#sdfs-kitchen

1 messages · Page 1 of 1 (latest)

vapid cradle May 23, 2025, 4:19 PM

I'm planning to use this space to casually share thoughts from time to time. I’ve tried a few different approaches before, but I think running a cooking channel might be the most fitting for now. If it doesn’t turn out to be effective, I’ll try something else—but for the time being, I just want a light, comfortable space where I can throw out ideas and spark some conversation.

I've tried to explain many things along the way, but I feel it hasn’t gone as well as I hoped. There are probably many reasons for that—but I think the situation became more complicated as we made contradictory choices, or added features and then later removed them. That made it even more important to clearly communicate what we were thinking, but we didn’t do that well enough. I’m sorry.

I’ll just start talking without much structure for now. Some of it might sound a bit strange or scattered—so I hope you’ll bear with me.

Where should I begin? Let’s start with the random modifiers. There are quite a few things tangled up in this topic, so it’s a bit tricky to explain—but I think this is a good place to start.

If we look back at the meeting notes from September 1st, 2022, when we decided to test random modifiers on items, I wrote something along the lines of: "I think simple items, like those in Ultima Online (T2A), suit this game better—but to add more depth (like build variety), we’ll need Diablo-style affixes. Let’s implement it quickly and test it out."

We could have developed the system without randomness, but designing each item individually would take a lot of time.
So, in the beginning, we chose a method where items are determined randomly because it was the simplest and fastest way to develop.

But this brought a huge amount of randomness into the game, meaning that obtaining what players wanted required a tremendous amount of time.
Also, since we didn’t yet fully understand the gear gap, the game ended up having a significant gear disparity. However, because we were still in a short playtesting phase at the time, these issues didn’t become very apparent.

The number of random modifiers that can appear on up to 11 pieces of gear at once grows exponentially, making balancing extremely difficult.
Especially since we designed the system to allow for sharply focused builds—like in Slay the Spire—to enhance fun, that part was achievable.
However, maintaining even a minimum level of balance was pure hell.

So, we tried reducing the number of random modifiers that could appear on items, and we also created our own scaling system called Attribute Bonus to limit the effects of items. However, all these knobs we created only pushed us further into the trap of complexity.

And during the process of adjusting the items, we received a lot of feedback from players who had already become familiar with the system and enjoyed acquiring powerful items, saying that the changes were moving in a negative direction.

We also got significant negative feedback from players who had already obtained many good items. The biggest issue was that we made those changes in the middle of the season. Although we anticipated this negative reaction, our wipe schedule was too far apart to keep up with the fast pace of testing.

So, we created a test server, but it wasn’t utilized effectively due to the small number of users, the difficulty of serious testing, and internal issues: the update process was very inconvenient, and migration to new versions had to be done manually. Because of these factors, it kept getting deprioritized in development.

But now, the test server is definitely easier to update and maintain, and going forward, we plan to make full use of it. I apologize for not doing so until now.

Early players were used to checking information through Discord, and during the early access phase, patch notes and other updates were available via the Blacksmith launcher. Because of this, creating an in-game feature to check patch notes and other information was deprioritized. I apologize that this feature is still missing in the game. We will prioritize developing it soon, so players can access this information without relying on Discord or other external tools.

Once the news feature is updated, we will also share test server updates and testing intentions through it. If something is updated on the test server, please participate actively and leave your feedback.

I will also consider whether there’s a way to allow users on all platforms to access the test server. Since access to this server has been an exclusive feature for Hold the Line owners, we’ve been hesitant to fully open it up so far. However, for the sake of better improvements, I’ll think about how we can make this change without upsetting the Hold the Line owners.

Let’s get back to talking about the random modifiers.

This feature brought several benefits to the game, but it also came with some serious side effects. One of the positives is that it allows for diverse builds without us having to manually create everything, letting players make interesting choices on their own.

However, it’s also highly luck-based, which can easily become addictive and time-consuming. Additionally, the many possible modifier lines on a single item can widen the gear gap even among items of the same rarity.

Because the gear gap is so large, people started complaining:
“We only have gray (Poor) items, but we got crushed by players who have purple (Epic) items in every slot!”

At the same time, since we only had trio queues and no tools to support them, we also received feedback like:
“If you don’t have friends, you can’t really play this game.”

We introduced solo queue and created the Goblin Cave dungeon. Soon after, since people also wanted duo play, we added duo queue as well.

I personally didn’t like this approach, but we wanted to satisfy the players. So, whenever I got the chance, I kept insisting every day that we should remove solo and duo queues.

Since I’m speaking from memory to keep things casual, the actual update order might be a bit different.
I hope you can understand that. I believe conveying the ideas is more important.

And then we took a closer look at the gear gap. At first, we thought it could be mitigated if everyone had powerful items. So, we tested things like adjusting drop rates, having merchants sell rare or epic items directly, or allowing players to buy items with their stats revealed. Sometimes these worked as intended, but other times unexpected side effects occurred.

But even so, players still often fell victim to the large gear gap. The measures we took weren’t enough to keep up with the ridiculously low chances of obtaining Best In Slot items. It ultimately required a lot of time.

And unlike typical FPS games, our game required players to learn many new skills to perform well in combat, so a lot of time was also needed to improve their skill.

In the end, those who could invest a lot of time tended to have both strong gear and high skill, while players who couldn’t play as much found themselves at a disadvantage in both aspects.

If we represent players who can invest a lot of time in the game versus those who cannot as a pyramid, the group that can invest a lot of time would likely be smaller.

Anyway, we saw that players who couldn’t invest a lot of time were repeatedly failing to get good results and eventually quitting the game. So we began thinking about how to keep these players engaged.

Players who were already doing well wanted more content and better balance, but we couldn’t invest much in that area.

Since we believed the game’s systems weren’t yet complete, we started developing various systems to better support the player base.

Especially after the introduction of the Adventurer Rank, we implemented systems like SBMM (Skill-Based Matchmaking), which used Adventurer Rank to separate matchmaking pools; GBMM (Gear-Based Matchmaking), which used gear score to match players of similar strength; and eventually, realizing that these dynamic systems required too many fragmented pools, we introduced a bracket system that separates dungeons based on item rarity.

We knew from the beginning that having solo/duo/trio queues, multiple dungeons, normal/high-roller modes, and various matchmaking pools would fragment the player base and risk turning the dungeons into ghost towns.

However, our priority was to see if players could have a meaningful progression experience—moving through brackets or pools step-by-step within a support system—even if things were split at first. That’s why we went ahead with these attempts.

But the loot system—what we call the “Lootdrop system”—was still in its infancy at the time.
We couldn’t properly tailor the loot drops for each bracket or pool, which meant players couldn’t really use those brackets as stepping stones.

Because the time it took to earn competitive items was still far greater than the time it took to lose them, most players ended up concentrating in the 0–24 bracket or the Common bracket—where the risk (and stress) was lowest.

We also introduced the CMM (Continuous Matchmaking System) to help alleviate the issue of dungeons constantly being empty due to the many matchmaking pools.

Before the introduction of CMM, we returned to Steam and, wanting more players to experience the game, we switched to a semi–free-to-play model and launched a large-scale(at our level) marketing campaign. In hindsight, it would have been better if we had promoted it as “Free-to-Demo.” But like fools, we made the mistake of advertising it loudly as “Free-to-Play.”

The biggest issue was that the normal dungeon had a restriction allowing only Common items. So, players who obtained higher-rarity gear had no place to use them. To those players, the game felt like a demo, and it gave the impression that we had hidden the real game behind a paywall in a sneaky attempt to make money. As a result, we earned a lot of resentment from players.

That was never our intention. We simply didn't have the breathing room to properly reflect on the side effects of the complex systems we had built.
In any case, the flood of negative feedback and review bombing began. In response, we introduced a bracket system based on gear score, allowing Squire Status players to use the Epic and Legendary items they had acquired.

Ultimately, the biggest issue was the gear gap and the excessive amount of time required to catch up with it. There was no perfect solution without directly addressing this core problem. We were able to achieve some level of control by introducing multiple brackets, but the system felt artificial. For example, the 25–124 bracket ended up being the least played because, due to the poor loot system, it wasn’t meaningfully different from the 0–24 bracket.

There were times when I wanted to remove the random modifiers altogether, but the game had already come too far for that to be an easy choice.

We also tried to drastically reduce the item gap with Patch #69, but since it happened mid-season and the items with dead rolls didn’t even maintain the minimal stat gap associated with rarity, it resulted in overwhelmingly negative feedback due to my poor preparation. Patch #69-1 was slightly better, but by then, most players had already lost interest, and I wasn’t able to explain the intention behind the changes as clearly and thoroughly as I am now.

Even I personally felt unsatisfied when acquiring items in that state, and so everything was reverted in order to take a step back, reconsider everything from the beginning, and make a better judgment.

Had I thought things through more carefully, done more research, and taken the time to implement things properly, players wouldn’t have had to go through such a frustrating experience. It’s entirely my fault. My heart aches too. I’m truly sorry.

Currently, if we were to remove random modifiers, I believe there still needs to be a slight difference based on rarity, and that rarity should function as true rarity. Legendary items shouldn’t feel legendary because of their power, but because of how rare they truly are.

I’m planning to test this on the test server sometime next week.

Of course, this does not mean the absolute removal of random modifiers. They still have their advantages, and especially recently, we have successfully reduced a significant amount of time consumption through the socket system.

I've also recently come to realize the other issues caused by the large number of systems and modes we've added. Perhaps I had already sensed it earlier, but I might have turned a blind eye, hoping that we could still find the right spot someday.

I’ve come to understand that, with our current team size, it's simply impossible to perfectly balance all these modes or maintain so many interconnected systems without bugs. It only increases the team’s burden. Because I’ve always pushed for fast development, we made a lot of trade-offs in the code and didn’t give ourselves time to properly reorganize the complex systems we built. As a result, they started to interact in unpredictable ways, creating massive bugs.

There have been bugs caused by recovery systems, the Dungeon Expressman, and CMM. There’s synergy (or rather, negative synergy) between the rarity and gear score bracket systems, and the Squire’s Quick Prepare system. The Druid, designed to handle everything decently, became overpowered in the solo queue environment due to the gear score system—since it doesn’t carry a weapon, it can wear relatively powerful gear and thus became an all-round powerhouse. But even so, it still struggles to perform in trio queue. These are just a few examples off the top of my head, and there are countless more I could list.

