top of page

Mount & Blade II: Bannerlord



What is Mount & Blade II: Bannerlord?

Strategy action RPG. Create a character, engage in diplomacy, craft, trade and conquer in this medieval sandbox. Raise armies to lead into battle and command and fight alongside your troops in massive real-time battles using a deep & intuitive skill-based combat system.

I interned as a level designer for 9 months and this was my first job in the industry.

Bannerlord Proprietary Engine Jira Project Planning 9-month Internship

 

Responsibilities


  • Concepted and developed siege and village levels (making specific adjustments for single- and multi-player modes) from start to finish using the proprietary engine and all the accompanying tools that I needed (Terrain Generation WorldMachine & Gaea). I have learned to use these tools for the needs of the task.

  • Developed multiple concept-to-prototype levels for the multiplayer Captain mode. Captain mode is a multiplayer mode where players fight over control points in a 5v5 whilst each player controls their own AI band of combatants.

  • Conducted research to further understand the vision of the IP and core gameplay, I was able to successfully understand the game's needs from the player perspective.

  • Created multiple game design concepts and prototypes on my own initiative which were greenlit by the team and will be built & published in the future content update for the singleplayer gameplay. The game design features I worked on were deepened by expanding reactivity and interaction players can have with existing systems.


My Work


Engine Familiriarization - Adurn Lake Scene

Fishing Hut on a lake near Adurn

At the start of my internship, to familiarize myself with the proprietary engine, I designed multiple scenes with the creative constraint that I must use environmental storytelling to tell a story.

At the start of my internship, I needed to get learn the proprietary engine, so I came up with an exercise to make it more interesting - I designed multiple scenes focused on environmental storytelling. Finding the right assets and tools to convey the story I wanted, while making sure the area is playable helped me remember the toolsets and UI layout by practice.

Environmental Storytelling

I designed a fishing hut that sits upon a lake. It uses Battanian faction assets, and the flora and fauna are based on the region where the faction (which is inspired by the Celtic people and the British Isles) is located. I have experimented with terrain layers and recolouring of objects that do not belong to that region and faction to make them appear to be made of same materials that would be available in the region, thus giving it a more diverse look than existing settlements in the game. This allowed more variety without increased the work load on the content art team.


The narrative of the scene is of a group of fishermen staying at the lake, there are clothes drying which implies that one of them fell into the lake (as there are bottles around the fire as well).

This is also the first time I used the terrain generation tool World Machine and created my own template for future use within TaleWorlds. It greatly helped me with cutting down on time when I was testing and making high-level changes to the terrain instead of hand-sculpting it.


Lessons

Even though the village in the back is not that far away, I gradually scaled down trees and the village itself to appear further, and by doing so, created an illusion of greater distance. This helped me to set-dress out-of-bounds areas in future levels at the cost of less time and assets. A simple trick, but a useful one. The scene took approximately a week, with constant feedback from colleagues, from design team to artists which led to technical and visual improvements. The goal to familiarize myself with the engine and tools was a success as I proceeded to make a few more scenes and then moved on to the Village level.


Singleplayer Level - Vinella Empire Village


Research

I had to research existing levels as the company grew rather quickly and little documentation was in place. Together with another level design intern we created a whiteboard where we shared our findings, analyses and gave feedback to each other on a regular basis.

Our research covered level design; game design of features present in them, to smaller details such as terrain layer painting, the use of parallax and more.

A small chunk of Bandit Hideouts resesarch

References

After understanding the level and game design, as well as other minutiae behind the village levels, I started researching medieval villages focussing on showing the production of the specific village I will design.

I realised that the salt-exporting villages in-game lack representation and I thought it would be visually interesting to have salt fields and I found it to be vastly distinct from existing levels in terms of the gameplay as well.


In-Game Salt Fields

Vinella Village is part of the Empire,a Late Roman Empire-inspired faction, and its exporting product is Salt.

Final Result

I tried to use different terrain generation tools that are standard in the industry, and made this level and the following castle siege using Gaea, a relatively young terrain generation tool.


The development was completed within 2 months, including research and prototyping. The level was thoroughly iterated upon, tested and it was approved and praised as a well-made level that will be released in a future content update.


Singleplayer Level - Sturgian Vladiv Castle


Blockout

Following the success of the Village level, I was given the task of designing a level for another single-player castle. After brief research of existing castle levels, I settled on a snowy Norse-Slavic-inspired Stugia as they have the least levels present in the game, which the lead designer approved.


Castle levels have 3 sub-levels (and each of those 3 have 4 different visuals for different seasons), so this was the longest project I worked on. A few larger and bolder ideas were discarded as I experimented with the level layout and ways to break it up into separate attack routes using caves, but through testing realised that the AI is not advanced enough to handle as many attack/approach routes. This was the longest part of the development, and after many blockout/blockmesh iterations, I was able to proceed to collision and set-dressing.


High-Level Collision & Set-Dressing


In TaleWorlds, level designers do environment art, set-dressing and level design, and I had to quickly adapt to do all these to see my level completed. I started with painting the textures and I regularly texted and spoke with artists who gave me feedback and taught me better ways of detailing the level, as well as where I can cut corners. Level designers gave feedback regularly on smaller segments of the level as they became better formed. I learnt that I can separate the level with terrain height and foliage, as the levels in Bannerlord are quite open and the foliage serves as a great soft edge to block line of sight (albeit, not as good during the in-game winter season).


Gameplay Logic

After receiving green-light on the level of detail in each of the sub-levels and each of the seasons, I moved to adding gameplay logic, such as siege rams and towers path splines, AI defense lines, and more. Through the implementation, I kept notes and shared them with my lead designer on ways to improve the system, as I believe a fresh pair of eyes can contribute better with this feedback than someone who got accustomed to the system.


NavMeshing

NavMeshing is largely done by hand, which requires a substantial amount of time spent on it. It took me two weeks to complete it for each of the sub-levels. I was also testing the level myself daily and sometimes hourly to make sure that there are no errors and to understand the fine inner workings of the system.


Detailing

Lastly, I spent a few more days detailing and implementing features that were in-development while I was working on the earlier stages of development (such as spline decals) and adjusting the level based on feedback from artists and designers alike.


Final Set-Dressing & Seasons

To complete the level, I finished adding season variation to the castle, from the obvious flora changes to smaller changes such as removing outside tables in front of houses during winter season to maintain the immersive feel of a living world.


Conclusion

As I worked on the most complex level in the current game, I was ecstatic throughout the entire development. This siege level will be released in the game with future content update.


It taught me valuable lessons that transcend this game, such as breaking line of sight with subtle terrain variation (if the level allows it), using foliage as a soft fence to segment the open play space, researching and prototyping with proprietary engine and tools I had to adapt to, and I learned how to use Gaea and World Machine.


Design Concepts - UNDER NDA

Besides the two above levels, I have worked on game design tasks that rework existing features that have, as of the time of writing, not yet been released and therefore under the NDA.


I also developed two other level concepts for Captain Mode multiplayer levels which are also under the NDA.


Bonus Content: Crane

I created a custom object on the left using existing meshes based on the reference to the right. I always seek to improve myself as a developer and a person, and I was inspired by all the medieval research and books I was reading to add to the game beyond what my role on the team was.

It was a personal challenge as I want to eventually learn modelling, and this was a fun first step.


Learnings & Takeaways

I have learned more about the game and structure of TaleWorlds, their production pipelines, met a tonne of great developers. I used many proprietary tools and Bannerlord proprietary engine, learnt to use terrain generator programes World Machine and Gaea, developed two complete levels, improved two game design features and developed several other contributions to one of my favourite game series.




Comments


bottom of page