A month ago, a playtest was held for 'Guns & Dragons', a new title from rookie developer AButton. As a "roguelite FPS"—a subgenre mostly explored in the indie scene with few attempts made domestically—it naturally drew interest, but the background behind its development was somewhat surprising. Guns & Dragons wasn't their debut title, nor had their first game even launched yet. With developer AButton actively running marketing campaigns ahead of the release of 'ZEUS: God of Pride', the playtest announcement for Guns & Dragons was certainly unusual.

Joining the test with light expectations while all eyes were on ZEUS, Guns & Dragons turned out to be a game with potential that shone far brighter than expected. Of course, based on the test build, there were unpolished systems, unintuitive mechanics, and plenty of shortcomings—but isn't finding the "right path" precisely what testing is for? The game's core logic and content structure were extremely intriguing, making the two-week test a genuinely enjoyable experience.

And we met them again at Gamescom 2026 in Cologne, Germany. Apart from ZEUS, we sat down to talk about the game with Director Junwoo Park and PM Iksong Kim, who are sprinting toward the Early Access and official launch of Guns & Dragons.

AButton is currently developing two projects simultaneously: ZEUS and Guns & Dragons. This is an unusual choice for a new studio. What is the reason behind running both projects together?

Director Junwoo Park: While I can't claim to know 100% of our CEO's exact thoughts, we believe that even if a new studio makes a great first game, they often fall into the "one-hit wonder" trap. That's why we view it as crucial to establish ourselves as a developer that continuously builds games by maintaining a continuous development pipeline.

To achieve this, we are broadly running our projects under a "Big & Little" strategy. In markets where you compete on quality and scale, we build "Big" projects like ZEUS that showcase what we do best through sheer scale and content volume. Meanwhile, with "Little" projects, we flexibly tackle novel ideas, relatively uncharted markets, and global audiences. Going forward, we envision alternating development between these two types of projects in their respective cycles.

Guns & Dragons falls into the "Little" category. Out of so many genres, why did you choose a roguelite FPS?

Director Junwoo Park: We set FPS as our core keyword right from the start. Our CEO is someone who couldn't let go of his passion for FPS games, and I have also been working on FPS projects for a long time.

However, most FPS titles today are cutthroat, hyper-competitive PvP games. So we thought, "Let's make a multiplayer FPS that we can genuinely enjoy playing." To achieve that, we tried using roguelite mechanics to set the pacing. Upon researching, we saw games like Gunfire Reborn and realized there was potential in this genre. Through R&D, we've been shaping the game into what it is today.

Many other roguelite FPS games involve constantly acquiring and swapping weapons during a run. In contrast, Guns & Dragons is structured so that you continuously evolve your starting weapon. Is there a reason you chose this approach?

Director Junwoo Park: We boldly discarded that element of fun. That's because our starting point was creating an FPS that players who struggle in PvP FPS games could also enjoy. In my case, no matter what game I play, I almost never use sniper rifles. However, in games like Gunfire Reborn or Roboquest, if a high-rarity sniper rifle drops, you feel forced into a situation where you think, "It's a good weapon, so I have to use it." In the FPS genre, we felt that might be pushing players too hard.

While not an FPS, Hades lets you pick a weapon from the start and build intricate setups around it. We thought that if a single weapon has enough depth, there's no need to force players to use a weapon type they don't want.

Of course, when players first experience the game, they might wonder, "Why aren't they showering me with various weapons right from the start?" However, we noticed that in other games, once players get familiarized, they eventually converge on a few powerful firearms anyway. We figured it would be better to let players dive deep into customizing a single gun instead.

In return, cores obtained from other weapons can still influence the weapon you are currently using. If you want to keep using an SMG but no appealing new SMG drops later, you can upgrade your existing SMG enough to keep playing with it. Through this, players naturally get exposed to other weapons little by little, gently encouraging them to try a different build in their next run if something catches their eye.