Because I was too long trapped in the desire to solve these fundamental problems, our team members, as always, have been doing their best to the point they can’t work any harder, but overall, our team has also slowed down.

At the same time, to address feedback from hardcore players who wanted items to be a bit stronger to enjoy acquiring them more, or who disliked “crow play” — looting the corpses of others while watching their fights distracted in the High Roller dungeon — we strengthened items and brought back circles and random portals in the High Roller dungeon, which likely caused confusion among players of other tiers.

This game has a wide variety of player types, and their voices often conflict with each other. Because of this, we had to listen to player feedback with extreme caution, and that remains a very difficult task even now.

We also wanted to create an environment where casual players could stay and enjoy the game, while at the same time hoping that hardcore players who had invested thousands of hours would continue to have fun. So we tried to support both sides, but in the end, due to our ambition and limited resources, it feels like we weren’t able to fully satisfy either.

Even within our own team, there are many people with completely different playstyles. For example, I prefer to be a hardcore PvP player—though I often lose—and mainly play in trios, never playing solo. In contrast, Terry—although that wasn’t the case in the past—has recently come to prefer more relaxed PvP. Since he may have to quit the game at any time, he avoids team play to prevent inconveniencing others and mostly prefers to play solo. Of course, this isn’t the only way we play the game. Please don’t misunderstand or overinterpret these examples.

Still, what we have in common is that we don’t have much time to play games, but despite that, we genuinely enjoy our game and try to invest as much time as possible into playing it.

Despite having such different personalities and playstyles, we respect each other deeply while passionately trying to persuade one another—and since we're both quite persuasive, we sometimes end up implementing features in the game that we don’t personally prefer.

For example, when Terry was trying to find ways to keep new players in the game, I advised him that the biggest source of stress for early players might be sudden PvP encounters, and that we could even consider removing them entirely in extreme cases. When the PvE-only Adventure mode was being developed, I personally didn’t want to include it because I was aware of the various side effects it could bring. However, I agreed that it could be a good way to onboard new players and provide the simplest way to press the “Play” button—so it was added.

Terry wanted to keep Fog of War, and we clashed several times over this as well. But he eventually agreed with my points: that for players who play repeatedly, brute-forcing their way through the map increased fatigue; that it gave cheaters a significant advantage; and that new players often couldn’t find the extraction points, making the game harder and actually lowering new player retention. So it was removed.

Since I'm speaking in a stream-of-consciousness style, it seems I've strayed a bit from the main topic—but this is that kind of kitchen, so let's just get used to it 😅

vapid cradle May 23, 2025, 6:58 PM

Anyway, to sum up, the large item gap and significant time investment caused by random modifiers, combined with our intention to support various player groups, led to creating many pools across regions with different dungeons, modes, party types, and brackets. This resulted in difficulties managing the game and caused players to be too scattered depending on the region, among other side effects. Moreover, the numerous experiments and attempts to solve these problems increased the development team’s workload and made it seem like we were stuck in place.

Back in October 2024, I declared that I would pursue the original vision and wanted to get closer to the game I originally set out to create. However, due to my limitations and various factors, I wasn’t able to get as close as I hoped.

There are many ideal aspects I wish for this game, but at least regarding dungeons and combat, I hope that inside dungeons, players in parties of various sizes and with different objectives can coexist more naturally. Also, I want the gear’s impact on win or loss to not be too overwhelming, so that players without the best items—though generally at a slight disadvantage—can still have at least some hope in combat situations.

Up until now, we tried to solve that issue by separating it into various modes, but that phase of experimentation is now over. I will reduce all modes, party types, brackets, etc. to only those with clear intent, and I want to properly refine and polish these minimized targets.

Terry, the design team, and I will study ways to support player groups who were previously supported by various brackets and modes, so they can perform well in a minimized bracket and mode environment.

For example, in the early days of CMM, exits opened too quickly, allowing players to escape too easily, which caused various issues. However, this also provided a means for relatively weaker players to farm near the exits and escape safely. While it’s not entirely clear yet, I believe the various insights we gained through these experimental tests will be helpful moving forward.

Additionally, I am determined to streamline the item system to be fun while eliminating the side effects mentioned earlier. Among the many knobs we have, I want to at least fix the item aspect to some extent and focus more on class balance and class development.

Our team is already fully committed to developing new systems and content, such as the ocean-themed dungeon named The Blue Maelstrom and the Gathering Hall, so I just need to stay focused and do my part well.

Until now, I hadn’t thought about explaining things in such detail, which caused a lot of misunderstandings and speculation. From now on, I will do my best to explain everything thoroughly through this channel.

As I always say, I am just an ordinary person who loves the game like many of you around me. So sometimes my thoughts or directions might be wrong. Because of that, I don’t think I can promise that everything will always go perfectly. You’ve been disappointed too many times to believe it anyway 😅

But I can promise that no matter the situation, I will always do my best.

Next time, I’ll talk about something you all might be curious about. And if it feels like I didn’t explain today’s topic well enough, I’ll make sure to clear things up. Catch you later!

vapid cradle May 26, 2025, 9:10 AM

We recently reduced Normal mode to Trio-only. At that time, we chose to set the item limit to Rare—even though we were aware that doing so might not be ideal given the current level of gear disparity, but that is what we wanted to test. The reason was to test whether randomly matched teams could still play well under such conditions.

Previously, the matchmaking pool was too fragmented, especially at higher gear levels, making it difficult to even conduct proper testing as random matchmaking often failed to find a party or required too much time.

Although the new Normal 'Rare' pool mechanically succeeded in forming parties, it did not performing as well as we wanted due to the still too large gear gap and the diverse objectives of the party members. Additionally, the lack of in-game communication tools further contributed to these issues.

As a result, we plan to change the item restriction in Normal mode to 'Common' rarity and maintain that setting throughout this season. The more limited power range for starting items should better align party expectations until we can directly address the item gap issue.

High-Roller mode is currently divided into Solo, Duo, and Trio queues, a structure designed to accommodate players participating in the Adventurer Rank. Once the Adventurer Rank ends, these queues will be merged into a single Trio pool. At that point, we hope to offer a wider range of communication tools to better support cooperative play.

vapid cradle May 29, 2025, 4:43 PM

Let's move on to another topic. This time, it would be good to focus on the changes made to the test server.

I'm not quite sure where to start... From my perspective, it's a complicated matter with many intertwined elements, but I'll try to explain it anyway.

Then let me start by talking about the removal of Adventure Mode and the ongoing squeezing of party types. What I’m aiming for in dungeon gameplay is a sense of naturalness. I often imagine what it would be like if we were truly adventuring in a medieval fantasy world—how would we appear inside a dungeon?

In such a world, it would be natural to see adventurers with different goals and groups of various sizes.

Of course, a real medieval fantasy world doesn’t exist, so we can only experience it through creative works like novels, movies, and games. But at least in the world I imagine, that’s how it would be.

From that perspective, the existence of dungeons limited to PvE only, or item brackets meant to reduce gear differences, or requiring fixed party sizes like solo/duo/trio to enter the same dungeon—without any kind of special magic or justification—feels very artificial.

vapid cradle May 29, 2025, 5:36 PM

That said, it's true—we don't actually live in that world. So at times, artificial adjustments can make the game more enjoyable. And we decided to explore that direction.

For example, when we chose to add the Goblin Caves for solo players, we could have also pursued a version that supported mixed party types.
But we were also genuinely curious about the potential of solo gameplay. Adding a solo queue and a dedicated dungeon was something we could move forward with quickly, and in many ways, it was an attractive option for us.

Both the players and our team enjoyed this direction for a while. However, as time went on and we tried to expand it further, we began to see some side effects emerge.

There were class balance issues specific to the solo queue—like with the Druid. Teaming became a frequent issue in solo play. And because we limited solo queue to just the Goblin Cave, we never built each dungeon with tailored versions for solo, duo, or trio play. This created friction when players tried to engage with content in ways it wasn't specifically designed for.

All of this began to interact with the gear brackets we had introduced to address gear disparities—and at some point, everything started turning into spaghetti.

That's when I realized something important: we simply couldn't maintain or meaningfully improve all of this in a sustainable way.
The cost of trying, testing, and supporting all these parallel systems was eating up too much of our actual development capacity.

To give one example—when building PvE dungeons, we had to audit every class ability to ensure nothing would disrupt the PvE experience.
Even with that effort, we missed things. We couldn’t deliver it perfectly. And even now, it still isn’t perfect.

So I’ve decided that, at least for the time being, it’s better for us to reduce the number of things we need to manage and focus our energy on what truly matters. I’m sorry.

For the time being, I plan to explore a direction where players with different goals and party types can coexist within the dungeon.

Some of you may remember our previous test in the High-Roller mode where we removed name tags from party members. That experiment moved in the opposite direction of making trios easier to coordinate, but it was ultimately an attempt to see if we could naturally reduce the advantages of trio play—so that various party types could more comfortably coexist in the game.

As for supporting well-organized trios and improving their experience, I’d like to explore features like enhanced in-game communication tools, better ways to find players with shared goals and language preferences through the Gathering Hall, and faster ways to connect with teammates through a guild system.

Now, let me talk about random modifiers. I think I’ve mentioned a few times before that I wanted to remove them—though I’m not sure if anyone remembers.
The current implementation of random modifiers doesn’t just contribute to gear disparity; it also has a significant impact on readability and the movement speed meta.

"Then why did you strengthen random modifiers this season? Isn’t that the opposite of what you're saying now?" So before I get into anything else, let me address that first.

vapid cradle May 29, 2025, 6:12 PM

As I mentioned earlier, random modifiers were introduced experimentally and have been a core part of the game for a long time—bringing both meaningful impact and many side effects.

On the positive side, they provided build variety and increased item value, giving players a reason to keep playing. Of course, we were also well aware of the negative side effects and received plenty of feedback and suggestions about them. Still, I wanted to spend more time working with the system myself to fully understand it.

Along the way, we experimented with adjusting the power of random modifiers—sometimes reducing or increasing the stats they provided.
Early on, when we lowered their power, we received a lot of negative feedback saying it made farming feel less rewarding and diminished the overall fun of the game.

