Coop TD: Verdant Bastion

Inspiration

Verdant Bastion began with a simple memory. Years ago, I was sitting on a couch with a friend, both of us focused on the same television while playing PixelJunk Monsters. One of us was rushing to build towers while the other sprinted around the map collecting gold before the next wave arrived. We barely spoke. Every now and then, one of us would grab the other's arm when things got close. That shared feeling of tension, teamwork, and excitement became the starting point for this project.

Over the years, many mobile tower defense games moved toward solo progression systems. Daily tasks, energy mechanics, and monetization loops became the focus. While those systems can be effective, they often overshadow what originally made the genre special: the satisfaction of creating a defense that succeeds because of your own planning and strategy. We wanted players to feel that their victory came from the choices they made, not from a purchased advantage.

At the same time, we found ourselves drawn back to Metroidvania games. What makes those games memorable is the sense of discovery. A path that seemed inaccessible early on suddenly opens hours later, revealing something meaningful. Progression feels earned because it comes through exploration rather than repetition. We wanted to bring that same feeling into a tower defense experience. Instead of simply unlocking towers after enough matches, players would uncover lore fragments, open new areas of the world, and gradually reveal a constellation of abilities that feels like reading a map of possibilities.

Another idea that kept returning during development was the concept of an opponent that learns. Not in a way that feels invasive or unfair, but in the way a skilled rival adapts over time. We imagined an enemy force that quietly observed how players approached challenges and responded with small adjustments in future sessions. If a particular strategy became dominant, the game would encourage players to rethink their approach. The goal was never to punish success, but to keep the experience fresh and engaging. This idea eventually evolved into the Adaptive Veil, an AI system that learns from player behavior while remaining carefully balanced and fair.

These three ideas became the foundation of Verdant Bastion: the warmth of cooperative play, the excitement of discovery, and the challenge of an opponent that adapts alongside you. Together, they shaped the experience we wanted to create.

What it does

Verdant Bastion is a cooperative tower defense game designed for two to four players on Meta Horizon mobile. Players work together to protect a living sanctuary known as the Verdant Core from waves of crystalline creatures called the Veilstorm. The world is built around a simple visual idea: life is represented by vibrant color, while the Veilstorm brings grey and decay. As players progress, they gradually restore color and vitality to the world around them.

The gameplay revolves around a cycle of preparation, defense, and recovery.

During the preparation phase, players move around the sanctuary collecting Essence dropped by defeated enemies. This resource is used to build, upgrade, and reposition towers before the next attack begins. A wave preview system gives players a clear view of the enemies that are about to arrive, allowing them to discuss strategy and make informed decisions before the action starts.

Once a wave begins, the focus shifts to defense. Towers automatically attack approaching enemies, but players remain actively involved throughout the battle. They can move their characters around the map to strengthen nearby defenses, deploy support abilities, activate traps, and help manage key chokepoints. Cooperation becomes especially important here. One player might focus on slowing enemies while another boosts the damage output of nearby towers, creating moments where teamwork directly influences the outcome of the battle.

After each wave, players enter a short recovery phase. This period provides an opportunity to evaluate what worked, collect rewards, review newly earned Lore Fragments, and prepare for future challenges. The cycle then repeats, with each wave introducing new strategic considerations.

Beyond its core gameplay loop, Verdant Bastion offers several progression and exploration systems designed to keep players engaged over the long term.

Players can choose from four unique characters: Rootcaller, Driftweaver, Thornwright, and Luminary. Each character has access to a distinct Skill Constellation, a progression system presented as a gradually unfolding star map. As players continue their journey, new abilities and specializations become available, encouraging experimentation and different playstyles.

The game world consists of five interconnected Sanctuaries displayed on a hand illustrated world map. Progression is not strictly linear. Certain areas are locked behind Lore Gates that require specific combinations of fragments discovered throughout the world. This structure encourages exploration and gives players reasons to revisit previous locations.

One of the game's most distinctive features is the Adaptive Veil. This system observes player behavior across multiple sessions and makes small adjustments to future enemy waves. If players repeatedly rely on the same strategy, the Veilstorm may adapt by introducing enemy compositions that encourage new approaches. These adjustments are carefully limited based on difficulty level to ensure the game remains fair and enjoyable. Casual players can enjoy a relaxed experience, while experienced players can encounter increasingly dynamic challenges.

