks like for CDPR. Tremblay's recent role, since moving up from being very much "in the trenches" on Cyberpunk 2077, has been to "pretty much oversee the tech development across all the products," as he described it. One aspect of that includes speaking with the various game directors across the studio's projects in North America and Europe, reviewing "what they are doing, challenging [them], approving when we go," and then making "recommendations to the board" about whether the studio is ready to press on with a project on the technical side of things.
This is a new process CDPR brought in as part of its response to the Cyberpunk launch, with the studio allowing more time for pre-production, and conducting regular reviews on whether the technology is ready for each individual project to move forward - including, in some cases, delaying that step where necessary.
"How it works in what we're doing," Tremblay explains, is "basically we have different stages of the project, where we have different requirements for tech. Then basically, with me and the tech directors, we nail down - we agree on 'this is what we want to nail' - and we work together to make sure this happens."
Then, "if we want to go into pre-production, we will prepare all the tooling to scale up the production and the [number of] content people. If it's not ready, or it's just a specific thing that is not ready, then we can discuss, okay, what can we do next? How are we going to approach it? And we can also discuss [whether] to delay - delay not as in delay release, but delay in terms of like, before we scale up production. And then we go to the board or the other people, and we discuss the reason why. And usually it's perfectly fine.
"It's easy to discuss in a very pragmatic approach, like: 'This is why. The reason we cannot go forward with different environments, with the production, is because this, this, they are not ready. And if we do so, we'll pay later - because everything they do will not work, or we'll have to go as [with Cyberpunk], catch up during production.' We do not want to do that again."
He suggested "a good example would be like, console [builds] not working. So having the console working super late, it's unacceptable anymore, and it's part of our process. So we do the reviews on console, so we know exactly where we are on all the platforms - the lower platforms we have - rather than, you know, 'Oh PC is fine, so we can go forward'. So we changed this approach to have a broader visibility on the other platforms we want to have."
"Most likely, we will not have a PC only launch"
When it comes to the next major game's launch, Tremblay also gave a rough steer on what the studio's strategy will be when it comes to those additional platforms. One of Cyberpunk 2077's major development struggles came from issues with last-gen consoles and their hard drive read-write speeds, for instance. I asked whether CDPR would consider an approach similar to that Larian took with Baldur's Gate 3, where it launched on PC first, and then consoles a few months later to buy a little more time. (Ironically, Larian boss Swen Vincke later explained this was partially also to avoid some major releases at the time, including Cyberpunk 2077's Phantom Liberty expansion.)
"It's something we discussed, actually," Tremblay told me, "but we're not 100 percent sure. Maybe - so it's hard to say if we want to go this way at this stage. I would assume that as long as I understand, like, the strategy we want to do - I'm just trying not to say too much, bear with me - most likely, the launch, we will not have a PC-only launch, for example. I don't think this is a strategy we want to adopt right now. That being said, will it mean that we will have - let's say there were 20 platforms available - we'll have 20 at launch? Maybe not. But at least PC only, and then scattered, it's not necessarily the approach we want to go forward, for sure."
Meanwhile, Paweł Sasko, who was lead quest designer on The Witcher 3 and is now the associate game director on Project Orion, Cyberpunk 2077's sequel, also shared a bit of extra context on CD Projekt's development team sizes and timeframes, firstly in the case of Project Hadar, the studio's new IP.
[Hadar]'s in its early stages, right?
I spoke with Sasko twice for Eurogamer's big Cyberpunk 2077 article, the first time as far back as spring of 2023, the second in the spring again of this year. Back in 2023, Sasko described the Hadar team as "incredibly tiny - well, not incredibly tiny; it's a small group of people that are putting like, a small concept board into that. And it will become a game, because you need to slowly get there, but it's in its early stages, right? It's not a full blown production where we have like a 600-people team on it."
Come spring 2024, there was a smidge more detail, noting Hadar was still "in early stages, so in that ideation phase, or however you call it - it requires constant communication, bouncing ideas."
The Boston-based team that's focused on the Cyberpunk sequel meanwhile is what he described as "a pretty tight core team composed out of the directors that have been shipping Cyberpunk and Phantom Liberty." Part of the studio's decision to set up there - alongside proximity to some top universities and tech companies for recruitment, and the ability to become "one of the biggest players" on the US' east coast - was to straddle the timezones of Vancouver, which houses a CDPR studio composed of "mostly engineers," and Warsaw, which is now "mostly art". The Cyberpunk sequel team remains "fairly small for triple-A," he said at the time.
Thirdly and finally with Sasko, there was a little discussion of The Witcher studio. "That team is probably the biggest team, the team that took all The Witcher, Cyberpunk devs on board, and is working on their game. I don't know, certainly, the exact number," he added. "It's like a couple hundred there," as of around late March this year.