We’ve long believed that the gear gap—especially how much time it demands from players—was a major issue. However, as I mentioned earlier, rather than tackling the root of that problem, we tried to respond by increasing item drop rates or dividing players into gear brackets.

Keeping that feedback in mind, I felt that this season—especially with the introduction of Adventure Mode and the highest number of modes and brackets we've ever had—was a good opportunity to test what would happen if we made random modifiers stronger. I wanted to see how players would respond to that.

It would be good to also explain and cover the All Attributes and Movement Speed options while we’re at it.

When I tested the All Attributes option multiple times (though it’s been a while and my memory is a bit fuzzy), I wanted to verify whether this option gave a relative advantage mainly to frontline classes. After conducting several tests, I agreed with the negative feedback and didn’t see much necessity for it, so I eventually removed it.

However, over time, I forgot about it (please keep in mind that I’m just a chimpanzee who cooks), and someone suggested that applying a very small amount of it to certain slots (like necklaces) might add some fun, so I decided to try it again.

That said, I’ve always considered this option to not so much enhance build variety as it simply boosts characters evenly, which in turn only widens the gear gap.

I can explain the movement speed option simply. Before diving fully into the movement speed meta, there was one thing I needed to verify.

Clerics, who lack ranged attacks and mobility, have generally been a less favored combat class outside their healer role (except during one period of buffball). They looked very helpless when losing to faster classes.

So, I wanted to check if clerics could still perform strongly even without ranged attacks when given maximum movement speed. To help with this, I also asked good cleric players who have reached the Warlord rank in our team.

As a result, in mid- to low-level play, combat clerics became very powerful and popular, but at the top-tier level, there was little difference.

To sum up simply, I wanted to see how combat would unfold if everyone had 330 movement speed. It would be ideal to test all these details internally, but unfortunately, we don’t have the capacity for that. Still, we plan to use the test server as much as possible going forward.

Let’s return to the topic of random modifiers. While they are currently removed from the test server, this is just a starting point.

We haven’t yet decided whether we’ll remove them entirely or keep them in another form—like the unique items we’ve experimented. The final decision will be made based on how testing goes on the test server. Our goal is to preserve the advantages while removing or minimizing the downsides. But one thing is certain—the final version will not be the same as what currently exists on the live server.

Once this round of testing is complete and the direction is finalized, I hope it will become easier to predict the general capabilities of an opponent based on what they’re wearing. Ideally, we’ll no longer see Fighters wearing full plate armor but running around extremely fast thanks to stacking movement speed on every gear piece.

And based on this new foundation, we’ll likely need to go through and re-tune the abilities of all classes once again. and I expect that item drop rates will also need to be adjusted significantly to match these changes.

My original vision for items was that gear should be a means to an end, not the end itself — and their value should never be too high.

Too many lines on each piece of gear and random modifiers drastically increase the rarity of BiS gear due to excessive RNG. This consumes too much of the player's time, widens the gear gap to the point where you can't compete without investing that time, and inflates gear value, further encouraging RMT.

This doesn’t mean I want to make everything completely flat like #69 — it just means that things have gone too far.

vapid cradle May 29, 2025, 7:16 PM

Now I have to go to sleep for tomorrow. I will deliver other topics later. Thank you.

vapid cradle Jun 4, 2025, 2:32 PM

In Test #11, we conducted internal testing using gear without random modifiers to evaluate the actual gap created by primary attributes (the white stats), and we confirmed that the gear gap remains significant.

For example, Warlord-tier players using rare items struggled to defeat—or only narrowly matched—players in the 1900–2100 rating range who were equipped with legendary gear. This is largely due to the relatively large disparities that still exist in weapon damage and armor rating. While one might argue that this simply highlights a lack of skill expression, when placed in an equal gear environment, Warlord-tier groups won overwhelmingly.

(A similar test was conducted internally during Patch #69, but at that time, the primary goal was to examine the gap caused by random modifiers. This time, I wanted to completely remove them and verify the results once more.)

Moving forward, we will likely continue narrowing these gaps—such as those mentioned above—until they align with our intended balance.

Additionally, in line with my original design intentions, we plan to adjust the base health so that the Fighter—who represents average stats across the board—has 100 HP.

We will also be adjusting the health scaling curve to ensure that high Vigor does not result in excessively high health.

Accordingly, if weapon damage, skills, or spells deal disproportionately high damage under this new system, we plan to make appropriate adjustments.

One of the biggest issues with the current random modifier system is that even the lowest possible value for certain attributes is still too high.

Because of this, other attributes are adjusted to match that high baseline, which leads to overall inflated values. When these inflated stats are spread across 11 gear slots, each potentially having up to 4 or 5 modifier lines, it results in an excessive gear gap.

While the multi-line random modifier system is certainly fun and adds variety, we have decided to shift that kind of diversity toward perks and skills instead. Going forward, we plan to test a more restrained version of the system in order to better control gear balance.

vapid cradle Jun 4, 2025, 3:07 PM

I am sorry. I fully understand the concerns of solo and duo players well.

However, I believe that for the game to truly move forward, the goal should be to have solo, duo, and trio queues work smoothly together within a single instance, as I mentioned earlier.

Within this mixed queue system, we will implement all indirect systems that help players enjoy playing even with unbalanced parties.

On the other hand, we will gradually remove the artificial features that were necessary to separate players into different queues or brackets based on their player type or the amount of time they invest in the game.

Accordingly, we have decided not to use the current system where interaction times change depending on conditions.

This feature was temporarily created to address complaints in 3v3 situations where players wearing expensive gear would kill an opponent and chase the remaining enemies, but then have the defeated player’s corpse stolen by others.

However, I believe that after a player killing occurs, protecting the corpse — the loot and reward of victory — from being stolen by other players should also be a strategic choice.

I also believe this can increase the survival chances of a party that has lost a member and create variables in 1v3 situations.

I believe that for the mixed queue system to succeed, players must have ways to escape even when being chased in combat, and various forms of 1v3 scenarios should be possible.

Therefore, we will bring back the “fast-opening fixed escapes” that was tested early in the CMM, and if the circle hinders this playstyle, it will be removed.

To explain the background behind the reintroduction of the circle:

When we introduced PvE and divided Normal and High Roller into various brackets, our goal was to offer a more casual environment for casual players and a more hardcore environment for hardcore players.

At the time, many hardcore players were suffering from “crow” players — those who would enter with minimal gear and play like hyenas. So we aimed to create an environment where these players would struggle to thrive, and where well-equipped, high-tier players would have to compete with one another to survive.

However, for the time being, I do not plan to artificially create or enforce such controlled environments.

Therefore, if the circle — which forces PvP in the mixed queue environment — ends up hindering that vision, it will be removed.

On a side note, there’s some sad news: our inventory guy has left the company. As a result, features like inventory sorting, search functions, and additional bags and quivers will be further delayed for the time being.

You might find it surprising — as do I — that even with around 80 people having joined and left the company over time, there’s currently no one to take over this task. But that’s the reality we’re facing.

We’re thinking hard about how to address this gap so we can ease your inconvenience as soon as possible. I’m always deeply sorry about these things.

vapid cradle Jun 11, 2025, 1:39 PM

Today, I’d like to talk a bit about the Circle. In the upcoming semi-wipe, the Circle will be removed. I’m sure many of you will suddenly wonder why it’s being removed, so I’ll note down a few reasons here.

There were several reasons why the Circle was introduced to the game initially, but during the early core PvP development phase, it was implemented as a tool to naturally encourage repeated PvP during internal testing.

During internal tests, it was not very helpful for speeding up the testing process if only a few surviving players couldn’t meet each other on the large map and ended up wasting time. Of course, I’m a big fan of PUBG and really like the circle mechanic, so it’s also true that among many other tools to encourage PvP, the circle was the first one considered.

Personally, since I found the circle mechanic fun, I tried mixing it with the extraction genre and created a system where escape portals spawn randomly. This feature was enjoyable and fit well, so we kept it for a while.

But as always, out of curiosity, we also wanted to see how our game would play without the Circle. Some segments of players requested this, and I personally understood well that the Circle could hinder the survival of relatively weaker players due to item disparity or different party types (solo/duo/trio).

So when we created Frost Mountain, we wanted to try out that kind of feature. And it worked well and we liked it too.

So far, from what I understand, players’ reactions to the Circle are quite divided. Generally, hardcore PVP players preferred the Circle because it causes more frequent PvP and makes it relatively uncomfortable for rat players or crow players.

However, not all players enjoy hardcore PvP, and especially until recently, we have been trying to broaden our target player base and support as many types of players as possible. Therefore, we created a session rule without the Circle for all dungeons, tailored to fit their concepts.

Since we decided to split the game into up to five modes(PVE, N1, N2, H1, H2) to separate players who don’t want forced PvP from those who want battles with extremely powerful gear, we thought introducing the Circle would no longer affect those who didn’t want it. Therefore, we wanted to reintroduce the Circle specifically for the High Roller 2 bucket, where relatively more PvP players gather.

Unfortunately, we didn’t yet have a way to apply separate session rules for High Roller 1 and 2. I should have developed that and applied it only to High Roller 2, but due to my misjudgment, it was applied to the entire High Roller and released to you all.

So players who wanted to grind Adventure Rank and didn’t want forced PvP, especially those in High Roller 1, probably felt uncomfortable with the Circle being applied there. I’m sorry about that.

Anyway, for a while now, we plan to explore offering less segmented modes and properly provide a mix queue where solo, duo, and trio all coexist. Therefore, we have decided to remove the Circle mechanic, which rather hinders achieving that goal.

In conclusion, the upcoming EA 6.5 season starting with the next semi-wipe will proceed with three modes: Adventure, Normal, and High Roller, and there will be no Circle mechanic.

vapid cradle Jun 30, 2025, 1:12 PM

While it would be ideal to present an impressive roadmap, we’ve often fallen short in preparation and lacked the ability to anticipate the future, which made us hesitant to make such attempts. Moving forward, we will make an effort to better prepare and take on the challenge of providing a clear roadmap.

Before we share a full-fledged roadmap, I’d like to take a moment to explain my thoughts on the current state of the game and the direction we are heading.