The last playtest ended once players defeated the first boss. How far along is that in an actual single run?

Director Junwoo Park: Looking at the current overall scale, it's about one-third of the way through. A single run is planned to span up to Chapter 3, and as players near Chapter 3, they will be able to craft final-tier weapons.

During the playtest, some felt that combat spaces were narrow or that map layouts were hard to read at a glance. How are you approaching feedback regarding level design?

Director Junwoo Park: We actually removed some environmental structures during the playtest itself. We received feedback that the density was much higher than we anticipated, making it difficult to parse the space, so we are currently tuning object density.

Some areas feel a bit complex despite being early zones. Caves are relatively easy to navigate, but entering urban areas introduces complexities, so we are discussing various possibilities.

Beyond level design itself, monster spawn distances play a role. Monsters rushed players from too close before they could adapt to a new space, and the moment players passed through a door, combat began with the rear door locking, creating a feeling of severe movement restriction.

Because of this, we are reconsidering options such as keeping doors open, allowing players to pass through freely, or adjusting monster spawn distances. Originally, during the initial blockout phase in design, it wasn't that complex. But as art assets like trees and objects were filled in, it grew cluttered. Going forward, we recognize the need to return to design after art passes to refine the spaces once more.

In fact, gathering this exact feedback was the primary goal of the playtest. The idea was to start small, collect feedback, and make fixes—and far more people participated than expected, allowing us to gather a wide array of feedback.

Indeed, updates were pushed out remarkably fast during the playtest period.

PM Iksong Kim: We pushed out eight or nine hotfixes over two weeks. We reviewed feedback almost immediately as it came in.

Director Junwoo Park: It was painful, of course. Still, I view it similarly to getting a question wrong while studying. Getting an answer wrong hurts, but you have to quickly check where you went wrong and fix it. Stubbornly arguing that "the question was wrong" gets you nowhere.

Guns & Dragons supports up to 3-player co-op. Can multiple players select the same character?

Director Junwoo Park: Yes, it's possible. Multiple players can choose the same character. We aim to offer two playable characters at Early Access launch and three characters by 1.0. Since it's a 3-player co-op game, we feel having at least three character types is essential.

Compared to playing solo, what are the advantages or distinct systems unique to co-op play?

Director Junwoo Park: Fundamentally, difficulty parameters like monster spawns and health adjust accordingly. Rather than calling that an advantage, the first thing players will feel is the psychological safety net of knowing someone can revive you during gameplay.

Additionally, when your build pops off, you can show off or carry your teammates. Systemically, while artifacts drop completely at random, players can trade them with each other. By exchanging artifacts that fit each other's builds, players can create powerful synergies.

There are also artifacts specifically designed for co-op. The whip artifact present in the recent test increases movement speed when hitting teammates. In unrevealed development builds, there are artifacts that transfer your ammo to teammates—every time you shoot a teammate, their ammunition increases. This enables playstyles where one player uses an SMG to constantly supply ammo while another fires heavy, ammo-hungry weapons like cannons.

We are also preparing multiplayer-exclusive events. In combat, we implemented backstab mechanics, which are uncommon in FPS games. Because monsters track players closely, attacking them from behind is difficult solo. However, if multiple players spread out and split aggro, they can strike from behind for bonus damage. We are building co-op combat synergies through mechanics like this.

Healing methods felt quite limited. Was this an intentional difficulty tuning?

Director Junwoo Park: In early builds, there were regenerating shields and health potions dropped much more frequently. However, that led to players tanking damage through sheer brute force to clear stages, which sapped the tension from the gameplay.

While armor drop rates can be adjusted, our direction was to keep health potion availability tight. However, that design turned out to be overly punitive during the playtest, so we increased potion drop rates and added partial health recovery when transitioning between areas. From around day four of the test, players should have had more breathing room regarding HP.

