1. This isn’t a commentary on WuWa’s gameplay, story, or content

This is mainly a collection of thoughts I have about the company Kuro Games and its process for managing the development of the game itself.

There’s been a lot of discussion about the game’s release already along with the salacious internet gossip about conditions at Kuro Games.¹Some of the more worrying commentary concerns things such as the firing of the Lead Engineer for the Game Engine, the company’s treatment of new hires, etc.

For the purpose of this quick note I’m writing, I don’t intend to cover that. Rather, this is mostly a set of observations about what we can imply about how Kuro is trying to manage its latest game and the worrying signs we can see.

2. Kuro is not a company that is doing okay and management is an important factor

The handling of the development and launch of WuWa implies a cascade of failures within the management team for Kuro Games.

It’s impossible to say exactly what the issue is without insider knowledge. But it’s possible to make several educated guesses about what these could be.

2a. Management don’t seem to have a clear design vision or faith in their own vision and their team

After the first Closed Beta Test (CBT), there was a lot of fan criticism of the story for WuWa. This led Kuro Games to make the decision to rewrite the story.

This is a terrible decision. I cannot understate how awful this decision was.

Live Services Games live or die based on their ability to maintain a constant stream of content and updates.

This is true for both:

• PvP type games (e.g. League of Legends, DotA, Valorant, Apex Legends, etc.) which rely heavily on a “create your own fun” approach from PvP and matchmaking; and
• Theme park type games (e.g. WoW, Genshin Impact, Destiny, etc.) where the developers are creating experiences for players. And developers need to keep creating new experiences to prevent the game becoming stale.

The less repeatable content your game has, the greater the emphasis on maintaining a strong pipeline of continuous updates and content.

In order to support the content pipeline, a company needs to be working on this content many months in advance. For an open-world adventure game such as Genshin, Mihoyo’s pipeline is about 12 months ahead of their current patch status.

WuWa CBT1 was held about 12 months before the game was actually released. During this time, by Kuro Games’ own admission, 90% of the story was completely rewritten.

That’s a LOT of work that’s NOT going towards your post-release content pipeline.

And forget just the writing effort involved here. This decision required completely redoing scripting, optimization, voice acting, etc that the team had worked on for nearly 3 years.

I just have to restate this because it absolutely boggles the mind how much work was thrown away. How little faith in your own product must you have to toss 3 yrs of work away?²Or have not identified that this was a fundamental problem with your game even at the alpha stage? A problem of this scale requires more than a basic misunderstanding but implies a larger failure behind the scenes.

Not to mention, if 3 yrs worth of planning led to a bad story… 6 months of panicking is not likely to deliver something higher quality anyway.

2b. Management deliberately chose to make the game development process harder

Kuro Games’ previous game Punishing Grey Raven was made in Unity. WuWa was made in Unreal Engine.³This decision was made well before the Unity royalty scandal in 2023 so that is not a relevant factor here.

This is a big choice to make. Choosing to go with a completely different game engine imposes serious pressure on your company:

• Your current staff may not be familiar with the new engine and technology;
• Your existing production pipeline, DevOps, internal and custom tooling, documentation, etc. are geared to supporting a different technology;
• Maintaining two games running different engines and code bases is massively more complex than if they all shared the same engine and code base.

There may be very good reasons to use a new engine for a new game.  However, when you take on a complex project, it helps to reduce as many sources of risk as possible.

This was Kuro Games’ first massive open world game. So that alone would have been a significant source of risk. Adding technology risk as well just makes your life harder.⁴It is notable that Mihoyo has stuck with Unity for all of its games. In fact, HSR uses a lot of the same underlying code base as Genshin. This was likely done to specifically minimize technical complexity when maintaining multiple games and ensure higher standards of QA. It’s not like Kuro Games couldn’t have chosen to do the same!

WuWa is a game designed to run on mobile devices. It’s not exactly trying to push the limits of next-generation graphics.

And what was the final result? Kuro Games released a game with a laundry list of optimization problems anyway and now have to manage an ever increasing backlog of technical debt across two different systems.