Although I believed I had already explained things from various angles, I’ve come to realize that many people are still unsure about what sdf’s vision truly is. This made me reflect on my lack of skill in communicating our intentions clearly. So this time, I will try to explain things more intuitively.

When we design a game, we fundamentally do not chase after currently successful formulas or create games specifically for target audiences we don't fully understand—just because that audience is large. The simple and honest reason is that we’re not particularly good at developing games in that way.

There are many great studios out there that are exceptionally good at that kind of development—but at the very least, I’m not that kind of person. That’s why, at a fundamental level, I make games for players who are similar to myself. I believe I’ve mentioned this before.

vapid cradle Jun 30, 2025, 1:53 PM

In the beginning, the game was developed based on the principles I mentioned earlier, targeting players with preferences similar to mine. However, as playtests progressed, we came to realize that a much wider range of players than we had initially expected were enjoying the game and seeing its potential.

As a result, instead of continuing to dig deeper in the original direction we had set, we decided to shift our focus toward trying to meet the growing demands reflected in the feedback we were receiving. As I’ve said before, I’m not the type of developer who excels at satisfying audiences I don’t fully understand. But I didn’t want to ignore the players’ interest and expectations—this felt like a major opportunity, and I agreed with the decision.

We discovered that our player base included solo, duo, and trio players; hardcore and casual users; those who enjoyed dungeon crawling; players who loved PvP; and others who preferred non-competitive experiences. This wide range of audiences meant we had to address a variety of needs and preferences.

Because we wanted to quickly build a structure that could satisfy these different groups to at least a certain degree, we ended up making a series of shallow choices and going through countless trial and error. As a result, development often felt like a series of reactions to the moment, rather than deliberate long-term planning.

In truth, we wanted to invest time in building deeper content that would support long-term gameplay cycles—things like training systems, personal housing, and guilds. But the shallow systems we had created early on became tangled together like a chimera, and we spent an overwhelming amount of time tuning and testing them instead.

vapid cradle Jun 30, 2025, 2:45 PM

For example, when we introduced the random modifier system, I personally had a lot of concerns. However, I wanted to realize the fantasy of creating powerful, synergy-based builds—like those seen in roguelike games—and playing the game through that lens. That’s why we decided to test the system during the playtests.

In the early stages of testing, it was actually possible to create “superman” builds with extraordinary damage, health, and movement speed. Weapon Damage, True Damage, and Additional Damage values were extremely high. Despite the presence of powerful options like All Attributes on multiple gear pieces, the TTK (Time to Kill) remained short enough, and with the right setup, players could strategically take on multiple opponents. At the time, the combat seemed promising to me.

However, not all players enjoyed this structure. In the community, there was considerable backlash against the “one-tap meta,” and many asked for longer TTK. Of course, some of that feedback stemmed from the gear gap, which often resulted in public stomping. But at the time, we wanted to listen to the players and try to improve things in the direction they were asking for.

Terry believed from the beginning that random modifiers were inherently difficult to balance and ultimately needed to be removed. I initially disagreed, as I supported the random modifier system. However, I eventually came to the same conclusion—unless we made every player into a glass cannon, there was no way to provide a meaningful sense of hope when faced with a gear gap.

At the same time, I was curious—would ideal combat really emerge if we changed the structure this way? So we gradually limited the impact of overly powerful modifiers, reduced the overall influence of random modifiers, and evolved the system to increase the TTK.

Of course, I still couldn’t fully let go of the original idea of powerful synergy-based builds. That’s why there were times when All Attributes made a return or TTK was reduced again—leading to seemingly contradictory updates. I apologize for the confusion caused by these shifts. They were the result of two conflicting design philosophies clashing with each other.

In any case, the result is that the recent structure has shifted to one where PvP is no longer decided by just one or two mistakes, but instead through multiple exchanges. However, even small differences in gear now feel much more significant, class-based rock-paper-scissors matchups have become harder to overcome, and winning while outnumbered has become virtually impossible.

Competing at a high level now requires a significant time investment. Leveling up your skills takes a long time, and without a highly skilled trio party, grinding and preparing the necessary gear also demands a lot of effort.

I came to believe that this structure hurt player accessibility and would ultimately hinder the game’s long-term growth. So we made the bold decision to rework the structure entirely. That led to Patch #69. Unfortunately, due to my lack of proper explanation and the poor timing of implementing such changes mid-season, the intended message was not clearly communicated and the patch caused unintended side effects. I sincerely apologize once again for that.

I said I'd keep it simple, but here I am rambling on again.

My inconsistent decisions and shallow responses to feedback from various types of players have led to the current state of the game, and I now believe that this outcome is not ideal. That was entirely my mistake, and I will not repeat it again. From now on, I will move forward with clear focus toward the final vision we have for the game. This does not mean that I will stop listening to player feedback. Rather, it means that I want to stay true to the direction we're aiming for, filtering feedback through that lens without being easily swayed.

To put our future direction more simply:

We will move away from the current model where investing time makes a player’s character increasingly powerful, and instead shift toward a model where players become increasingly wealthy. Of course, players who invest time will still feel a certain amount of power growth and advantage, but I want to ensure that opponents never feel completely hopeless when facing them.

vapid cradle Jun 30, 2025, 3:19 PM

To speak a bit more specifically about the current state of the game, I’m generally satisfied with where things are now, and I plan to use this point as the foundation for further tuning.

While not yet fully confirmed and therefore not included in our internal roadmap, I would like to introduce systems such as a mastery system that grants additional bonuses based on your proficiency with individual weapons or armor—for example, gaining +1 True Damage when reaching max proficiency with a Stiletto Dagger, or gaining a slight increase in Armor Rating when maxing out proficiency with a Fine Cuirass. Additionally, based on our previous multiclass testing, we’re also considering a training system resembling a subclass concept, where players could allocate levels across multiple classes within a 20-level cap. To make this possible, multiclass setups—particularly those involving Sorcerer—will need to be tested again, possibly as part of a special event.

As for the quiver, development has been paused because our original idea was to support various types of arrows—fire, poison, ice, etc.—and we haven’t yet had the time to fully explore and finalize that concept. I understand that many players simply want larger arrow stacks or a quiver that works like a gold container, but in my view, Rangers are already strong enough. So for now, I would prefer to keep some element of inventory resource management in place. I ask for your understanding on this.

I’m also well aware that many players are dissatisfied with the sense of reward from item acquisition. We have no intention of keeping things in their current state forever. The reason why the single-line random modifiers being tested on the test server were not included in this semi-wipe is because I wanted to first establish a clear foundation for the kind of combat style we aim to deliver.

That said, this doesn’t necessarily mean that random modifiers will return in the future. However, we believe that each class must have multiple viable builds, and for now, we are focusing on making those builds work properly within the current baseline. After that, we’ll consider whether it’s possible to create something interesting through the Goldsmith’s services using gems.

This semi-wipe will serve as a solid opportunity to establish and refine the standards for combat and game modes in preparation for the upcoming EA7 season. I welcome criticism as a sign of interest, but I also hope you’ll continue to provide feedback so we can build a better game together.

It saddens me to hear claims that we’re intentionally ruining the game. I remain full of hope for its future and still have plenty of energy. All of us will continue doing our absolute best to help the game grow and evolve. We are truly grateful to everyone who continues to support us through all the challenges. Thank you.

vapid cradle Jul 1, 2025, 4:24 PM

As we communicate more, I feel saddened that misunderstandings seem to increase due to my lack of communication skills. However, despite making mistakes, I created this channel with the intention of speaking more often and sharing my thoughts with all of you. So I'd like to take this opportunity to explain in detail what we are planning over the next month or so.

As mentioned previously, we want to use the semi-wipe as a time to organize and streamline many of the things that have been developed so far. However, this is not about turning the game into something that merely reflects my own personal preferences.

I believe that for this game to gradually grow its player base, the gameplay loop—from the moment someone first starts playing—must be smooth, minimizing any friction or discomfort players may feel along the way.

From that perspective, there are currently three key areas we believe are most important to address: improving the experience for solo players, strengthening the meaning and satisfaction of looting in the current setup, and properly tuning the combat system so that players can always maintain a sense of hope for victory, regardless of the situation.

The problem lies in the fact that the approach we've been taking to achieve this has been too extreme. Right now, everything is so entangled that trying to find the right balance feels like tuning a chimera, as I mentioned earlier. That’s why we want to first reset everything to a completely plain state, clearly define our direction and guiding principles, and then carefully and precisely reintroduce the elements we already have or have developed. We no longer want to waste time in this area unnecessarily.

And in that process, we want to accurately identify and address the things we haven’t been able to verify or fully examine until now.

vapid cradle Jul 1, 2025, 4:49 PM

We are fully committed to finding a solution that can satisfy solo players. If the current matchmaking system—where various group types play together—cannot provide that satisfaction, then we will confidently bring back the solo queue. In fact, we would then shift our focus toward fine-tuning dungeon balance and design specifically for solo play to make it more engaging and enjoyable.

We also plan to enhance the sense of satisfaction in looting, and to give the act of obtaining loot more meaning. Whether that means reintroducing random modifiers or adapting the approach previously used for unique items in a new form, our goal is clear: to ensure that players can feel truly rewarded through loot within a structure that doesn’t require adding more gear brackets beyond Normal and High Roller, yet still allows for meaningful gear progression and gaps.

We are also working to build a combat foundation that allows for hope even in situations where a player is outnumbered. I want every player in a party—regardless of their gear or skill level—to have the potential to enjoy the game. This doesn't simply mean ensuring balanced matchmaking between strangers, but creating situations where, even if your friend is less skilled or under-geared, you still have a real chance to carry or clutch the fight, making the overall experience more hopeful, enjoyable, and positive for everyone.

However, if carrying or clutching is only theoretically possible but rarely achievable in practice, then parties with weaker player will end up repeatedly failing, and that group will inevitably struggle to enjoy the game over time.

To enable real moments of carry and clutch potential, I believe there must be tools or mechanics that allow players to reduce incoming damage to zero. Whether it’s consistently creating 1v1 situations through map positioning, eliminating threats before taking hits, or using abilities like guarding to fully block damage—these are the kinds of mechanics that make such combat possible.