To strengthen the social aspect of the game, Verdant Bastion also includes Daily Seeds and Weekly Bloom Events. These activities encourage players to return regularly and share experiences with friends. Rather than focusing on competitive rankings, these events are designed to create a sense of community and shared discovery.

At its core, Verdant Bastion combines strategic tower defense gameplay with exploration, progression, and meaningful cooperation. Every system is designed to support the central fantasy of restoring a living world while working together to overcome an evolving threat.

How we built it

The development of Verdant Bastion was divided into four major phases. Each phase focused on validating a specific part of the experience before moving on to the next. Rather than trying to build every feature at once, we focused on creating small, playable milestones that allowed us to learn from testing and iterate quickly.

Phase 1: Building the Foundation

The first phase was entirely focused on proving that the core gameplay loop was fun. We implemented enemy spawning, pathfinding, basic tower placement, and the health system for the Verdant Core. At this stage there was only a single enemy type, the Drifter, and one basic tower called Thornbriar. There was no proper user interface, no progression system, and almost no visual polish.

The key question we wanted to answer was simple: if a player places a tower and defeats an enemy, does that feel satisfying?

The answer came surprisingly quickly. During early testing, one player spent more than ten minutes experimenting with a build that had almost no graphics, sound, or progression. Watching someone engage with such a basic version of the game gave us confidence that the core loop was strong enough to support a larger experience.

Phase 2: Expanding the Experience

Once the foundation was proven, we began introducing the systems that would shape the game's identity. Additional tower types were added, along with a mobile friendly user interface, a dedicated recovery phase between waves, and the wave preview panel.

This stage marked the transition from prototype to game. Players no longer relied purely on instinct when placing towers. Instead, they began planning around the information provided by the wave preview system. Test sessions showed a noticeable increase in strategic decision making, with players actively preparing for upcoming threats rather than reacting after enemies had already appeared.

The introduction of the recovery phase also improved pacing. Players had time to reflect on their performance, discuss plans with teammates, and adjust their defenses before the next challenge arrived.

Phase 3: Cooperation and Progression

The third phase introduced multiplayer functionality and many of the systems that define Verdant Bastion's cooperative identity. Character movement, online sessions, the Skill Constellation progression system, and the first synergy mechanics were all developed during this period.

This phase revealed one of the biggest challenges in cooperative game design. Simply placing multiple players in the same world does not automatically create meaningful teamwork. Early playtests often felt like parallel single player experiences, where participants happened to share the same map but rarely interacted with one another.

To address this, we focused on creating mechanics that rewarded collaboration in visible ways. One of the earliest examples involved the Driftweaver's slowing aura. When a Rootcaller stood nearby, the aura expanded significantly, making the effect immediately noticeable to both players. The mechanic encouraged cooperation naturally because players could instantly see the benefit of working together.

The Skill Constellation system was also introduced during this phase. Instead of presenting progression as a traditional list of upgrades, abilities were arranged within a gradually unfolding star map. This approach reinforced the game's themes of exploration and discovery while giving players long term goals to pursue.

Phase 4: Polish and Integration

The final phase focused on refining the overall experience and preparing the game for its target platform. Visual effects, animations, sound design, environmental transitions, and social features were all added during this stage.

One of the most impactful additions was the color restoration system. Areas affected by the Veilstorm appear grey and lifeless, but gradually bloom back into vibrant color as players progress. This visual transformation became one of the strongest ways of communicating player achievement and world progression.

The Adaptive Veil system was also finalized during this phase. The AI observes player behavior across sessions and tracks patterns such as preferred tower types, common defensive layouts, and response times to different threats. Using this information, it can make small adjustments to future encounters while remaining within carefully defined limits.

A significant amount of work during this stage involved deciding what not to include. Many ideas that showed promise were postponed to future updates because they were not essential to delivering the core experience. This disciplined approach helped keep development focused on the features that contributed most directly to the game's vision.

By the end of the fourth phase, Verdant Bastion had evolved from a simple tower defense prototype into a fully playable cooperative experience built around teamwork, exploration, and adaptive challenge. Each development phase helped shape the final game while providing valuable insights into what players found most engaging.

