In case you’ve missed it, balance has been a big community topic lately. It’s no surprise considering an entirely brand-new classline was released late last year, with 22 celestial classes following it last month.
As a reminder, we decided to hold off on any balance adjustments in Orna 3.4/Aethric 1.1 so that we could tackle additional balance concerns in the newer celestial content as well.
With conversations aplenty this time around, many members of the community have been sharing their experiences with, against, and alongside other classlines. This is great community banter, and we definitely encourage it (while productive), but has led to many difficulties getting our finger on the subjective pulse regarding game balance this time around.
For the 2023 H1 Balance & Mechanics Patch, we’ll be taking a much more objective approach – ensuring all classes are within a state that they can consume game content at an equal difficulty, but perhaps not all types of content as efficiently. This aligns with Orna’s baseline design choices, wherein:
- All classes should have equal viability for player progression
- Not all classes should be able to consume all types of content as efficiently (ie: dungeon clearing vs raid slaying vs horde content vs PvP)
- Some classes will trade damage for defensive options
- Some classes will trade defensive options for damage potential
To ensure we are at a state where we can meet these design goals, while remaining relatively objective, while leaving room for subjective experiences via player feedback, we’ve been hard at work at developing a new live metric – calculated with realtime player data – named the Class Content Score (CCS). The CCS allows us to grade character classes in a way that:
- Buckets class content performance by activity type (Dungeon vs PvP vs Raid damage limit reaching, etc)
- Applies studio determined weights to these buckets, depending on the importance of the particular content and reward
- Provides a final score (the CCS) for the class, tallied from the individual activity type scores
From here, we can apply a CCS target (for example, at 90% of the highest CCS value). This target value allows us to:
- Apply nerfs, where appropriate to the extreme outliers (for example, Summoners in PvP at the moment)
- Buff all other classes in specific areas to bring them to the desired CCS
- Increase base endgame difficulty to meet the new equalized CCS values
Of course, we’ve always used data to inform and back up our decisions — the CCS is a solely new, more intelligent, more realtime metric that will further improve our ability to look at classes holistically.
At this time, we have over 10 million samples of class performance activity across many different in-game activities. I believe this sample size gives us great opportunity to talk about, publicize, and justify the goals of any upcoming balance adjustments.
Where we’re at
Let’s take a moment to examine some of the larger offenders from an objective standpoint. This post will not go into further detail about any exact upcoming changes, but rather will establish the themes that we will be tackling.
Although specializations, followers, gear, and more are also within our realm of analysis, we’ll keep this document focused on top-level class data.
Tier 9: Overall Class Performance
For the most part, we see the base Tier 9 classes within the acceptable range of variance when looking at their ability to consume content. The Tier 9 Celestials then pull ahead, particularly due to their ability to clear Towers well, and the Summoner line’s current PvP prowess. If we dive into PvP, we do see Summoner’s pull ahead by a large margin.
Tier 9: PvP Class Performance
This theme of Summoners in PvP is well documented by the community, and does justify our goal of reducing some of its defensive abilities in PvP only. Remember, this patch will make some PvP and PvE separations:
Tier 9: PvE Class Performance
Looking at the PvE side, we do see less variance with class performance, with dungeon clearing as an example:
And Tower battles as well:
Tier 7/8: PvE Class Performance
In the middle tiers, we do see the summoner classline pull ahead in PvE, mainly due to their safety when raiding:
Tier 10 Class Performance
Looking at overall Tier 10 class performance, we do see some familiar trends.
Taking the lead on content performance, we are seeing a top 5 of: Gilgamesh Hercules, Beowulf Hydrus, Heretic Ara, Realmshifter Corvus, and Deity Ara.
And at the bottom, we are seeing the following classes beginning to greatly trail behind: Realmshifter Dorado, Grand Summoner Hydrus, Beowulf, Realmshifter
When looking at some PvE data, we do see a few of these classes begin to trail behind of the others, starting with dungeon clearing:
But, strangely enough, most classes are clearing Towers with little variance:
On the PvP offense side, we see less variance in the classes, but the Hydrus’s start to shine more:
And on the top end of PvP defense, we see the summoner classline’s true problem – it’s a nightmare to deal with!
The main, non-exhaustive list of goals with 2023’s balance patches are below. There will be many more changes, but the goal of this article is to highlight the higher level themes, based around the data above, intersected with player feedback.
- Bring down the summoner classline’s dominance in PvP (while keeping it viable in PvE)
- Bring down the summoner’s classline’s earlier dominance in PvE (without affecting its current T10 state)
- Reduce the interactions that can affect Ultima (ie: Bulwark, Elysian Balance), without affecting its mainstay purpose as an endgame, sought-after ability
- Further improve both Realmshifter and Beowulf’s abilities in PvE content
- Further improve some of the underperforming Celestial Classes (Realmshifter Dorado, Beowulf Auriga, Grand Summoner Hydrus, and so on…)
- Address the known balance issues with Titan Augments and Celestial Weapons
- Improve the one-shot meta in PvP areas that matter most (Kingdom wars), without affecting ability to farm the Arena
- Bring all character class’s CCS to levels in an acceptable level of variance