Right now, the system might feel somewhat extreme, but I’ve confirmed that this kind of combat is indeed achievable, which is why I’ve said I’m somewhat satisfied with the current state. That doesn’t mean it’s perfect or finished. It simply means that we’ve reached a point where I see the potential to evolve this combat foundation in a way that aligns with the ideals I’ve been pursuing—while also creating an experience that players can enjoy and find meaningful and fair.

We will do our utmost to successfully accomplish the tasks mentioned earlier within the remaining period. Additionally, we will strive to complete the promised content and new systems with high quality and deliver them to you accurately.

I sincerely apologize to those who have experienced inconvenience due to my extreme approach to problem-solving. Please understand that our ultimate goal is to create a comfortable and enjoyable environment for all players who love this game.

vapid cradle Jul 7, 2025, 7:03 PM

During EA6.5 and through our internal tests and reviews, we’ve been closely analyzing how elements like player skill, gear, and HTK(Hit to Kill) interact within the combat system.

Balancing skill vs. gear is incredibly difficult to me because what feels "fair" or "just right" varies from player to player—and even among developers. This is especially true in a game like ours, where both mechanical skill and long-term progression play critical roles.

What we’ve been missing is a shared language or tool to evaluate balance objectively, rather than relying on vague impressions or gut feelings.

In the past, I sometimes used the term TTK (Time to Kill) to describe how fast combat plays out. But over time, I realized that TTK can be a bit vague and doesn’t fully fit the way our game works.

Instead, we’ve found that HTK (Hit to Kill)—the number of hits it takes to kill someone—is a clearer and more useful metric for us. It’s easy to measure and directly reflects two key factors: how strong your gear is, and how well you aim (like landing headshots).

Because of that, HTK gives us a solid foundation for balancing combat in a more structured and meaningful way. Of course, HTK doesn’t explain everything about combat. There are many factors that influence the outcome of a fight, but here, I want to keep the explanation simple.

During the EA6.5 test, we intentionally lowered the overall HTK values and simplified the model by removing additional factors like random modifiers.

As a result, we observed positive effects such as an increased possibility of 1v3 comebacks and more opportunities for players with lower-tier gear to overcome opponents with stronger equipment through superior skill.

However, there were clear downsides. Instant deaths were frequent, which is not suitable for our game where players carry items accumulated over multiple rounds. New players often died from monsters in one hit, increasing the likelihood of quitting before learning the game.

Through this test, I confirmed that the solution I sought was not simply a lower HTK, but rather sufficient room for players to reduce HTK through their own skill.

The basic principles we aim for in combat balance going forward are as follows:

Both skill and gear must matter. Only when these two elements are well balanced can players feel rewarded for their growth and effort.

If skill becomes too dominant, the meaning of collecting and upgrading gear disappears. Conversely, if gear becomes too powerful, the game turns into a simple stat-check, reducing opportunities for skilled players to make comebacks or showcase their abilities.

HTK (Hit to Kill) should be maintained at an appropriate level, allowing players who perform well (e.g., landing headshots) to reduce HTK and increase their chances of victory.

Gear differences must have clear meaning, as this creates the fun of loot progression in the game. However, if the gap is too large, it requires too many gear brackets, making it impossible for skilled players to overcome the difference, so it needs to be tuned to a reasonable level.

Summary
Gear aspect

Maintain an appropriate HTK in fights between players with similar gear levels.
Recognize HTK differences caused by gear gaps as a unique feature of the game.

Skill aspect

Reward precise actions like headshots with higher damage, allowing skilled players to gain meaningful advantages.
Even with gear gaps, provide skilled players opportunities to catch up without undermining the value of gear.

We have prepared an appropriate HTK value for the upcoming patch, and adjusted it so that players can reduce HTK through skillful play to a satisfying degree. I hope this will satisfy most of the players.

Also, since there is no longer a need to undermine the gear power fantasy, we plan to keep the random modifiers in place until someday when new and exciting systems are introduced. These random modifiers will be managed at a balanced and enjoyable level without being overly restricted, and will be maintained at a scale that does not require too many brackets.

I’m satisfied with the internal test results and eager to start adjusting the abilities of underused classes right away, but I plan to spend more time researching the remaining challenge: enabling solo players to reliably play in the mixed queue at a practical level.

This will be the final challenge I want to verify through this opportunity, and I will do my utmost to find the answer. I don’t know what the outcome will be, but please bear with me a little longer. Thank you.

Oh, I almost forgot to mention one more thing. The Arena Test 5 will be conducted in an SSF (Solo Self-Found) format, and players will be able to use items up to Legendary rarity. Additionally, to enhance the value of dungeon farming, the socket system will be adjusted so that players cannot reach maximum stats solely through sockets. This change is scheduled to be applied in this week’s patch (Hotfix #90).

The reason SSF was not maintained recently in the arena was due to accessibility concerns. However, since we have now created more opportunities for players to overcome gear deficiencies with skill, we plan to retest the SSF format.

The arena must be clearly linked to the dungeon farming loop, and the significance of dungeon farming should be preserved at a high value.

vapid cradle Aug 17, 2025, 3:56 PM

It’s been a while. I’ve been caught up with various matters and haven’t had much time to focus on the Kitchen. Today, I’d like to take a moment to share what has happened so far and the conclusions we’ve reached.

First, let me once again summarize the original vision. The original vision is to create a combat environment where parties of various sizes and goals can naturally coexist, where gear does not excessively determine the outcome of battles, where items function as a means rather than an end, and where time investment leads to gradual accumulation of wealth rather than overwhelming power growth — ensuring that no one feels hopeless in combat.

Coexistence: An environment where parties of different sizes and goals can naturally play together.
Balanced Gear Impact: Gear should not excessively determine victory or defeat.
Role of Items: Items are a means to achieve goals, not the goal itself.
Shift in Growth Model: Time investment results in gradual accumulation of wealth, not excessive power scaling.
Combat with Hope: A structure where no one feels despair in combat and everyone has at least some chance.
Long-term Goal: A sustainable game that can be enjoyed for more than 10 years.

In summary, this means building a game where, even within a competitive structure where winners and losers are inevitable, players never feel hopeless in defeat but are instead given the hope to try again — ensuring continuous engagement. On this foundation, the goal is to create a game that can be enjoyed and sustained for more than a decade.

This was shared to explain how things have evolved starting from the original vision, not to represent the entirety of our future direction.

And I believe that sharing my personal way of thinking with you is the best way to help convey understanding.

To properly address any phenomenon or problem, I believe it is essential to first identify and fully understand its root cause — its essence. This is similar to a programmer’s mindset: instead of simply using a piece of complex code left behind by many people out of habit, one seeks to fully understand its structure and then use or improve it correctly.

So when I sense, whether through intuition or surface-level observations, that there may be a problem with the game’s foundation or its core, I cannot just leave it as is. I must dig in, analyze, and organize until I can truly convince myself and gain certainty. I believe this attitude is ultimately what gives us the power to create a truly great game.

That said, this does not mean that every element of the game must be 100% perfect. What I mean is that the parts that are truly important must at least surpass a certain “critical threshold” of quality. That threshold can vary depending on the nature of the game or the element in question. In some cases, 100% completion may be required, while in other cases, 50% may be sufficient.

Problem Recognition: What was the core issue?
From the playtests through early access launch and the initial stages of service, we developed rapidly while trying to accommodate a wide range of player requests. In doing so, however, we should have taken more time to deeply understand and resolve the underlying issues. Instead, within the fast-paced development environment, we often chose the easier path, and as a result, additional layers of complexity and problems were created.

Of course, the biggest issue was that there was simply not enough to do within the game. Because there was a lack of diverse ongoing content and endgame goals, each season ended up repeating a similar gameplay loop. However, I will not be addressing that issue here.

1) Party Type Issue
Feedback

In the trio-only playtest environment, solo and duo players found it difficult to enjoy the game.
Solo and duo players felt it was unfair to go up against trios.
It was also difficult to find people to play with.
Final Action
Added solo, duo, and trio options to all maps.

2) Gear Gap Issue
Feedback

Large gear differences often led to hardcore players slaughtering casual players, making the game difficult to enjoy.
Final Action
Introduced up to 5 GBMM (Gear-Based Matchmaking) brackets.

3) Excessive Time Investment Issue
Feedback

Securing competitive items required excessive time, making it hard for players with limited playtime to compete.
With large gear gaps already in place, the low probability of obtaining valid gear due to random modifiers made farming take too long.
Even in the Trading Post, obtaining desired items was cumbersome and time-consuming.
As a result, casual players often couldn’t keep up with the gear gap and tended to quit.
Final Action
Again, introduced up to 5 GBMM (Gear-Based Matchmaking) brackets.
Adjusted item drop rates, added higher-tier shop sales/crafting, and implemented the gem & socket system.

4) Combat Hope Issue
Feedback

In 3v3 situations, losing one member would drastically swing the match, making victory chances extremely low.
Playing with teammates who were less skilled or had weaker gear often led to repeated losses, discouraging team play and ultimately shrinking the community.
Final Action
Adjusted HTK (Hits to Kill) across weapons and armor, and introduced 200% headshot damage.

Core Issue: Matchmaking Pool Fragmentation
Issues 1–3 ultimately stemmed from the demand that “players want environments suited to their own conditions.” However, by providing all of these as separate spaces, the matchmaking pools became excessively fragmented — leading to degraded game quality and significantly higher server costs.

Resolution and Direction: What did we decide to do?
We made many attempts over a long period to resolve the various issues above, but in the process, new core problems kept arising. As a result, we lost confidence that we were truly addressing the fundamental causes of these issues. Ultimately, we decided to simplify each element in order to uncover and focus on the root causes.

In addition, with new dungeons on the way, we anticipated that matchmaking pools would only increase further. We judged that hasty fixes would simply cause the same core problems to repeat, so it became clear that we needed to establish a properly consolidated and complete system to prevent this from happening.

Conclusion: What did we gain as a result?
After simplifying and reviewing each element one by one, we were able to confirm whether each was truly necessary for the game, and if so, to what extent it should be maintained.