Challenges we ran into

Developing Verdant Bastion brought a number of interesting challenges. Many of them were not technical problems, but design questions that required repeated testing and refinement. In several cases, solutions that seemed obvious at first turned out to create new issues that only became visible once players interacted with the game.

Creating Meaningful Cooperation on Mobile

One of the earliest challenges was designing cooperative gameplay that felt natural on mobile devices. Unlike couch co op experiences where players share a screen, every player in Verdant Bastion has their own device and their own perspective. This creates a risk that players feel isolated even when playing together.

Our first attempt at encouraging teamwork was to assign specific tower types to specific players. While this created distinct roles, it also limited freedom and often frustrated players. Testers felt restricted because they could not contribute in the ways they wanted.

We eventually shifted to a more flexible approach. Every player can build any tower, but each character specializes in certain interactions and abilities. This preserves player freedom while still creating opportunities for cooperation. Players feel empowered to help wherever they are needed, while their chosen character continues to offer unique strengths.

Designing the Wave Preview System

The wave preview panel almost never made it into the game.

In early versions, the recovery phase simply displayed statistics from the previous wave. While informative, this caused players to focus on what had already happened rather than what was about to happen. Many players repeatedly made decisions based on past threats instead of preparing for future ones.

We experimented with showing information about upcoming enemies before eventually creating the wave preview panel. Presenting the next enemy wave visually gave players a clear understanding of what was coming and encouraged meaningful planning discussions between teammates.

The challenge was not simply adding the feature but integrating it at the right moment. Through testing we discovered that players needed to see the preview immediately after a wave ended. Once we adjusted the timing, strategic decision making improved significantly.

Balancing Challenge Without Creating Stress

Many tower defense games increase difficulty by making enemies faster, stronger, or more numerous. While effective, this approach often leads to a more stressful experience.

Verdant Bastion was designed to feel challenging while maintaining a calm and welcoming atmosphere. To achieve this, we separated pace from complexity. The overall rhythm of the game remains consistent regardless of difficulty level. Instead of overwhelming players with speed, higher difficulties introduce more varied enemy compositions, smarter routing, and stronger Adaptive Veil behavior.

This approach allows experienced players to face deeper strategic challenges without losing the relaxed tone that defines the game.

Finding the Right Balance for the Adaptive Veil

The Adaptive Veil presented a unique design challenge. If the AI adapted too aggressively, players felt punished for developing effective strategies. If it adapted too little, many players never noticed its presence at all.

Several early versions of the system were too reactive. Players often interpreted adjustments as unfair countermeasures rather than intelligent adaptation. Through testing, we learned that the AI needed clear limits on how much it could change from one session to the next.

The introduction of difficulty based adaptation caps solved much of this problem. The Veil can learn and respond, but only within boundaries that preserve fairness. As a result, players tend to view it as an attentive opponent rather than an adversarial system designed to stop them from succeeding.

Bringing Metroidvania Style Discovery Into a Tower Defense Game

Another challenge involved incorporating the sense of exploration commonly found in Metroidvania games. Traditional Metroidvanias use movement abilities to unlock new areas of the world. Tower defense games do not naturally support that kind of progression.

We needed an alternative that would create a similar emotional experience. The solution became the Lore Gate system. Instead of unlocking areas through movement abilities, players collect fragments throughout their journey and use specific combinations to access new Sanctuaries and story content.

Although the mechanic works differently from traditional Metroidvania progression, it creates many of the same feelings. Players encounter paths that are initially inaccessible, gradually uncover the requirements needed to unlock them, and experience the satisfaction of returning later with the knowledge and resources needed to move forward.

Learning When to Simplify

Perhaps the most important challenge throughout development was learning when to stop adding features.

Many ideas appeared exciting on paper but added unnecessary complexity once implemented. Every new mechanic had to justify its place within the overall experience. If a feature distracted from cooperation, strategic planning, or exploration, it was either redesigned or removed entirely.

This process taught us that successful design is often about choosing what to leave out. By focusing on clarity and purpose, we were able to create a game that feels approachable while still offering meaningful depth for players who want to master its systems.

