← Home

AA War Lane


Itch.io

Gameplay Synopsis

The battlefield awaits! Created in Unity, AA War Lane is a single-player auto-battler featuring…

  • Army Formations — Inspect the enemy army, plan your unit composition to counter it and organize your army into an effective fighting formation.

  • Unique Challenges — Beat increasingly difficult levels that offer new hurdles to overcome.

  • Unit Variety — Unlock new units as you beat more and more levels; learn the advantages and drawbacks they each have to offer and how to best utilize them in your formations.

  • Tactical Gameplay — Learn which units are best at countering which other units through a steady process of trial and error.

  • Unit Upgrades — Win upgrade points after beating each level and invest them into specific unit attributes to strengthen your army.

Personal Contributions

My contributing roles were production, design and engineering. I proposed the game’s concept and consulted with our unit designer to make subtle refinements. I was responsible for all engineering-related development necessary to power the experience (player input, platform compatibility, GUI, etc.).

  • Game Concept — Initiating the earliest stage of development, I proposed the concept for a stylized, 3D tactical auto-battler playable on desktop and mobile. I came prepared with a setting, gameplay loop and a list of ideas for units with descriptions of their roles.

  • NPC Combat System — I designed and programmed every feature of NPC behavior from the team functionality, targeting algorithm, attack mechanics, healing mechanics, all the way to the functionality for armies maintaining a tactical formation while moving.

  • Placement System — I handled the implementation of a grid-based unit placement system that is compatible with both desktop and mobile control schemes.

  • Camera Transition System — I developed an animated camera system that smoothly transitions between crucial perspectives to give the player key views of the battleground.

  • Upgrade System — I designed and programmed an upgrade system that complements the gameplay loop by adding an extra layer of planning and strategy to every level.

  • In-Game Level — The layout I created for the in-game level serves as a practical setting for every battle to unfold. It has a background landscape featuring mountain ridges I sculpted and models I arranged.

*Art assisted by AI.

Game Concept

I derived the concept from the premises of different successful contemporary games in the mobile market. Clash of Clans was my primary reference; a game where units are dropped into a level and left to automatically attack enemies. Since the traditional real-time army management of desktop strategy games is almost impossible to replicate on a touch screen, these kinds of games turn real-time strategy into a viable genre on mobile by removing the responsibility of micro-level decision making from players. Recognizing this, I thought up my own take on an auto-battler ported into a feasibly scoped, single-player concept.

Design Goals

  • Create a gameplay concept for an auto-battler.

  • Create a distinct and consistent theme and style for the game.

  • Keep gameplay and controls simple enough for cross-compatibility on desktop and mobile.

Design Process

My gameplay concept is based around the idea of pre-placing an entire army formation and sending it into battle against an enemy army formation. What spoke to me about this specific gameplay format was its broad potential for tactical scenarios and distinction from other auto-battler games in the contemporary mobile gaming market.

I wanted to give AA War Lane an identity as a tactics-focused strategy game set in an exotic place far away from conventional urban society. For this, I settled upon adopting a vague military aesthetic and a vast tropical desert atmosphere for the setting. To give it its own flavor and make room for technological progression, I decided there should be a range of different types and qualities of weapons, even if they wouldn’t necessarily make sense in a real-world military setting. Some of the game’s initial features were downscoped out of the final result in order to keep a realistic development workload and sharpen the gameplay loop. However, the overarching tactical vision remained intact in the final product.

Concept Board

I narrowed down the initial list of units to six core members in the final version of the concept.

  • Grunt: The cheapest, weakest infantry unit and the only one with a melee attack.

  • Minute-Man: The low-cost, standard infantry unit shooting with bolt-action firing speed.

  • Medic: The low-cost, healing support unit and the only one with no attack.

  • Humvee: The high-cost, lightweight vehicle unit that deals rapid-fire, medium damage; it is highly effective against infantry.

  • Rocketeer: the medium-cost, slow-firing, anti-armor specialist infantry unit; it is highly effective against vehicles.

  • Tank: The most expensive, strongest heavyweight vehicle unit.