For example, the conclusion regarding gear differences may seem obvious: if the gap is too large, it feels unfair, and if it is too small, players lose their sense of purpose. However, my philosophy as a director is not to simply accept such “obvious” conclusions at face value, but to thoroughly verify them within the actual system. So we tested, “What if gear differences did not exist at all?” The result was that players lost their motivation to pursue better gear, and the meaning of gameplay itself became diluted.

Through this, we confirmed that gear difference is not just an optional element, but a core factor that supports the game itself. The key is not whether it exists or not, but where to set the proper threshold.

Party Type Issue
It is essential to provide an environment where players outside of trio parties can also enjoy the game. We reviewed the possibility of allowing solo and duo play within a mixed queue, but found it difficult to achieve in a short period, and even if possible, it seemed less enjoyable than having a dedicated queue. Therefore, we concluded that in order for solo and duo players to avoid being forced into “rat play” and to freely enjoy the game, queue separation is essential.

Gear Gap Issue
At first, due to a misanalysis, we perceived gear differences themselves as a problem, but in reality, they were not. If gear differences were eliminated entirely, players would lose their motivation to pursue better gear, and since gear functions as a core progression goal in this game, a certain level of difference was necessary.

However, if the gap becomes too wide, it would require too many brackets or create excessive disparities between rarities, causing some groups of players to lose hope. Therefore, it is important to maintain gear differences at a reasonable level. In addition, we determined that in the future, a minimal bracket might be necessary to protect new players or to give them a natural sense of progression.

Excessive Time Investment Issue
Initially, due to a misanalysis, we believed excessive time investment itself was the problem, but that was not the case. The essence was ensuring that players with less time to invest could still narrow the gap and enjoy the game without losing hope, by providing appropriate support systems.

Combat Hope Issue
The core of the problem was that when players lost, they did not feel “I could have won if I had played better,” but instead blamed the system. This issue was also connected to gear differences: when the disparity in power was too great, it became nearly impossible to retain hope, and the chance of overcoming such situations through skill alone was too small.

Direction Going Forward

Solo, duo, and trio will be fixed as part of the game system, just like Normal and High Roller. This will apply to all dungeons, as long as the player population allows.
Gear differences will be maintained at a level where players can feel progression and value. While extreme gaps may inevitably determine outcomes, the balance will be adjusted so that in normal competitive ranges, players do not lose hope.
To ensure that players with limited time do not give up on keeping up, systems such as sockets and improved item management/trading will be provided to narrow gaps, while reducing repetitive and unnecessary time sinks.
Even in defeat, players should feel “I could have won if I had played better.” To ensure this, HTK adjustments, hitbox compensation, and other methods will be used to guarantee at least a minimal chance for skill to overcome disadvantage.
Matchmaking pools will be kept as minimal as possible to avoid excessive fragmentation of the player base, while still maintaining enough size for the balances defined above to function properly.

We have gone through many trials and errors amidst rapid development and countless experiments. At times, it may have seemed as though we had lost direction, and we caused instability and confusion for you. However, these processes were necessary to confirm whether each element was truly needed for the game, and if so, to determine to what extent it should be maintained. For me personally, it was a process that had to be undertaken in order to provide this game as a service for more than ten years. I am truly sorry for the difficulties this has caused you.

During the Early Access period, experiments and adjustments will continue, but unless there are exceptional circumstances, we will do our utmost to uphold the standards we have now established. Based on these standards, we will shape future changes so that you can enjoy a game in a stable environment, where you do not lose hope and can continue to challenge yourselves.

Thank you sincerely for your interest and support for Dark and Darker.

vapid cradle Oct 20, 2025, 2:29 PM

It’s been almost a month since the start of EA Season 7. I think it’s time to once again take a moment to freely share some unorganized thoughts about how things have been so far and what lies ahead.

Where should I start? Let’s just pick something that comes to mind. I briefly mentioned the newly established balance part in a previous interview, but this time I’d like to go into a bit more detail about it.

Until Season 6, Terry and I handled balance adjustments and internal testing ourselves. However, since we were managing many different tasks at the same time, it was difficult to conduct enough repeated testing to properly validate each change. We were able to establish the overall direction, but we clearly recognized that fine-tuning required dedicated personnel and more specialized tools.

In the meantime, various members of the design team helped out when they could, but since everyone had their own primary responsibilities, it was hard to maintain a consistent testing process. Based on that experience, we established a dedicated balance part this season to enable a more systematic and focused approach going forward.

Thanks to this effort, we’ve been able to improve several abilities that were previously underused, even if the work is still ongoing. We plan to continue in this direction and strive to bring steady and positive changes in a consistent and sustainable way.

Building on this, our goal for the next season is to introduce new abilities to as many existing classes as possible, while implementing ideas that can inspire more exciting styles of play. For example, the Barbarian might gain a skill that allows them to throw their weapon, or a perk that lets them wield a two-handed weapon in each hand.

And just imagine, how fun would it be if you could throw the two-handed weapons you’re holding in each hand? I’m not exactly sure what will happen, but I’m looking forward to seeing some interesting situations unfold.

That said, there’s a bit of disappointing news. We initially planned to introduce the Monk class next season, but we decided to delay its development in order to further refine and enhance the existing classes.

There are many topics I’d like to talk about, but for today, I’ll focus on just one more. the Faustian Bargain.

The Faustian Bargain is currently implemented as a fully experimental feature. Our ultimate goal is to make it possible for players to start a new season thinking, “This time, I’ll play only as a monster.” Ideally, players should be able to join matchmaking as a monster from the very beginning, and progression elements such as Adventure Rank should also be achievable through monster play. With a sufficiently deep mastery and progression system, we hope that playing as a monster can evolve into a meaningful and rewarding experience of its own.

Of course, it should also be available in the High Roller.

The Faustian Bargain is currently in a testing phase, exploring what elements are needed to make playing as a monster truly enjoyable.
Our goal is to create an experience that’s fun both for those playing as monsters and for those facing them.

To achieve this, monsters that move too slowly or feel overly powerful will need adjustments—but rather than simply tweaking numbers, the focus is on making monster gameplay smoother and more natural, allowing the playstyle itself to evolve.
At the moment, monster movement still feels a bit rough and clunky, so there’s still some distance to go before reaching the ideal state.

That said, we decided to release it early in an experimental form, as we wanted to quickly expand the range of playable experiences and refine the system together with players through active feedback and iteration.

Ah, there’s one more thing I’d like to talk about. I’d like to share a few thoughts on PvE.

Since it’s not separated by region, the results may differ from what players in each region actually experience. So please consider this simply as a fun player distribution graph. From this, we can see that PvE players make up over 20% of the total player base.

Based on this, we’ve decided internally to continue supporting the PvE mode going forward. Even during EA Season 6, the share of PvE players remained around 20% until the end of the season, showing consistent engagement. In fact, we expect that additional modes—such as a PvE High Roller—will be needed to provide new challenges for players who primarily enjoy PvE gameplay.

That said, you may have noticed that in this season’s PvE mode, there are once again quite a few elements that allow players to interfere with one another. The reason for this is that when we first launched the PvE mode, we developed the system rather quickly — while it functioned as intended, its internal structure wasn’t as well-designed as we would have liked.

As we’ve been working to reorganize and rebuild those hastily made parts, this temporary state of increased player interference has occurred. This will soon be improved, and the mode will be updated to minimize player-to-player interference as much as possible.

vapid cradle Oct 20, 2025, 3:24 PM

What I personally enjoyed most about this season’s work was that, through Seasons 6.5 and 6.9, we were finally able to fully resolve our long-standing questions regarding the direction of solo/duo/trio and item gap design. Because of that, everything we worked on this season felt like a step forward — progress that pushes the game ahead.

Of course, until we reach version 1.0, there will continue to be experiments in new areas, but at least for the aspects we’ve already tested and verified, I think we can finally move forward without lingering doubts.

Tomorrow, I’d like to share my thoughts on the current game modes based on this direction, as well as the changes planned for next season’s modes. I’ll also talk about the structure and reasoning behind the changes to High Roller matchmaking, The Blue Maelstrom, and several other topics that come to mind.

Lastly, I want to express my sincere gratitude to everyone who continues to play and support the game despite all the challenges along the way. See you again tomorrow!

vapid cradle Oct 21, 2025, 1:13 PM

Now, let’s talk about our thoughts on game modes, both past and future directions.

In previous seasons, we faced several challenges as the number of dungeons increased, along with the introduction of Normal and High Roller modes, different party types such as solo, duo, and trio, and gear brackets based on item gaps. These overlapping systems led to a fragmentation of matchmaking pools and various related issues.

At the start of this season, we removed the world map and replaced dungeon selection with a book and bookshelf interface reminiscent of a campaign book. This change was made because we are preparing for a future where certain dungeons or modes may retire each season.

With this seasonal dungeon rotation system, we plan to continue adding new dungeons to fulfill the fantasy of medieval dungeon exploration. However, since each dungeon varies in scope and completeness, our first step will be to balance the overall content volume across all dungeons before finalizing the retirement rules.

The High Roller mode was originally designed for hardcore players seeking greater risk and greater rewards, naturally dividing the player base between Normal and High Roller modes while maintaining the smallest possible number of matchmaking brackets.

After extensive testing, we concluded that the game benefits from having a variety of party compositions, and that reducing gear gaps too much can undermine the core enjoyment of this genre.

That said, aside from the truly hardcore players who thrive in extreme, competitive environments, most players find it difficult to enjoy the game when the gear gap feels unfair or punishing. For that reason, we believe it is important to use gear brackets to create space for more casual players to play comfortably and grow naturally, allowing them to gradually escalate into higher levels of play.

From that perspective, we wanted to meet the requests of players who wished for the return of the 0–24 bracket, but at the same time, we also had to consider the limitations of the matchmaking pool based on the total player population.

With the understanding that the 0–24 bracket carries relatively low risk, we designed the "Squire to Riches" mode. This mode features randomly assigned dungeon selection and trio-only matchmaking, allowing players to enjoy light PvP experiences without overly fragmenting the matchmaking pool.

However, this mode is still not in its final form. At the moment, various support systems are not functioning smoothly enough to allow players to repeatedly play the Squire to Riches (S2R) mode. Since the process of selling items after extraction and re-entering the game is quite cumbersome, we hope to improve these aspects so that, by next season, players can enjoy a smoother and more seamless experience when replaying the S2R mode.