In co-op, breaking a potion creates a healing zone where multiple players can stand together to recover. Rather than explicitly handing out all these mechanics via tutorials, we want to preserve the joy of discovery through gameplay and community information sharing.

Some of the perk choices presented upon leveling up were hard to intuitively understand. Are there plans to rework this?

Director Junwoo Park: We plan to revise them. We got feedback from the playtest, and while running the playtest simultaneously with ChinaJoy, I spent entire days watching people play on-site, which convinced me that changes were needed. That was especially true for skill upgrades upon leveling up.

The moment players see an option, it should immediately click: "If I choose this, this is how I can play going forward." For example, if a node says, "Finishing an enemy with a grenade permanently increases grenade damage by 1%," anyone can see that finishing enemies with grenades continuously builds toward a devastating grenade down the line.

We are currently working to shift toward more intuitive effects like that, allowing players to easily envision their future builds.

How do you plan to expand content after entering Early Access?

Director Junwoo Park: We are designing it so that distinct dungeons appear in each chapter, with multiple dungeons randomly selected within that chapter pool.

Currently, Chapter 1 doesn't have a vast variety of dungeons and represents an early build. Our focus right now is on expanding content volume. Maps currently share similar flow patterns, so we plan to introduce more variations and add two or more boss types even within Chapter 1.

We are producing this content sequentially and will keep adding more content toward 1.0 even after Early Access launches.

When are you aiming to launch in Early Access?

PM Iksong Kim: Our original plan wasn't to go all out right now, but rather to gather player feedback, make adjustments, launch Early Access on a larger scale, and listen to feedback again. However, far more people showed interest than expected, and we received a massive amount of feedback.

Once we faithfully incorporate the feedback received, expand content volume, and judge that the game can meet player expectations at release, we will announce the date via Discord, Steam, and other channels. Both the Director and I continuously monitor Steam and Discord, where we also share development updates and progress.

Director Junwoo Park: Personally, I want to deliver it as quickly as possible. We place huge importance on fast, agile development. However, because we received far more feedback than anticipated, we are currently undertaking major overhauls.

In that case, what level of content volume do you consider the baseline before starting paid Early Access?

PM Iksong Kim: Early Access is the point where players actually purchase the game. Therefore, we believe that what players expect and what we want to deliver must be reasonably aligned and in place by then. Until we are certain that standard has been met, it's hard to commit to a specific timeframe.

Director Junwoo Park: For example, if we are charging players money, they should at least be able to defeat a dragon. Chapters 1 through 3 must naturally be included, and the experience must be fully prepared to complete a whole run cycle before we can consider it worthy of a price tag.

Building mechanics are also present at the base. What made you include a building system in a roguelite FPS?

Director Junwoo Park: The initial concept was meant for meta-progression. We recognize that base building is a polarizing mechanic—you either love it or hate it. So our direction was to make it completely skippable for players who dislike building, while giving players who love it a stage to proudly show off.

In multiplayer, you start at your island, head to the lighthouse, and board a ship. When matchmaking connects players, their islands anchor next to the lighthouse in real time. That makes each player's custom structures visible, allowing everyone to showcase the spaces they've built.

At the same time, because the game incorporates extraction-like elements alongside roguelite mechanics, we are focusing on a resource gameplay loop where materials brought back from runs are put to use. Moving forward, we plan to add minigames and various systems to the camp to build clear motivation for bringing home valuable resources from runs.

If the game successfully establishes itself, what long-term content would you like to keep adding?

Director Junwoo Park: From the beginning, our goal was to steadily expand classes and weapon categories. Internally, we've discussed introducing a Mage class. We could create a Mage and magic staves, equipping staves with various spells to offer a mechanic that feels distinct from firearm crafting while maintaining depth.

At first, we considered selling such content as paid DLC, but upon reflection, selling new classes as paid DLC felt like it posed a significant psychological barrier to players.

