Children of Morta is a narrative-driven, roguelike action game with pixel art visuals about a family of guardians. This is our sixth released game at Dead Mage. This project initially started with the goal of making a little game to be developed in six months. We selected pixel art for the visualization and roguelike for the genre to manage the scope and meet our milestones. The more we worked on the project, the more we realized there is potential in it. We decided to try a Kickstarter campaign after nine months of development. The campaign launched in February 2015, and its success raised our expectations from the game. Two years after the project's start, we had a publisher, 11 bit studios, onboard that was as passionate about the game as we were. Once more, the expectations were raised. Eventually, after five years, we launched the game on PC, and soon after that, on PS4, Xbox One, and Nintendo Switch. 

The game has received praise from the community and the media. Looking back at it today, we are happy that we had the opportunity to work on the game for a relatively long time.

A year has passed since the launch of the game. This article will describe the main lessons we learned from the game's development process and a year post-launch. Reading game post-mortems has always been inspirational and educational for our team. We hope sharing these points will be helpful for other developers.

What Went Right

1 - A Cohesive and Supportive Team

Our project started with an initial team of five developers, and mid-way through the development, the core team size increased to 18. 

All creative projects depend heavily on the members' quality and how they can interact and work as a team. Our team was dedicated and passionate, but the most important feature was that they all found great purpose in making this game. Children of Morta was the first large game project for many of the members. We had many ups and downs in the five-year development process; however, the strong sense of meaning we found in the project enabled our team to bear the hard days and push the project forward every day. Throughout this project, we learned that we, ourselves, needed to be a family to make Children of Morta. 

If I wanted to describe the team culture in one word, that would be "Supportive." Supporting team members is vital in two ways. The first, which is the apparent functionality, is getting help from others to finish a task, technical or non-technical support. The second, which is the more critical functionality, is the meaningful act of helping others. When team members got stuck in their work and didn’t feel the progress, assisting others increased the morale and regained the team’s sense of purpose.

Support is the central theme in the story of Children of Morta. We had to learn its value along with the guardian family in the game, the Bergsons.

2 - Protecting the Main Game Pillars

Our initial idea was to make a game about a family with a complete storyline in the roguelike genre. The challenge was adding a story to a roguelike. These two elements proved to be interesting, based on early feedback. We knew we had to keep the two main pillars, an emotional game about a family and a story-based roguelike. We knew that every other aspect of the game needed to support these two pillars. The same two features were praised after the launch of the game.

We spent a significant amount of time, making sure a story could be added to the roguelike game structure. Furthermore, our main concern was to avoid ludo-narrative dissonance, and we paid careful attention to blend the emotional narrative elements with the gameplay systems properly. Our main goal was to create a cohesive experience where the narrative and gameplay supported each other. Hundreds of meetings and discussions were needed, and many ideas and solutions were thrown away in the process, but we were able to reach the destination we were aiming for.

Specifying the game's vision initially, getting feedback on it, and then protecting that vision until the end worked really well in our experience.

If you would like to read more about the design challenges we faced, check this detailed article written by our game and narrative designer, Hamid R. Saeedy:

The Design Challenges of Children of Morta

The Bergson Family

3 - Quality of the Pixel Art and Animations

Visuals are always the first requirement for being noticed. This was the case for Children of Morta from the initial days until the launch of the game. Our visuals benefited from pixel art with high details and various color palettes for the backgrounds. On top of that, advanced lighting techniques, post-process effects, and highly detailed and smooth animations made the game look different from other pixel art titles. These aesthetic decisions and our artists' hard work and dedication to creating the massive content helped the game get noticed from a visual standpoint.

Due to 2D pixel art, a short art pipeline, where our concept artist’s creations would be precisely what we saw in the game, provided excellent control over the game's final look and feel. 

Family house and Mount Morta on the horizon.