Unit variety from left to right.

Key Takeaways

  • Market Research — Even when not monetizing (like in the case of this game), it is still important to study successful examples from relevant gaming markets to learn actionable design lessons about how their business models succeeded and what they innovated that worked.

  • Feature-First Scoping — Since gameplay features are the bulk of the work on the engineering side of game development, a feature-first mindset should be adopted to avoid overscoping and achieve a more accurate judgement of the development workload and timeline.

  • Layering Balance Dynamics — When crafting a list of unit/class types, it helps to structure them into layers that each have their own compatible dynamics. If downscoping becomes necessary, layers of units/classes can be peeled off without ruining the overall balance.

NPC Combat System

The NPC combat system is the driving force behind the auto-battler gameplay. It is the brain of every unit, enabling them to fight out battles and fulfill their role in the army. A lot of thought was invested into making it function at a production-ready level. The result was a highly optimized system that works exactly as intended.

Design Goals

  • Create a closed-loop logic system for NPC units to act independently from the player.

  • NPCs acquire targets that make the most sense for their physical position and role.

  • Armies maintain a pre-placed tactical formation while moving.

  • Ability for developers to maintain different properties and visual/audio effects related to combat.

Design Process

I knew that if every NPC in the game needed to act autonomously, then that could only be achieved through a closed decision-making loop that runs for the duration of a unit’s lifetime. To accomplish this, I created my own code structure for the combat system which allows units to ignore friendlies, pick sensible targets and maintain the player’s tactical formation while moving to engage the enemy.

As the game was always intended to be played on desktop and mobile, I made it a particular point of focus to optimize the target finding algorithm as much as possible. To achieve this, I considered the mathematical dynamics involved in comparing distances. I learned that while finding a discrete distance obviously requires getting the magnitude of the vector between two points, only square magnitude is necessary to numerically sort a collection of distances from least to greatest. This is because the square and square root functions both possess a monotonic property in the positive domain, as in, they are both monotonically increasing at all values greater than zero (0 < x² < y² ⇒ 0 < √x < √y). Given this relationship, the more expensive step of calculating the square root for magnitude can be completely skipped without changing the algorithm’s result for the closest target.

Algorithm for finding closest target using square magnitude.

To make target finding even more efficient, I took the added step of partitioning all active units into different object lists based on the team they belong to. The thought process behind doing this was to eliminate all redundant comparisons performed on every frame. Since this ensured only valid targets were being searched through for every active unit, the quantity of distance comparisons per frame was cut down to the absolute bare minimum. By combining this organization structure with the square magnitude optimization, I was able to create a sharply performant system for every unit to identify their own ideal target at all times. This organization structure also doubled at achieving my desired functionality for ignoring friendly units by strictly limiting the set of valid targets for every unit to the opposing team.

Unit Organization System

With the functionality for identifying targets tested and verified to work performantly, all that was left to do was integrate it into an encompassing algorithm for movement and attacking. I began by writing out all the conditional statements necessary for an NPC to decide during any frame: whether it should march in formation, disengage from formation, move toward a target or attack that target if within range. This is the step that brought it all together and completed the auto-battler gameplay. Every active unit was now capable of moving and fighting properly until death or the conclusion of the battle.

Overarching auto-battling structure.

After getting the auto-battling necessities out of the way, I decided the next step was to ensure the continued maintainability of unit types. I began by exposing their combat-related properties to the inspector so they could be readily modified for balancing purposes. Alongside this choice, I created exposed fields for assigning the specific visual and audio effects of combat to every unit.

Exposed fields in the inspector.

Key Takeaways

  • Math Comprehension — Having a good grasp on mathematics relevant to virtual 3D environments (vectors, quaternions, etc.) can allow for optimizations to be found by understanding redundant calculations.

  • Prudent Implementation — Choosing the right way to implement a mechanic within a system can simplify the overall logical structure and reduce development workload. In this case, my unit organization system assisted with performance and satisfied the team differentiating mechanic at the same time.