Was this really worth it in the end?

2c. Management created an organization that released a broken product

The games industry has managed to gaslight people into thinking that releasing a broken game is normal and acceptable.

It’s not.

For the game to release in the state it did implies at least one of the following happened:

• Lack of manpower;
• Lack of overall management control and prioritization;
• Haphazard design documentation leading to internal design conflicts and wasted work;
• Lack of infrastructure and tooling to efficiently triage and remediate issues;
• Lack of experience with the Unreal Engine⁵As I mentioned earlier;
• Lack of funding to support business operations
• etc.

We probably won’t know the exact reason unless Jason Schreier gets a scoop on the exact issues that went wrong inside Kuro Games. But these are all issues within the control of management.

These things are the root cause of business failure. And most importantly: These issues aren’t things you can fix overnight.

Business and operational processes, workflows, company culture, etc. can be changed. But it takes time to change these.

Take something really simple such as hiring for example. Generally speaking, the rule of thumb I have consistently heard across multiple companies is that the time-to-productivity⁶How quickly a new employee becomes fully productive is roughly 12 months.

So even if Kuro Games could literally pick the perfect developer off the street tomorrow and get them working, they still would require 12 months before their development team is properly at capacity.

This is before you factor in second-order effects such as the fact hiring more employees distracts your current employees from doing their existing job.

Kuro Games management could have created an organization with internal processes, people, and capabilities that had the ability to release a polished game. They created a structure that wasn’t up to the task.

3. Isn’t this a pretty harsh judgement of management? Sometimes things just go wrong.

That is true. Sometimes things just go wrong and even if you do your absolute best.

However, the scale of the failure here is significant. For WuWa to have had such a troubled launch there can’t have been just one thing that went wrong.

Failure within organizations and systems typically requires the accumulation of multiple errors all at the same time.

⁷This may be intuitive if you have read NTSB Investigation Reports, worked in resilience engineering, etc. It’s often not enough for a single component to have failed for a full system to collapse.

It is the responsibility of management to create robust business and operational processes that can self-correct and monitor these processes to prevent them from degrading.

Furthermore, many of the problems were self-inflicted. It’s absolutely the fault of management for letting self-inflicted problems occur.This goes beyond the choice to use Unreal Engine over Unity and increase complexity of development. Or choosing to rewrite 90% of the story versus smaller scale changes or delaying the release is another such decision.

For example, where was the review process before customer emails were leaked? For this failure to occur, it’s not enough for an intern to press the wrong button. It requires a broader cultural failure to even let an intern be in the position to make this mistake in the first place.

4. Is WuWa doomed to completely fail?

No.

It is absolutely possible for games with poor releases to later improve and succeed. But there’s a reason this is uncommon.

The same organization that already has not proven the capability to launch something successfully, is now going to have split attention trying both to fix bugs and develop new content at the same time.

And as far as we can tell, it’s the same management in charge whose oversight allowed these problems to occur. And what makes many of these problems so disappointing is that were self-inflicted.

It’s not impossible for things to improve. But it’s going to be an uphill battle to do so.

• 1
  Some of the more worrying commentary concerns things such as the firing of the Lead Engineer for the Game Engine, the company’s treatment of new hires, etc.
• 2
  Or have not identified that this was a fundamental problem with your game even at the alpha stage? A problem of this scale requires more than a basic misunderstanding but implies a larger failure behind the scenes.
• 3
  This decision was made well before the Unity royalty scandal in 2023 so that is not a relevant factor here.
• 4
  It is notable that Mihoyo has stuck with Unity for all of its games. In fact, HSR uses a lot of the same underlying code base as Genshin. This was likely done to specifically minimize technical complexity when maintaining multiple games and ensure higher standards of QA. It’s not like Kuro Games couldn’t have chosen to do the same!
• 5
  As I mentioned earlier
• 6
  How quickly a new employee becomes fully productive
• 7
  This may be intuitive if you have read NTSB Investigation Reports, worked in resilience engineering, etc. It’s often not enough for a single component to have failed for a full system to collapse.