As mentioned earlier, we plan to further expand and support the PvE mode, and a PvE High Roller mode is scheduled to be introduced either during this season or in the next. We expect that it will require several rounds of balancing and tuning before this mode can fully stabilize.

In this mode, although specific details may change, the Goblin Recovery feature will be disabled to suit the high-difficulty nature of the mode, and it is expected to include circles as well as random dungeon selection.

We also understand that some players who mainly enjoy other modes may be concerned that the PvE High Roller could offer rewards that feel too generous compared to its level of risk. However, at this time, there are no plans to restrict the sale of items obtained through PvE, though the drop tables will be set to a lower value compared to the standard High Roller mode.

We still believe that the step from Normal to High Roller is too steep, so we are currently considering dividing the High Roller into HR1 and HR2. This plan has not been finalized yet, and we expect to share more details once internal discussions progress further. For now, you can think of it as a concept similar to the previous structure of 125–224 and 225+, with two separate difficulty tiers.

In addition, the division will no longer be based on gear score but rather on item rarity, and it is likely that the cap for HR1 will be set at Epic rarity.

However, since introducing additional brackets would require a sufficient player population to support stable matchmaking, we plan to review this idea carefully before making any decisions.

Now, let’s talk about the current changes to the High Roller.

When we revamped the dungeon selection system, the High Roller mode was changed to use random dungeon selection. This adjustment was primarily intended to reduce matchmaking pool fragmentation, and it was a decision made with confidence in the skill and adaptability of High Roller adventurers. This structure is expected to remain in place for the foreseeable future.

However, not all dungeons were given the same selection probability because each dungeon offers a different number of floors. For example, Forgotten Castle has three floors, so it was assigned the highest selection rate to account for players gathering as they progress to deeper levels.
After that, Frost Mountain and Goblin Caves followed with progressively lower probabilities.

Recently, however, internal analysis showed that Frost Mountain’s difficulty and rewards were not properly balanced with its risk level, so in a recent patch, the selection rates of Frost Mountain and Goblin Caves were temporarily made equal.

Currently, Frost Mountain is undergoing balance adjustments to its monster density, debuff difficulty, and other parameters, and if everything proceeds smoothly, the update is planned to be released with Hotfix #99.

This season, the High Roller matchmaking system has undergone some changes. We have introduced a hybrid matchmaking approach that combines the strengths of the previous Bucket Matchmaking system and the Continuous Matchmaking (CMM) system.

We once tested a Bucket Matchmaking system that categorized players based on their gear score when creating dungeon instances. In simple terms, once enough players were gathered, the system would sort them by gear score and then generate multiple dungeon instances, allowing players with similar gear scores to play together.

At that time, the system was applied globally across all matchmaking pools and modes. It worked well during peak hours or on highly active servers, but outside those periods, it caused excessively long matchmaking times.

Even so, one of the strengths of this system was its ability to create multiple instances at once and distribute players across them, which made it relatively effective at reducing pre-teaming and stream sniping.

With the introduction of random dungeon selection in High Roller, we were able to significantly reduce the matchmaking pool, and given the sufficient player population, we decided to combine the bucket system with the existing matchmaking system. Our goal was to prevent pre-teaming and thereby improve the overall quality of gameplay within each dungeon instance.

However, based on our past experience and the current player population, we determined that keeping the same system active during off-peak hours would not be ideal. Therefore, when server activity decreases, the system automatically switches to Continuous Matchmaking (CMM), allowing players to enter dungeons more quickly during low-population periods.

The Bucket System waits until enough players have queued to create multiple dungeon instances, or until the maximum matchmaking time is reached. We are continuously monitoring data and fine-tuning these parameters as needed.

The hybrid system operates so that if the bucket fails to fill even once within the maximum waiting time, it automatically switches to Continuous Matchmaking (CMM). While in CMM mode, the system continuously monitors server activity in the background, and if it determines that bucket matchmaking can operate stably again, it automatically reactivates it.

For this reason, some of you may have experienced longer matchmaking times in High Roller this season compared to previous ones.

Moving forward, we plan to continue improving this system so that it can adapt more flexibly to different situations, while minimizing player inconvenience and preventing unfair behavior such as pre-teaming as much as possible.

We suspect that there may be a bug in the spawn logic related to Continuous Matchmaking (CMM), and we are currently investigating it. We also do not want players to spawn too close to each other, forcing them into combat immediately without time to prepare, or causing them to be caught in endless fights.

We will continue to track down this issue and improve the spawn logic to prevent such situations from occurring.

Ultimately, we believe that the limited map size is one of the root causes. Just as we provided a 7x7 map in the Blue Maelstrom, we are working hard on optimizations to support larger map layouts. Hotfix #98 will include several successful server optimizations, and we plan to build on these small successes step by step to deliver more expansive content in the future.

I was planning to talk about the Blue Maelstrom today, but I woke up early and I’m feeling a bit tired, so I think I’ll call it a night.
See you next time, and as always, thank you so much for enjoying the game.

vapid cradle Nov 20, 2025, 10:27 AM

Today, I’d like to pick up where we left off with The Blue Maelstrom and share a bit of my thoughts on Random Dungeon High-Roller and the Circle as well.

As many of you know, this map has been a huge challenge for us, and to be honest, that challenge still isn’t completely over. I know a lot of people wonder why we felt the need to create this map at all, and some believe it was unnecessary. I understand those opinions, and I empathize with them as well.

vapid cradle Nov 20, 2025, 10:57 AM

But even so, there were two reasons we felt this map needed to be made.

The first reason is that, whenever we create new content, we place a lot of value on giving players a truly new experience. We imagined adventurers crossing a storm-torn sea, sailing through whirlpools, and diving into the depths in search of long-forgotten treasure buried beneath the waves—overcoming dangerous sea creatures along the way.

The fantasy of the ocean has always lived in people’s hearts, right alongside pirates and seafaring legends. We wanted to bring that sense of adventure to life in a way that you could actually explore and enjoy.

The second reason is that we wanted to explore the limits of what we can express in this game, and in doing so, give the team a chance to break through those limits and grow.

As history has shown, underwater combat is notoriously difficult to implement. When we first considered it, we had plenty of concerns as well—but that’s exactly why we felt it was worth challenging. And if we could make underwater combat feel good, we believed it could become a genuinely fun new experience.

Personally, I’m quite satisfied with the initial implementation, and I’m confident that with continued tuning, we can bring it to a really solid place. We’re aware that certain classes struggle, such as not being able to use fire magic underwater, and we’ll work on ways to make sure they still have viable options. That said… unfortunately, fire is still not going to work underwater.

In addition to that, we carried out a lot of research and development while working to implement underwater movement for both players and monsters, as well as mounts and ziplines. I’m sorry that it still doesn’t look completely polished yet, but we’ve made meaningful progress internally, and I’m confident that the experience and knowledge we gained through this process will be a valuable asset for our future development.

Based on this foundation, we’ll continue refining the rougher parts of player and monster movement and work toward completing them with much better precision.

And someday, I really want us to explore space as well.

There are still some serious issues in The Blue Maelstrom. Performance problems, water cross-sections becoming visible, rough transitions when moving from water to land, and even physics bugs that occasionally launch players into the sky… many bugs that I’m honestly sorry to even mention. But we’ll continue fixing them one by one, steadily and carefully.

I understand and am well aware that many of you have negative feelings about the Circle. So, I’d like to talk about it for a moment.

It seems there may be a bit of misunderstanding. My appreciation for the Circle isn’t because I’m a fan of PUBG, but because I believe it has the following advantages.

Random extraction portals
Having a controlled number of random portals creates situations where players must compete with each other to escape. And since higher-tier gear provides an advantage in that competition, it gives more meaning and value to actually wearing good gear.

Additionally, because the portal locations keep changing, players end up fighting in many different level modules instead of the same spots over and over. Combined with the random dungeon system, I believe this helps reduce the feeling of repeating the same adventure endlessly.

Preventing “Crow Play” in a Natural Way
I’m well aware that in High Roller, highly mobile classes with low gear can follow geared players around and swoop in to steal rewards during chaotic fights—and that this creates a poor experience for top-level competitive players.

I still consider this a valid style of play, so I didn’t want to restrict it with artificial rules, like forcing slower interaction speeds or other hard limitations. At the same time, I also didn’t want this to become the most efficient strategy to the point where everyone feels pressured to play this way.

That’s why, within the combination of the Circle and random extraction portals, my hope was that players would naturally equip higher gear to gain an advantage in the competition to survive and escape. And even highly mobile classes would have their movement somewhat constrained by the Circle, helping prevent this strategy from becoming overwhelmingly dominant.

“There were also other reasons, such as wanting to give veteran adventurers a more hardcore adventure experience.

However, after seeing how the Circle works in the Blue Maelstrom and how players interact with it, I was able to clearly understand the exact reasons why players dislike the Circle.

vapid cradle Nov 20, 2025, 12:00 PM

The layout of Blue Maelstrom, where the bosses are placed in two opposite corners of the map, ended up being very unsuitable for boss grinding when combined with the Circle. As we reviewed the new modes and analyzed the different player groups and their preferences, we found that S2R and Arena players generally had little issue with the Circle, while players who enjoy any of the extraction modes had much stronger concerns about it.

From the perspective of an extraction player, it is completely understandable to feel that a design element which only disrupts one of their core goals, such as boss grinding, does not fit the genre. I fully agree with that sentiment.

For these reasons, the Circle will be removed from High Roller in the future. However, we will look for ways to preserve the strengths mentioned above in a different form of design. There is a good chance that the extraction system itself will undergo a significant change, and this change may apply to both Normal and High Roller. Because of this, it will take some time before the Circle is fully removed from High Roller.

Now, I would like to talk about Random Dungeons in High Roller. I know that many HR players have been unhappy with this system.

As mentioned earlier, we understand that for extraction players, choosing a specific dungeon is very important. Many players farm certain dungeons for boss grinding, and the choice of dungeon also affects the strengths and weaknesses of the classes they prefer to play. Because of this, we would like to let High Roller players freely choose their dungeon, just like in Normal mode.