4 - Publisher Relationships 

Like every independent team, we expected to find a publisher that would help us make a better game, promote it, and handle the sales processes. We got all that we expected in our partnership with 11 bit studios that happened halfway through our development process. Getting regular feedback about the game while having total freedom to decide on creative matters and a standard quality assurance process were essential factors that helped us improve the game's quality. When developers interact with their game every day, it is easy to become blind to specific issues. External feedback, especially from experienced developers that also happen to be publishers, is immensely helpful. Although many times, the feedback we received after focus tests delayed the launch a few times. Those decisions were challenging for a team that had already worked on the project for a very long time. However, knowing that the decisions to delay were opportunities for us to make a much better game, the team could endure. Today, we are glad we did not rush into releasing the game. The delays have paid off.

Managing social media, community management, media relations, and attendance in the main game events and shows were examples of marketing activities that were important for Children of Morta. Having a partner to help in those areas while the team was entirely focused on making the game was a blessing.

Finally, selling the game on multiple platforms and in various regions of the world while running all kinds of promotions and providing support for the community is the type of help that significantly helps independent teams. It all went really well in our experiences post-launching Children of Morta.

Looking back at this collaboration, I believe the publisher relationship’s essential parameters are trust, empathy, honesty, and excellent communication. We were lucky enough to have these in our relationship.

5 - Internal Tools and Technology Development

During development, we needed to decide about making customized tools at various stages. Many of these tools proved to be very valuable. One of the most important of them was a Hierarchical Finite State Machine tool that we created inside the Unity Engine editor. This tool would let our game designers define the behaviors they expected using a graphical user interface, and it would generate code in the background for efficient execution. These state machines were utilized for NPC AI and PC behaviors, user interface flows, player progress systems, and narrative management. It is hard to imagine how we could have made the game without these tools.

In-house Hierarchical Finite State Machine Tool

Resource management was another area that required our own tools. Levels and every object in the game are procedurally generated, demanding complete control over all assets. The tools we created for resource management helped us make our own customized asset lifecycle management on top of what Unity provides by default.

Procedural Level Generation

The final area which went right regarding tools and technology development was a customized rendering pipeline that helped us integrate some unique lighting techniques with a standard 2D pixel art pipeline. 

If you would like to read more about the details of these tools and technologies, check out this interview on Unity blogs with Reza Hooshangi, our lead programmer. 

Children of Morta, Bringing Order to the Chaos of Procedural Generation
 

What Went Wrong

1 - User Experience Design

We did not have a person dedicated to working on the UI/UX during the first half of our project. The importance of user experience (UX) tasks was something that we were unconsciously ignoring. Initially, the game was relatively simple, and those who tried the game were all developers and had the creators next to them to describe what is what. The game got more complicated when new characters were added. Skill trees were more extensive, many items were added, and various upgrade systems were implemented. Anyone outside the development team had a hard time understanding how those systems worked. The game's feedback to the player was not very clear. 

This was something that we realized in the last quarter of development. After every focus test, the pressing issues were mostly related to UX. The gamers did not understand what the game was trying to communicate with them. The requirement to support three consoles other than PC made this an even more significant challenge. Fixing the user experience issues required many iterations, and it demanded many hours from our core team members. At first, the design and implementation of UI and UX related tasks were shared between the existing team members. When things didn't go well, we had to add new team members for UI design and implementation and had one of our designers focused and dedicated to UX. Overall the game user interface was recreated at least three times from scratch. 

Character selection UI was recreated many times.

2 - No Clear Lines Between Pre-production and Production

When we look back at the game's development process today, it is impossible to point out when pre-production ended, and production started. This project continued growing with various feedback stages, and new features and ideas were being added to the game at every step. There are many drawbacks when this happens, a lot of work will be thrown away, many assets will be created but will not be used because, by the time they are ready, the main requirement for them has changed and also there will be a lot of stress on the psyche of the team. Our development process was more like a colossal pre-production phase.

When the development process is more chaotic than orderly, as much as there are opportunities for finding new and exciting ideas, there is a chance to forget about simple game requirements. One of these areas for us was the game story. We always knew that our story was supposed to be about a family. We were so drowned in the daily design, art, and technical challenges that we realized the game story was not really about a family towards the end of the project. This required us to go back to the drawing boards for the narrative and tweak and change it, which led to a lot of work for almost everybody on the team.

Almost all of the story was finalized after our nominal pre-production phase.

