Hello Lumia Island Survivors!
We’re back with the second entry of the Lumia Dev Journal. In this entry, we want to talk about one of the most widely discussed topics of this season: character improvements and reworks.
The goal of character improvements is to refine skills and create an enhanced experience in Lumia. Unfortunately, the improvements carried out in Season 8 exposed several issues in both process and results, becoming one of the most disappointing areas for both the dev team and our players.
Overloaded Changes & Balance Delays
Overall, the quality of Season 8’s character improvements fell short of expectations. The biggest reason was that we tried to handle too many improvements in a single season. Attempting to increase the volume of adjustments excessively was a mistake, and looking at examples from other games, it is far more common to focus on a smaller number of characters. We now know it was not a good choice.
Managing so many changes at once made it difficult to ensure quality for each one internally. When all those changes were applied to the game at once, they caused significant issues across the play environment, required lengthy stabilization, and ultimately led to feedback such as, “Why make character improvements at all if this is the result?”
Chiara’s case illustrates this problem well. Even though the goal was a relatively simple improvement, the process was not smooth, and the end result created bigger issues than those we were trying to solve.
The aim of Chiara’s adjustment was to reduce her over-reliance on Last Judgement (R2), which made her too powerful when it succeeded and too weak when it failed. To fix this, we made her transformation persist after using Last Judgement (R2).
However, this led to a new issue: Last Judgement (R2) was now being used as a gap closer. As a targeted teleport from long range, it felt unfair for opponents with no real counterplay. We had preemptively reduced the skill’s range to control this playstyle, but player stats worsened, forcing us to partially restore it, which created new problems.
There was also feedback suggesting that the ultimate’s second hit should only activate when Chiara has a certain amount of Stigma stacks. But since this had previously been changed due to widespread complaints about clunkiness, reverting it wasn’t feasible.
In the end, adjusting range, base damage, and Stigma stack’s damage could have solved the issue numerically, but we failed to respond quickly. The reason for this delay was that we were working on too many characters at once. If we had focused on just one or two characters, including Chiara, she could have found her balance much sooner, and the same would have been true for others.
Going forward, we will invest far more care into character improvements. We will move away from the mindset that we must rework a certain number of characters each season. Instead, we’ll take the necessary time, approach the process transparently as we’re doing in this journal, and ensure that adjusted characters stabilize faster after changes.
Felix Rework
Next, we’d like to talk about Felix, the character who received the most feedback this season.
The reason we decided to rework Felix was because of the extreme performance gap between the highest tier and the levels just below. At the very top, he was overwhelmingly powerful and needed nerfs. But just below that tier, he was often among the weakest. Such extreme experience disparities are very rare, which is why only a handful of players could truly make use of him.
It is important for a game to have characters that require heavy study and practice. But the degree of statistical anomaly shown by Felix was almost unprecedented, which is why we don’t expect any characters to undergo reworks for the same reason in the near future. We’ll discuss criteria for selecting rework targets in more detail below.
We chose to rework Retail Spearman (P) because we judged that its range-control mechanic was the main reason Felix became “a character only usable in top-tier queues.” After Season 8’s changes to individual hit detection, the impact of his passive on performance increased dramatically. Keeping the mechanic would have made his statistical skew even worse.
However, this also made Crescent Wave (E)’s basic attack evasion easier to use, which amplified its problems. After the rework, Felix’s pick rate skyrocketed across all tiers, destabilizing the meta. In the 8.3 update, we removed the Spearweaving stack bonus granted on successful evasion to ease this, but as with Chiara, the fix came too late, leaving players with a long period of unfair experience.
Even now, we acknowledge internally that Crescent Wave (E), particularly its basic attack evasion, remains problematic. We plan to restructure the skill by adding conditions or limitations to its evasion mechanic. If you have any specific ideas or directions for this adjustment, please share them with us in the survey.
Question 1: What would you like to see improved in Felix’s E skill, Crescent Wave?
Future Direction for Character Improvements
Just as important as correcting past mistakes and ensuring rollback options is defining the direction of future improvements.
Character reworks, especially those with large-scale changes, inevitably dismantle players’ familiarity with that character. Even if the rework results in a stronger character, players can still feel a sense of loss if the playstyle no longer matches the character they used to enjoy.
Hyejin is an example of this. Her current skill design is somewhat misaligned. While she could be called versatile, the more common assessment is that she feels unfocused. As such, she has become a top candidate for an upcoming rework.
We even prototyped a rework for Hyejin, but debates over whether the direction was correct became heated. Many team members struggled to accept the new playstyle, so none of the proposed versions have been implemented.