Accomplishments that we're proud of

The thing we're most proud of isn't a feature, it's a moment. In playtesting, we watched two friends playing co-op for the first time. They'd never seen the synergy mechanic. About eight minutes in, the Driftweaver player dropped their slowdown aura on a chokepoint and the Rootcaller player happened to walk through it on the way to boost a different tower. The Rootcaller's proximity doubled the aura radius, and they both noticed at the same time. They weren't looking at each other. They just both went "...oh." And then immediately started doing it on purpose. That moment accidental discovery of a co-op mechanic that neither player was told about is the exact experience we were designing for. Beyond that:

The colour-shift tile system works more beautifully than we expected. Watching a grey Sanctuary bloom back to full colour as you clear levels is genuinely moving in a way that a progress bar never is. The wave preview panel changed playtester decision-making in measurable ways, average tower repositions in the Prepare phase went up by about 40% once the panel was in place. Players were actually thinking ahead. The Skill Constellation reads as intuitive even without a tutorial. Every playtester who encountered it immediately understood "I can see where this is going, I just can't get there yet", which is exactly the right mental model for a Metroidvania progression layer. The Adaptive Veil at Veilbroken difficulty, after ten sessions with the same player profile, produced wave compositions in testing that genuinely felt designed for that specific player. Not punishing being attentive.

What we learned

The biggest lesson is about constraints as design tools. Every time we removed a restriction, made the HUD richer, gave players more simultaneous choices, added more enemy types earlier, the game got harder to read and less fun to play. Every time we added a constraint, one wave preview at a time, action bar with only four contextual options, recovery phase with a clean single purpose, the game got clearer and more satisfying. Related: mobile-first design is not desktop design minus a keyboard. We redesigned the HUD three times. The final version puts almost nothing in the centre of the screen, because that's where players' thumbs and eyes compete. The game area is the garden. The HUD is the frame. That sounds obvious in retrospect; it took months to internalise. On co-op: shared visibility beats shared control. The most effective co-op moments weren't the ones where players coordinated explicitly. They were the ones where the game's visual language made one player's action immediately legible to another, the Driftweaver aura glowing, the Rootcaller's radius visibly expanding. We designed a lot of explicit co-op mechanics that we eventually cut. The implicit ones, built from visual clarity, did more work. On AI design: a learning opponent needs to feel curious, not hostile. The difference between those two things is the escalation cap. An uncapped Veil feels like it's trying to beat you. A capped Veil feels like it's interested in you. That's a meaningful distinction in a game designed to be relaxing. And finally: the first 30 seconds decide everything. We spent more time on the opening colour bloom, the moment the grey garden becomes Verdant, than on any individual tower type. It's the only thing every player experiences in exactly the same way. It needed to make a promise. We think it does.

What's next for Coop TD: Verdant Bastion

The current build is a complete, playable first Sanctuary with everything the MVP promised. What's next is the vision we built toward but scoped out: Sanctuaries 2–5 — each with distinct visual palettes, enemy types, and Lore Gate configurations. Rimeglade (frost and blue-white light), the unnamed desert sanctuary we're calling Ashveld, the deep-ocean Tidalroot, and the final Sanctuary: the Veil's Heart itself, a level that begins entirely grey and, if fully cleared, reveals something more vibrant than anything the player has seen before. Asymmetric co-op roles. Right now the four characters have different Constellations but similar moment-to-moment play. The version we want to build gives each role a genuinely different interface: the Driftweaver might see wind currents that other players can't, placing barriers rather than towers. This is technically substantial but would make co-op feel like four people experiencing different games simultaneously. Full Adaptive Veil escalation. The MVP caps the Veil's learning to protect new players. With a proper session data set, we want to remove that cap at the highest difficulty tier and see what the Veil becomes when it's had twenty sessions to study a specific group of four players. The Veil's Voice. A narrative layer where the Adaptive Veil gains a written voice between levels, short messages in the game's lore language that describe what it observed. Not threatening. Curious. The way a good chess opponent might say "interesting choice with the rook" between games. Player-built Sanctuaries. A simple level editor scoped for mobile, letting players share custom sanctuary designs as Daily Seeds. Meta Horizon's social infrastructure makes this particularly compelling, a community-built library of challenges is, honestly, the version of this game we'd most like to play ourselves.