Ultimately, if the game does well and sales of the base game generate enough revenue to reinvest into continuous updates, we want to keep adding classes and weapon types. However, we don't have any set DLC plans at this moment.

Are weapon categories locked to specific characters?

Director Junwoo Park: No, they are interchangeable. Just because an Elf character currently uses a bow doesn't limit them to bows—they can wield guns as well. Other characters can also equip bows. Even if a Mage character and staves are added later, the Mage could wield guns, and other characters could use staves.

The reason we added bows in the first place was to offer diverse combat styles within a first-person perspective. Because of that, we are quite obsessed with continually increasing the number of possible combinations.

With so many possible combinations, overpowered builds might emerge beyond expectations. What is your stance on balance tolerance?

Director Junwoo Park: Our game isn't PvP. So we don't intend to enforce overly tight balancing. If players discover a truly overpowered combination, we want to celebrate it and say, "Great find!"

Are there any other new weapon categories you want to add in the future?

Director Junwoo Park: Cold weapons (melee/throwing weapons) are next on our list. We are considering area-of-effect attacks with swords or blades, as well as throwing weapons like shuriken.

However, rather than implementing deep melee combat like Chivalry, we are aiming for something closer to Overwatch. Embedding ultra-deep melee mechanics into our current system structure would be difficult.

Boss battles felt noticeably distinct from mob encounters in terms of spatial design and strategy. What criteria do you prioritize when designing bosses?

Director Junwoo Park: Fundamentally, we prioritize creating a sense of continuity by co-designing boss encounters alongside the elite monsters in that zone.

When designing boss patterns, our baseline rule is that as long as players observe calmly and possess basic mechanical reaction ability, they should be able to clear a boss even on their first encounter. At the same time, even if mechanical execution is a player's weak point, we make sure to provide ample tactical solutions to overcome mechanics.

For the first boss, while dodging attacks with jumps and reflexes is valid, players can also avoid damage cleanly by utilizing surrounding pillars. There are also platforming elements designed to dodge via jumping. Future bosses will build upon this foundation to offer distinct experiences, prompting players to ask, "How should I tackle this one?" We plan to invest even greater care into later boss encounters.

During the playtest, players felt heavy pressure from blacksmith enhancement costs and failure rates. How are you adjusting this system?

Director Junwoo Park: Gold is a run-exclusive resource used only within a single run. Conversely, resources like wood and stone are brought back to camp for base building and other purposes.

Regarding blacksmith costs, there was a lot of player dissatisfaction during the playtest. So we drastically lowered the price. What was originally around 200 Gold has been reduced to roughly 50 Gold.

Blacksmith enhancement is virtually the only mechanic that allows you to upgrade a weapon beyond a single rarity tier—taking a Normal-rarity weapon all the way to Legendary as long as you have the funds. Because of that, we felt high cost and risk were justified. However, players often got only one attempt before failing. So we reworked the system to allow multiple attempts and adjusted pricing accordingly.

You received a massive influx of player feedback during the playtest. What criteria do you use to evaluate situations where player feedback conflicts with the dev team's vision?

Director Junwoo Park: Some team members internally worry, "Aren't we giving in to every piece of feedback?" But we don't intend to blindly adopt every single opinion.

We do what we believe is necessary, and for things we disagree with, we aim to clearly explain our reasoning to players. For example: "It would be difficult to implement this feature due to X reason. Instead, we will refine other areas or enable you to play in Y manner."

I stay active on Discord continuously, trying to answer as many player questions and concerns as possible.

Finally, do you have any last words for players eagerly anticipating Guns & Dragons?

Director Junwoo Park: Far more players participated in our playtest and added the game to their wishlists than expected. However, domestic interest in Korea was much lower than anticipated. While we expected that this art style wouldn't appeal to many players in Korea, interest was lower than even our estimates.

Still, our game has challenging mechanics and plenty of depth to dive into. We hope you will keep an eye on us. We are working hard to build it, and we will deliver a great game.