Another major downside for a pre-production style development is the inability to plan in detail, estimate the tasks, and have fixed deadlines. When the requirements are fluid, new tasks pop up, and following a rigid project plan becomes a significant challenge. Because our team grew from 5 to 18, all of the above became exponentially more challenging. Besides the human resource challenges, our tech pipeline required significant efforts to adjust to the new team size. Areas included version control and tracking tools, asset pipeline, and the build system.

New ideas have many potentials for indie games; however, there is always the danger that those ideas might not resonate with the audience. We were lucky enough to have a continuous feedback process throughout development. One of the first activities we incorporated into our process was weekly game tests by an external member. As early as the second week after starting making the game, when we only had a few capsules that would move on a plane, we held weekly play-test sessions, inviting other members of our studio and friends to try the latest state of the game. The second phase of feedback came when we shared the early builds with our Kickstarter backers. They would play the game and give us feedback on various aspects, which helped us better understand the game. The third wave of feedback was after we signed with our publisher. We got regular feedback every week. In addition to that, we received significantly helpful feedback after the focus testing sessions. Gamers of all sorts were invited to play the game. We were trying various ideas in the game but what prevented us from getting lost were these continuous feedback loops.

As stressful as a pre-production style, chaotic development is for any team, indie developers have a chance to compete in a challenging industry by finding creative ideas during the pre-production phase. A more prolonged pre-production phase would mean more opportunities to find those unique selling points of the game. Like everything else, this opportunity needs to be considered, along with all the other drawbacks of not having a very organized production phase where everything can be planned very well. So, at the end of the day, I have mixed feelings about this type of development. As much as I remember the days that I wished for more order in our daily development, I know very well what gems we found from inside the chaos.

Character workshop was a progression system that was added after our nominal pre-production.

3 - Late Addition of Online Multiplayer

We had very early plans to add multiplayer to the game, however, local multiplayer. When we were in the middle of the development process, we decided to add online multiplayer to the game. This was a no-brainer for a top-down hack and slash game with multiple characters. Still, we immensely underestimated the technical challenges that it would add to the project. Our engineer side knew that this would be quite challenging, but our gamer side could not resist forgetting the idea. We felt brave and foolish enough to decide to challenge ourselves and go for online multiplayer. 

This feature introduced a lot of work for the team. The code base’s architecture had to change, specific standards for online multiplayer on consoles needed to be implemented, and testing proved much more challenging. A few months before release, we had to delay it until after the initial launch. It did not even make it to the first year after the release. We hope to have it available in the coming months.

Online Co-op Multiplayer

4 - Management of the Localization Process 

Another area that we had heavily underestimated was the efforts needed to handle the localization requirements. Children of Morta has close to 60 thousand words that needed to be translated into eleven languages. This process needed to start months before the actual release of the game. When we sent out the first batch for translation, we kept getting feedback regarding the game, especially the narrative. As a result of these feedback sessions, we felt the need to change many words in the game. This would require the submission of new batches to localization teams. We would get back the localized content at different times. Soon, managing the content’s status and all the languages became a nightmare for a team already under a lot of stress.

5 - Underestimating 2D Pixel Art and Animation Efforts

We selected pixel art for the look and feel of the game because we believed it would be a better decision for a small team and require less work. A few years into the project, we fully experienced all the specific challenges, partly because of our own ambitious ideas. We had selected a relatively high-resolution pixel art look with high keyframe animations to make every motion smooth. Every frame of every character needed to be drawn by hand, frame by frame. Considering that we had selected 6 initial playable characters and 3 other family members available in many cutscenes, a lot of work was imposed on our two-person art team. 

Another area that we had underestimated was the energies needed to animate the boss characters. They were giant characters, and the more the pixels, the more effort was required to animate.

2D pixel art limited the visual variety for game items and character abilities. A new weapon for each character would require a new set of manually hand-drawn frames. This was not feasible.

Sample animation frames for John.

All animation frames for John.
 

Conclusion

We learned a lot and grew as a team during the creation of Children of Morta. We wished we would have done many things differently, but the best that we can hope for is contemplating our failures and remembering what we learned, and utilizing them for the next games that we will work on.

Children of Morta Twitter
Dead Mage Twitter