Placement System

The placement system represents the medium through which players express their own tactical styles and philosophies while playing the game. Its implementation was a noteworthy challenge that taught me many lessons in player input, grid logic, conflict handling and overall game data management. Although this could be said about every feature in every game I develop, I especially feel that I continue to use what I learned in the creation of this system to this day.

Design Goals

  • Create a placement system capable of hosting units of varying grid dimensions.

  • Limit unit placement to the grid.

  • Disallow placement conflicts on the grid.

  • Handle unit costs and disallow exceeding the budget for a level.

Design Process

The obvious first step was to decide what kind of grid would work best for gameplay. I concluded that since the game’s focus is two tactical formations clashing into one another head-on, the placement grid should be rectangular in shape with its length along the horizontal axis relative to the opposing army. I settled on an 18x10 grid since it fit the shape profile I needed and provided plenty of room to devise a formation.

Naturally, I had the grid rely on raycasts from the player’s on-screen cursor/tap position to determine the absolute spot on the placement grid being selected. Doing it this way allowed for the absolute position from the raycast to then be converted into a specific cell position within the grid. For all units, the bottom right cell is treated as the anchor point for placement. I made this choice so that when playing on mobile, the player’s finger does not obscure the unit from view while placing.

Placement Grid

I made every placed unit hold a set of essential data associated with it. If the placement grid itself is the foundation, then this data serves as the central pillar which supports all other placement mechanics. A List of vectors is used to contain all occupied grid positions for identifying and preventing placement conflicts. An ID and cost value are stored to grab the correct unit prefab and subtract the correct amount from army points. Lastly, an index is recorded to reliably remove the associated unit from the List dedicated to storing all placed units.

Essential Placement Data

For the placement mechanics themselves, I set up a series of critical checks which must be valid before a placement can execute. These checks ensure the cursor/tap selection is on the grid, no placement conflicts exist, the placement is within the grid’s boundaries and the unit can be afforded. A unit will only be placed onto the grid when all conditions are validated. Upon placement, the unit prefab is offset away from the bottom right corner and centered onto its proper position on the grid. Once situated, the unit is accounted for in both the List of all placed units and the unit organization system, and the cost of the unit is subtracted from the current amount of army points.

Unit Placement

Key Takeaways

  • Single Responsibility Principle — SOLID design principles are just as important in game development as they are in software development. In this case, ensuring classes focused on a single responsibility greatly helped the readability and maintainability of a complex system.

  • Cross-Platform Interaction — A well written input system can allow for differing interaction mechanics on different platforms without any need for changes to the underlying software infrastructure of a gameplay feature.

Camera Transition System

The camera transition system serves as the player’s eyes for every level of the game. My goal was to create a system that did more than just the bare minimum of showing players what they need to see by snapping between different perspectives. Instead, what I designed is a vehicle for delivering a cohesive sense of spatial presence in the level and a feeling of dramatic impact.

Design Goals

  • Create a camera system that switches between key perspectives.

  • Smooth transitions that allow players to see and admire background details.

Design Process

I knew the game would be missing out on a polished feel if camera perspectives transitioned instantly with no real-time movement. Compared to other quality-related additions I could have made to the game, I identified this one as contributing a significant level of polish with a relatively low effort cost and prioritized it.

My idea for the camera transition system was for it to act as an API dedicated to moving the player’s camera view. Upon inputting a value for position and rotation, the camera view script handles the transition on its own. By doing this, I allowed for all GUI button-connected methods to avoid direct manipulation of the camera and safely orchestrate transitions with the same consistent movement functionality.

A camera point aimed where enemy army spawns.

After considering different options, I decided interpolation logic was the best way to implement camera movement. Thanks to the predictable and reliable behavior of interpolation, it leaves no room to worry about the camera escaping its logic constraints due to a floating point error and drifting away from the battlegrounds. I included the option to use either linear interpolation or spherical linear interpolation to provide more ways to manipulate camera motion.

Overarching transition logic.

To streamline the development and testing of camera transitions, I made the system work using dedicated camera points set up within the level. References to these points are used for all position and rotation values. By using this technique, I created a quick and interactive way to manipulate camera transitions and experiment with various framings.

