Refining size, controls, and monetization for Train Valley 2’s mobile port

Flazm and META Publishing formed a partnership in 2015 to develop Train Valley, a train management puzzle game that launched on Steam that same year, and released on consoles in 2022. The teams followed that up with the sequel, Train Valley 2, which had its full launch on Steam in 2019, on mobile in 2022, and on consoles in 2023.

Bringing the game from PC to mobile brought many obstacles, but they persevered. Over the past three years, they’ve worked on maintaining a strong live ops strategy by testing different tactics and finding what works best for them and their community.

We interviewed Ilya Grabelnikov, chief publishing officer at META Publishing and Evgeny Bushuev, lead developer at GameOn Production, the studio they worked with to port to iOS and Android. They discussed the difficulties they had in reducing build size, adapting controls, building the optimal monetization strategy, and shared the steps they took to overcome them.

Tell us about Train Valley 2?

Ilya Grabelnikov: Train Valley 2 is a train tycoon puzzle game. Players can take their railroad company from the days of the Industrial Revolution into the future, meeting the needs of the valley's cities and industries. They can build railroad networks, manage traffic, upgrade their locomotives, and keep everything on schedule without delays or accidents.

How has developing the game for mobile differed from PC and consoles?

Evgeny Bushuev: Overall, it was more challenging. There was an additional hurdle, especially for Android, since there is a limit on the launch file size. You either connect the resource download after the launch or optimize the game to fit the 160 MB limit. It was hard to optimize the sources without losing the overall game graphics quality or cutting part of the game.

It was also difficult to optimize for touchscreen and small-screen usage. Since the game was created for PC and didn’t have controller support, porting to mobile was a challenge. 

Monetization and content updates were also a hurdle. Mobile users are completely different from those on PC and consoles, and it requires a totally different monetization strategy. We tried all possible monetization types, including Freemium, Free-to-play, and several hybrid strategies, including full Premium. We finally decided on ads and in-app purchases (IAP).

How did the team optimize the build size for Android?

Evgeny: When we got our hands on the project, over 85% of the build size was taken up by 3D models. Compression was completely disabled on .fbx files. After a series of tests, we decided to use medium compression for all game models. This resulted in a significant reduction in build size with an unnoticeable loss in quality.

The next largest category was audio files – they took up about 20% of the total size. We converted all music from .wav to .ogg, which saved an additional 10-20 MB. For sound effects, we simply increased the compression level a little, and this saved a few more megabytes.

In addition, we optimized textures, removed unused assets, made sure that all sprites were included in atlases, and cleaned out the Resources folders from unnecessary content. Despite all these measures, we still couldn’t meet the size limits set by the stores.

The final stage of build size optimization was level icons. In Train Valley 2, each level icon was a separate 3D model in .fbx format. After adding new level packs, their numbers more than doubled, which further affected the final build size. We decided to replace 3D icons with static sprites.

To do this, we wrote a script that automatically places a 3D model of an icon in the scene, renders the image, and saves it in .png format. After replacing all 3D icons with sprites, the build size decreased by several more megabytes – this was enough to meet the requirements of mobile stores.

What steps did the team take to overcome the touchscreen and screen size issues?

Evgeny: We went through several stages to adapt the controls for mobile devices. In the PC version, zooming and camera movement are implemented via the keyboard, and the construction of the railway is done by moving the mouse with the key pressed. On the phone, such controls won’t work for a number of reasons: It requires a lot of buttons, continuous finger movement is not the most accurate solution, and in mobile games, swipes are usually used to move the camera.

We tested different control options: Building a track by moving a finger across the screen, moving the camera with two fingers, zooming via separate buttons, removing rails one section at a time, controlling the camera using separate buttons, and more.

As a result, we came to the following solution: The construction system works on the principle of two taps – the first tap sets the beginning of the track and the second tap signals the end. At the same time, the player can freely move the camera and change the scale during the construction process. This method turned out to be convenient and natural for playing on a touchscreen.

Later, we began experimenting with the rail removal mode. In the first iterations, it worked piecemeal – you could only delete one section at a time, and in some situations, such interaction was very tiring. We tried an alternative: In deletion mode, the camera is blocked to allow you to delete blocks like an eraser, moving your finger across the screen. This option turned out to be much more convenient and remained in the final version.

To adapt the controls for touchscreens, we had to rewrite the existing rail construction logic, but such fundamental mechanisms as finding a path for rails between two cells, calculating the cost, etc., remained untouched.

Describe the game’s initial freemium model and why the team pivoted to an ads-supported model?

Ilya: When we decided to port the game, we saw it only as a Premium title. For quick adoption, we felt that the Freemium option, where players unlock the first 10 levels for free, and can pay for additional content would be best. Unfortunately, mobile users didn’t appreciate this approach and we received negative reviews since the game was located in the free game sections in the stores.

We then decided to move into the free-to-play model, where players either move level by level or use IAP to unlock content and select any level. This approach didn’t work well because the game is pretty hard on later levels, and players were unable to reach five stars on the levels above 17.

Through trial and error, we came to the ad-supported model. Players can continue playing without ads if they reach five stars by themselves. If they have difficulty, they can watch the rewarded ads to move to the next level. We will continue to analyze this strategy and iterate on it to maximize player engagement.

What were your content management and live ops strategies?

Ilya: We have a huge community on PC because that version includes a level editor. Every player can create a new level and share it with their community. We have seen that users are very hungry for more content, and we started producing level pack downloadable content (DLC). These packs include modified and rebalanced versions of the most played user-built levels, which were released with their approval.

Since we’ve seen strong engagement and positive sentiment from the community, we’ll continue offering new levels and mechanics via the DLCs. Our live ops content is centered around that approach – we set up weekly or monthly level-building contests for the creators and the community.

How did the team make use of Unity?

Evgeny: Unity was really helpful during the port from PC to mobile. All the PC changes and updates were merged into the mobile branch, and Unity Build Automation helped accelerate the submission process significantly. Once the code is updated, the builds are almost instantly ready for testing or submissions.

We’re still studying Unity Analytics as we’re not that experienced in mobile publishing in general. But it will definitely give us the required feedback for game performance and in-game bottlenecks.

What has been the biggest lesson the team has learned while operating the game on mobile post-launch?

Ilya: More updates – the more updates you do for the game, the better. Also, every mobile game is a work-in-progress. You constantly need to add, upgrade, optimize, and analyze the progress and performance.

To read more about projects made with Unity, visit Unity’s Resources page.