Notably, that's a fair bit earlier than this recent announcement that full-scale production had begun. It's tempting to infer that fairly significant pre-production work has been going on for a little while before that public confirmation, but then it's worth remembering Tremblay's explanation of how that new process works: it may be that this time has been spent getting the technological side of things ready before full development started in earnest. Likewise, a few hundred isn't as big as it sounds for CDPR these days: while around 300 people worked on The Witcher 3, 350 developers worked on Phantom Liberty alone. The studio now employs something in the region of 1400 people worldwide.
"There's no magic, only hard work and making sure you stay on the right course"
Sticking with inferences, there's also one more bit of information we might be able to glean, this time from a conversation with joint CEO Michał Nowakowski. Like a lot of developers I spoke to from CD Projekt Red, he described his mindset on the studio's future releases.
"Optimistic. I feel optimistic because I'm, in general, an optimistic person," he said. "But on the on the more realistic side - I think that's actually part of why I'm optimistic - is I think we are aware, as an organisation, that, let's say what was one [example] with Phantom Liberty coming back to 95 percent, currently 94 percent [positive Steam reviews], nothing is granted forever. And I think it's something that it's important for us to understand. And I think also for all the people in the company: there's no magic. There's only hard work and making sure you stay on the right course."
With future games, he added, "with the Witcher or you know, with Cyberpunk in the future, it's only down to us. Everything can fail, but it's our role to make sure it doesn't. And remembering and understanding that I think is the biggest reason for me to be optimistic, because in the company everyone's constantly worried - but worried in a good way, if you know what I mean. Worried that we are going to do it in a good enough way? 'Are we, for sure, not letting something slip through our fingers?' And if you're that kind of worried, I think that gives us better chances of being successful in the future than thinking we can't fail because we're too big, or too whatever - you know, noses up our somethings."
One of the specific lessons the studio took from Cyberpunk, which has particular relevance here, to bring things back to the next announcement, was to be "smarter in how we want to announce and kick off marketing campaigns," Nowakowski told me.
"To be honest, when we were kicking off the marketing campaign officially with pre-orders, which was the Keanu on stage [moment, at E3 in June 2019], the plan was actually to launch roughly one year later. It just didn't really work. So we didn't really plan for like a two-year campaign, and I still think that one year would really be fine [in terms of] time for a promotional campaign of that game."
"For a new game, expect a slightly longer - but not two-year - campaign"
The studio "learned a lot of good practices from that experience: so announce the date when you're like really, really sure of it. And now I think we have much better tools to be sure of that date, which we - on a smaller scale - proved to ourselves with Phantom Liberty." While that campaign lasted around six months, given it was only an expansion, Nowakowski added that "for a new game, we would still expect a slightly longer - but not two-year - lasting campaign."
"Having said that," he added, "I want to stress: it doesn't mean we would not tease or drop some cool assets before [a full reveal]. Because the marketing campaign, slightly earlier before the launch of the game, that's different than the actual, say, 'mass attack'. Mass attack is when you announce the date, you start collecting the pre-orders and it really is that race from that point, that moment, to the moment you launch the game."
For example, he continued, "If I ask you, what do you know about The Witcher 4? The answer is not much, probably - yes, there's theories and so on. But there's nothing really specific. So we want to drop the crumbs here and there so that people - and the media as well - can, you know, pick up on it and try to figure out what it is we're trying to say this time. So that [is something] we can start doing a little bit earlier." That would involve aspects such as "having some video assets, or something to whet the appetite of the fans - not even to create the hype so much," he said, as to simply get the early conversations flowing.
Another note on platforms, too: another thing the studio took from the mistakes of Cyberpunk 2077 was what Nowakowski described as "the focus on the target platform" during marketing from the off. "I think what put us into a lot of trouble is how late we started with PS4 and the Xbox [One] back in the day, and we don't ever want to be in that same spot again."
What can we take from that? Well, expect at least a year - but most likely less than two years - between The Witcher 4's proper reveal and its actual launch. Expect, possibly, some light teases or short trailers a little earlier than that. We probably shouldn't expect a firm commitment to platforms early on, nor a PC-only release - although plans can of course always change - and hopefully, if Phantom Liberty is a guide and if those lessons truly have been learned, expect CD Projekt Red to launch The Witcher 4 a lot closer to the first release date it gives.
Purely to speculate: the Game Awards, known for their hefty emphasis on major reveal trailers and teases, are just around the corner, too, and CD Projekt has put some weight behind Keighley-powered events such as Summer Games Fest before, most notably with Phantom Liberty. We might see the first of those breadcrumbs quite soon.