Currently, Hyejin functions as a burst assassin centered on her Fear CC. To assassinate an enemy DPS, she must land the non-targeting skill Charm of Despair (Q), then use Charm of the Nomad (E) to bypass the frontline and follow up with her ultimate and another Charm of Despair (Q). This requires risky, high-difficulty play.
We have been considering the following directions for Hyejin:
1) Hyejin should become a clearer burst assassin.
Add penetration to Charm of Despair (Q)’s first hit, allowing Hyejin to contribute more in standoffs while also raising her potential after engaging.
Optionally restore the second hit of Charm of the Nomad (E), making it easier to utilize Feared. However, we would adjust its strength to prevent guaranteed combos with Charm of the Prophecy (W), lowering its solo power compared to the past.
The next option has often been suggested by the community: shifting Hyejin into a Ranged DPS, a playstyle that fits more intuitively with her current Q-skill-centered kit.
2) Hyejin should become a Ranged DPS.
Add On-Hit detection to Charm of Despair (Q), solidifying its role as her core skill.
Increase projectile speed of Charm of the Nomad (E), making it a short repositioning tool fitting for a ranged DPS.
Change the form of Five Omens (R) to allow projecting power further or blocking enemy approaches more intuitively, firmly setting her primary combat range.
Optionally add a basic attack enhancement mechanic if needed.
The third option is to avoid a full rework and instead attempt gradual, partial improvements. Hyejin has long been considered a difficult character to use, but recently she has reached a more stable position, and in terms of pick rate and usability, the urgency of a rework has lessened somewhat. In fact, among the dev team members who main Hyejin, many preferred keeping her current state as much as possible, with only limited changes such as improving the hit detection of Charm of Despair (Q), rather than pursuing the larger shifts outlined in the first two directions. For this reason, we have set this as the third option.
3) Hyejin should continue the same with just a few adjustments.
Partial improvements such as refining Charm of Despair (Q)’s hit detection, while keeping the core of her playstyle intact.
The options shown above may change if better ideas emerge during development or if unforeseen problems are discovered. Please share your opinions in the survey to help us move in the right direction.
Additionally, no matter how much effort is made, it is impossible to predict or prevent all feelings of loss. One lesson from recent reworks is that running system-level improvements in parallel with reworks created unnecessary complications. To avoid this, we will ensure the following for all large-scale character reworks moving forward:
1. Character reworks will be carried out independently and not overlap with large-scale system changes.
2. Rollback-capable code will always be secured as a priority.
Questions 2 & 3: Do you play Hyejin often? What kind of character would you like Hyejin to become?
Character Improvements & Reworks Criteria
Finally, let’s discuss the question most players may have right now: What criteria do we use to decide which characters to improve?
One of the most frequently mentioned candidates for rework in the community has been Adina. Adina is indeed a difficult character to play, but unlike Felix, she performs well even in lower tiers. As noted earlier, we do not expect to rework any other characters purely due to experience gap for the time being.
Experience gap is not the only criterion. For example, Jackie’s rework was chosen because her old, basic-attack-focused kit felt outdated and linear. The goal was to let her play in a more versatile, fun way, but due to poor balancing across weapon groups, the reason for the rework was not communicated effectively to players.
In the short term, we will continue reflecting on feedback about balancing processes that didn’t come out well. In the long term, we are considering fundamental improvements for multi-weapon characters, which we will cover in a future Dev Journal.
A major piece of feedback we have received is the question: “Why improve this character now?” We believe this arises because we have failed to earn agreement from players on the scale and necessity of improvements.
In the future, especially for large-scale reworks, we will establish stronger consensus with players before proceeding. While we have read all community feedback, opinions are widely split: some say older, solo-era characters should be prioritized, while others believe newer or overly versatile characters should come first.
At this stage, we believe the best way forward is to have a clear purpose and hear directly from you through this survey.
Question 4: Which character(s) should be prioritized for adjustments?
In this second edition of the Lumia Dev Journal, we shared survey results from the recent Gadget Vault UI/UX improvements, as well as our thoughts and future plans for character improvements.
We believe that for difficult and sensitive topics, taking time to prepare, and sharing detailed, organized thoughts with you is the best approach.
The next Dev Journal will cover Trolling and Smurfing, followed by fundamental structural topics such as Skill Haste, Healing Reduction, Farming, and Multi-weapon Characters.
Through consistent communication and by listening to your feedback, we will continue working to regain the trust that was lost.
As always, thank you for sticking with us.