However, with the current system, doing so would split the matchmaking pool too much. This would result in many underpopulated ghost dungeons, which would not be a good experience for players and would also create a significant increase in sever costs ~~maintenance burden~~ for us. We also tested rotation-based approaches in the past, and concluded that they lead to similar issues, so they are not being considered at the moment.

vapid cradle Nov 20, 2025, 12:35 PM

With 4 dungeons, 3 party types, and 2 main modes, allowing players to freely choose among all of them would create 24 separate matchmaking pools. With the current number of players in each region, it is not possible to support that many pools.

Before we decide whether to keep or remove random dungeon selection, we first need to explore ways to continue introducing new dungeons while still keeping the matchmaking pools at a manageable level. This system also needs to include spaces for new players to learn the game and for players who need protection from large gear differences.

We will do our best to find the most appropriate solution and deliver an improved system as quickly as possible. I am always sorry that we have not been able to provide the best possible environment for you to enjoy the game, and I sincerely appreciate everyone who continues to play despite that.

vapid cradle Nov 26, 2025, 2:51 PM

In tomorrow’s hotfix #101, the High Roller dungeon will once again become a selectable dungeon. We have explored multiple solutions for this issue, but we concluded that we need more time to reach the goals we are aiming for.

In the meantime, we believe that temporarily handling a larger number of matchmaking pools is preferable to restricting players who want to choose a specific dungeon.

Therefore, until we finish researching and preparing a better system, the High Roller dungeon will remain selectable.

vapid cradle Mar 8, 2026, 5:18 PM

Now that some time has passed since the start of Early Access Season 8, we wanted to take a moment to talk about the goals we had for Season 8 and the direction we’re looking to take Dark and Darker going forward.

Near the end of Season 6, we ran our final aggressive experiments with the semi-wipe and semi-semi wipe. After that, we declared that the experimental phase was over. With that in mind, we began preparing the subsequent seasons with the goal of providing a more stable experience that players could feel confident about.

We also knew that many players were still skeptical because of the decisions we’ve made in the past, or more accurately, the decisions I’ve made. Rebuilding trust takes time, so our focus has been on showing better results going forward and doing our best to earn that trust back from the players.

Recent Development Direction: Shifting Toward Stability
The new content we prepared for Season 7 followed the same development philosophy we’ve had for Dark and Darker so far, similar to systems like underwater mechanics and mounts. Rather than focusing on safely expanding what we had already built, we chose to push the boundaries ahead of time to prepare the game for the future. Because of that, many of these features were completely new experiments, and some of them didn’t reach the level of polish we ultimately want yet.

With Season 8, we decided to shift our focus. Instead of continuing to add more new experiments, we wanted to make better use of the systems we’ve already built and refine them further. We quickly built Firedeep by applying existing design approaches, and we worked to organize the many systems and rules we had been testing so they could move in a more consistent direction.

At the same time, we also decided to actively improve and introduce features that players have been requesting for a long time, but that we had previously delayed because we didn’t want to release them in a half-finished state.

vapid cradle Mar 8, 2026, 5:59 PM

One of the changes that helped stabilize the game the most was the creation of a dedicated balance team within the game design group. For a long time we didn’t have a team focused solely on balance, which made it difficult to consistently test and validate changes. Because of that, too many adjustments ended up being implemented at the idea stage without enough verification, and that had been a long-standing concern of mine.

Even though we were short on people in certain areas, I ultimately decided to establish a balance team. We brought together several members from other internal teams, and I asked Paul, who has played Dark and Darker for over 2,000 hours(Sorry I haven't played as many hours as many of you.) alongside me, to lead the effort. As mentioned earlier, he is also our CTO, and he was one of the people who designed our in-house anti-cheat system IRONSHIELD together with the head of the anti-cheat team.

Under his leadership, the team gathered balance feedback from players across multiple channels and tested changes more carefully, which led to some great results.

Thanks to the stability brought by the development direction we chose, along with the countless efforts of many team members who aren’t all mentioned here, we’ve finally started to see positive signals in our metrics with the arrival of Season 8.

For the first time in a while, those improvements are now visible in the objective data as well.

Our Original Expectations
When we first started developing this game and thinking about launching it as a live service, we never expected it to receive this much love. Even in the pitch decks we showed to investors, we estimated around 3,000 concurrent players. We believed we would simply build and grow a small but meaningful game of our own with a small team.

As a bit of a behind-the-scenes story, we met with a lot of investors early on, but perhaps because of how modest our goals were, none of them decided to invest at the beginning. Because of that, during the early days of the company there were times when salaries were delayed, and we had to rely on small contributions from dozens of friends-and-family angel investors to keep going. At one point, we even took on outsourced work from a 3D precision mapping company, building 3D buildings just to keep development moving forward. We remain deeply grateful to everyone who supported us during that time.

Looking back now, it was certainly a difficult period, but it was also a very special time for us.

While we were repeatedly running playtests, we suddenly started receiving far more attention than we had ever expected. But for a team that had just cleared the floor mats out of Terry’s jiu-jitsu dojo and squeezed a few desks into what was still basically a dojo, we were in no way prepared to handle that level of attention.

Because of that, even when our metrics reached very high numbers during the honeymoon phase, we had a pretty good sense that they would eventually drop off quickly over time.

Our Original Plan
Because we were short on capital, we had no choice but to develop the game at a very fast pace. As a result, many systems were incomplete, most of them didn’t even exist yet, and the game launched into Early Access with a severe lack of content. It truly was “early” access in the most literal sense.

Our goal was to buy as much time as possible while the many metrics that had rapidly risen thanks to the unexpected love from players gradually came back down, rather than collapsing too quickly. We wanted to use that time to further complete the game, and ultimately bring those metrics back to a healthy recovery and rebound.

Because of various issues, we even had to build our own platform, Blacksmith, and simply getting the game released was an incredibly difficult task for us. In many ways, it felt almost like a miracle. But we managed to make it happen, and over the course of the eight seasons since then, we’ve experienced countless ups and downs.

Right after launch, both we and many of the players were excited. It truly felt like a celebration. But as expected, over time it became clear that many parts of the game were still lacking.

During the process of continuing to build and complete the game, my rather aggressive development approach also caused disappointment and frustration for many players. I want to sincerely apologize for that once again. As I’ve mentioned before, that phase is now behind us, and we do not intend to repeat that approach going forward.

After going through all of these twists and turns, we’ve finally reached Season 8, and we believe we’re now arriving at the point where the changes we originally expected at launch are finally beginning to take shape.

Looking Ahead
With the shift toward a more stable development direction that I mentioned earlier, we’re now able to make much more structured plans going forward. As one example, we’ve already internally determined all of the wipe schedules for 2026. While there may still be some minor adjustments, this finally puts us in a position where we expect to be able to share the roadmap that many of you have been asking for.

After deciding to run the final major tests for the solo/party queue around April 2025, we felt that spending marketing resources during that period wouldn’t make much sense. At the time, our new player onboarding systems were still lacking, which meant that the experience for new players wasn’t very good. Because of that, we chose not to spend any meaningful marketing budget, aside from Twitch Drops.

In Season 8, we also put a lot of effort into helping new players better understand what content to start with and where to begin, so they can approach the game in a more natural progression. While the changes were not dramatic, we did see meaningful positive improvements in the metrics. Because of that, starting with the next scheduled wipe, we plan to begin promoting the game more actively.

Combat System Changes in Season 8
As usual, I tend to prefer speaking to you in a more spontaneous way rather than delivering something overly prepared and polished, so the order of things may have gotten a little messy. When that happens, please just remember that I’m essentially a chimpanzee cooking in the kitchen.

That said, there are still a few things related to player combat in the Season 8 update that I really want to talk about.

vapid cradle Mar 8, 2026, 7:38 PM

In this season, we made adjustments to the jump system, attack slow, and hit slow, and added new parry animations to several weapons.

Using jumps to avoid movement speed penalties could create moments where players felt like, “I really outplayed that.” When executed perfectly, it was a very satisfying mechanic, and personally it was one I liked a lot. However, I always had concerns that it was too difficult for new players to naturally discover, and even when they did, it wasn’t always easy for them to intuitively understand why it worked.

Among the party members I often play with, there is one player whose PvP skills are a bit below average. That’s mainly because they prefer dungeon crawling over PvP in our game, which means they tend to lose most PvP encounters. But because of that, I’ve been able to indirectly observe how new or less PvP-experienced players tend to behave in actual combat situations through that player.

We’ve spent thousands of hours playing other FPS games together and that player has continued to improve over time while still achieving small but meaningful results. We’ve also played Dark and Darker together for close to 2,000 hours. However, despite all that time, that player rarely used the technique of avoiding movement speed penalties through jumping.

To better understand this, I began monitoring the playstyles of many players who were less familiar with PvP. Over time, I became increasingly convinced that while this mechanic could be a fun skill expression for highly experienced players, it was also something that most players would struggle to understand naturally.

Because of that, this season we adjusted the jump-related mechanics to make them more intuitive. At the same time, to preserve ways for players to express skill in combat, we introduced a parry system along with several new combat animations that can be used directly during fights.

These systems are still in their early stages and will continue to be adjusted together with our balance team. In particular, elements such as weapon block impact power and parry animation speed are still quite rough and will require further tuning. We will also continue to explore whether we can introduce additional intuitive systems that allow players to express skill differences at an appropriate level.

In other words, our goal is to make the combat system something that anyone can naturally understand and intuitively grasp, allowing players to anticipate what actions will lead to better outcomes, while still ensuring that the execution of those actions reveals meaningful skill differences between players.

Personally, I don’t think we’ve achieved this perfectly yet, but I’m confident that we’ve taken a solid first step in the right direction. We’d greatly appreciate your feedback and thoughts so we can continue improving it going forward.

I think that covers everything I wanted to share today. As always, thank you for taking the time to listen to my somewhat scattered thoughts.

Going forward, we’d like to make Tavern Table more active by running Q&A sessions where we can directly answer players’ questions. We also plan to use it to share updates about upcoming seasons, insights into what’s happening within the development team, and even some of the fun events happening within the community.

Up until now, we haven’t done the best job of maintaining our own communication channels, and we’ve often relied on the generous help of others to share information with the community. But we now feel that we’re finally in a position to communicate these things ourselves in a more organized way.

We hope this channel can gradually grow into something more established and become content that players can enjoy as well. Thank you for your continued interest and support.