Landon McDowell
We at Believer are building games better. Our industry has been plagued with bad practices for decades, leading to crunch culture and bad player outcomes. We intend to avoid that path.
So many game studios have been reluctant to adopt some of the practices pioneered in other corners of tech that help with these problems. It is understandable why. Large budgets, tight secrecy, and tough schedules drive a high degree of conservatism. There is reluctance to experiment with tools and techniques because it can introduce risk.
It is a false economy.
For us it is far more important to get feedback from players and playtests quickly. It is vital that we can quickly adapt and incorporate those learnings into our work. The process of game development is inherently one of discovery, so we try to capture every opportunity to learn what players love and don’t love about our games.
We are developing a highly iterative approach to game development. This includes continuous deployment so we can get feedback on work in minutes, not weeks or even months that is so often the case in other studios. This approach allows us to make a code change and conduct a playtest in less than 20 minutes. It will be hard, but we intend to keep it that way throughout the game development.
Additionally, automated testing and validation are critical to this goal. We must have confidence that the work we are shipping works as intended and that regressions won’t occur. This practice allows us to ensure higher quality experiences in our testing as well as mitigates the need for big polish cycles that often lead to release crunch.
Our approach to tools has needed to change to accomplish this. We adopted Git and Github, for example, because we want a fast branching and merging strategy. This was a huge decision and one met with no small amount of skepticism as it has and will require a large investment on our part. As a result of that investment, it is easier to run a clean main, ensuring that anyone can see and provide feedback on the current state of the game at any time.
This approach isn’t without drawbacks. We have had to invest heavily in workflows and development tooling. Few people with game industry experience have worked in such an environment, making onboarding slower. We get a lot of blank stares from our peers when we talk about what we’re doing.
It has all been worth it, though. It will also continue to pay off the further along we go.
As much of an advantage as this is, we do not want to go it alone. We believe that both the game industry and the players we serve will benefit from adopting this approach. We believe more teams using and iterating on these tools and this approach creates ever better outcomes for developers and players alike. That is why we are committed to open sourcing tools. It is also why we are opening up about this approach and engaging more openly with fellow developers. For starters, we have created a Discord server where we can discuss all of this, open to all.
We have already opened up the improvements we have made on Git LFS (available on Github). There is much more to come. Join us in our Discord! We believe we are onto something important for the game community, both for developers and players, and are anxious to have others join us on this journey.