Camera Transitioning

Key Takeaways

  • Interactive Development — Designing a system to be driven by in-engine interactivity can be extremely valuable for efficiently creating polished work.

Upgrade System

The upgrade system is what ties the whole game together. Upgrades serve as the reward players receive for their conquests and are what imbues them with a sense of progression throughout the levels. Without the upgrade system, levels would lose an important dimension of strategy and lack the feeling of being stacked on top of one another.

Design Goals

  • Create an upgrade system based around unlocking points after beating a level.

  • Have it impact a variety of meaningful stats related to unit strength/performance.

  • Ability to transfer points between stats for unfettered strategy and decision-making.

Design Process

My intent for the upgrade system was to have it be supplementary for earlier levels as the player was still unlocking new units, and then consequential for later levels when winning becomes hard without the right kind of upgrade balance. I decided three upgrades per level beaten was a balanced rate to keep players competitive against the increasingly difficult enemy armies.

To create straightforward dynamics, I divided upgrade points between health, armor, attack damage and attack speed. Upgrades apply to all units identically. (Attack-related upgrades for medic translate into heal amount and heal speed.)

  • Health: Increases health by one per point. More impactful for infantry than vehicles since they have significantly lower default health.

  • Armor: Decreases damage of incoming attacks by one per point (guaranteed minimum of one damage per attack). More effective against grunts, minute-men and humvees since their default damage is low.

  • Attack Damage: Increases attack damage by one per point. Impactful to grunts, minute-men and humvees since their default attack damage is low and their default attack speed is quick compared to rocketeers and tanks.

  • Attack Speed: Increases attack speed by one percent per point. Notably impactful to all units if set high enough.

Upgrades Menu

Key Takeaways

  • Stat Dynamics — Understanding why units/classes of various roles are impacted differently by the same stat changes is important for making informed balance decisions. It allows one to understand key balancing concepts such as overkill and hits-to-kill.

In-Game Level

The in-game level is the facade that pulls players into the experience and invites them to suspend their belief. It is what gives the game an atmosphere, elevating it beyond the feeling of an empty prototype and into having the impression of taking place in a world of its own.

Design Goals

  • Create the in-game level in a way that adheres to the creative decisions of my concept plan.

  • Designate a clearly defined, unhindered central space for armies to battle in.

  • Limit details to places the player can see.

  • Maintain visual balance and a believable composition in the foreground and background.

Design Process

My thought process behind composing the game’s map layout was to give it a stylized setting that provided a visual identity and told an open-ended story at the same time. To keep it visually pleasing but performant on mobile, I needed to strike a balance between detail and minimalism.

First of all, I began development of the level with the battleground. I needed an area to test gameplay functionality during development, so this was the natural first step. It didn’t take me much time to conclude that a small plateau was the perfect shape for two armies to clash on top of. Without ambiguity, it clearly communicates where gameplay will and will not be taking place.

The Battleground

For the background, I began with the mountain ridge. I reasoned that a hard backdrop would help define the space in which everything else will reside. Once the peaks, dips and ridges looked believable, I got to work filling in the background with suitable props to communicate the tropical setting. My model variety consisted of only two buildings, two trees and an oasis. But by sizing, rotating and positioning them in different ways, I pulled off a much more varied appearance that makes this fact hard to notice. As the cherry on top, I gave all the buildings and trees various tilts to imbue the scene a with subtle personality.

Foreground and Background

Key Takeaways

  • Intentional Design — Knowing what players need from your level before getting creative is always the smart course of action. Make a blockout of the playable space, then fill it with interesting ideas.

  • Clustering — Randomness is not uniform. Knowing that natural and spontaneous things tend to cluster is crucial for making many types of scenes appear believable.

Up Next


Jobert: An Explosive Adventure

Side-Scrolling Platformer

Platforming Mechanics Engineer

Handled programming for movement and platforming mechanics. Participated in a team environment to help create the vision.

View Details

*Art not representative of my work.