Erel Bugs - PTS v.269

5th job Enhancement Nodes that grant +120% Final Damage are not registering the upgraded SHINE versions of their target skills in the psdSkill (passive skill data) property.

<dir name="psdSkill">
  <dir name="181001000" /> <!-- Spear of Lugh (1st job) -->
  <dir name="181101003" /> <!-- Enhanced Spear of Lugh I -->
  <dir name="181111002" /> <!-- Enhanced Spear of Lugh II -->
  <dir name="181121000" /> <!-- Enhanced Spear of Lugh III -->
</dir>

<dir name="psdSkill">
  <dir name="181101000" /> <!-- Fury of Roan -->
  <dir name="181101001" /> <!-- Second Fury of Roan; carries values for [Shining Sentinel].
                                Has 0 animation (used when chained with Lugh). -->
  <dir name="181121004" /> <!-- Enhanced Fury of Roan I -->
  <dir name="181121005" /> <!-- Enhanced Fury of Roan I duplicate; 0 animation chain variant. -->
</dir>

<dir name="psdSkill">
  <dir name="181121001" /> <!-- Sting of Roan; 0 animation chain variant. -->
  <dir name="181121002" /> <!-- Sting of Roan. -->
</dir>

When the player unlocks SHINE Fury of Roan and SHINE Sting of Roan, the Shining Sentinel sub-skills they trigger should be upgraded to a stronger version. They are not.

Intended values (level 30):

Reality: only one skill exists for each, and it never changes.

• 181101002 - Shining Sentinel (Fury of Roan)
• 181121003 - Shining Sentinel (Sting of Roan)

<dir name="181101000"> <!-- Fury of Roan -->
  <dir name="extraSkillInfo"> <!-- Skills cast when this is used -->
    <dir name="0">
      <int32 name="skill" value="181101002" /> <!-- Shining Sentinel (Fury of Roan) -->
      <int32 name="delay" value="0" />
    </dir>
    <dir name="1">
      <int32 name="skill" value="400011171" /> <!-- Sentinel Captain: Fury of Roan -->
      <int32 name="delay" value="0" />
    </dir>
  </dir>
</dir>

<dir name="181141002"> <!-- SHINE Fury of Roan -->
  <dir name="extraSkillInfo">
    <dir name="0">
      <int32 name="skill" value="181101002" /> <!-- Shining Sentinel (Fury of Roan) - SAME ID as above -->
      <int32 name="delay" value="0" />
    </dir>
    <dir name="1">
      <int32 name="skill" value="400011174" /> <!-- Sentinel Captain: SHINE Fury of Roan -->
      <int32 name="delay" value="0" />
    </dir>
  </dir>
</dir>

<dir name="181101002">
  <dir name="common">
    <int32 name="maxLevel" value="20" />
    <string name="damage" value="50+5*x" />        <!-- (50 + 5*20) = 150%; no path to 200% -->
    <string name="mobCount" value="3" />
    <string name="attackCount" value="4" />        <!-- Hardcoded 4 hits; no path to 6 -->
    <vector name="lt" value="-617, -490" />
    <vector name="rb" value="721, 170" />
    <string name="time" value="10" />
    <string name="attackDelay" value="2900" />
    <string name="s" value="80" />
  </dir>
</dir>

The Eternal Guardian tooltip states that Captain's Fury of Roan and Captain's Sting of Roan should each hit 8 enemies, 13 times, for 556% at level 30 (formula 256 + 10*x).

The actual data is wrong for both - and they're wrong in different ways, suggesting their values got crossed.

<dir name="400011172"> <!-- Sentinel Captain: Sting of Roan -->
  <dir name="common">
    <int32 name="maxLevel" value="30" />
    <string name="cooltime" value="10" />
    <string name="damage" value="256+6*x" />   <!-- 436% at lvl 30; tooltip expects 256+10*x = 556% -->
    <string name="attackCount" value="13" />   <!-- Correct -->
    <string name="mobCount" value="6" />       <!-- Incorrect; should be 8 -->
    <vector name="lt" value="-652, -467" />
    <vector name="rb" value="445, 84" />
    <string name="x" value="600" />
    <string name="ignoreMobpdpR" value="100" />
    <string name="u" value="-60" />
  </dir>
</dir>

<dir name="400011171"> <!-- Sentinel Captain: Fury of Roan -->
  <dir name="common">
    <int32 name="maxLevel" value="30" />
    <string name="cooltime" value="10" />
    <string name="damage" value="256+6*x" />   <!-- 436% at lvl 30; tooltip expects 256+10*x = 556% -->
    <string name="attackCount" value="6" />    <!-- Incorrect; should be 13 -->
    <string name="mobCount" value="8" />       <!-- Correct -->
    <vector name="lt" value="-652, -467" />
    <vector name="rb" value="445, 84" />
    <string name="x" value="780" />
    <string name="ignoreMobpdpR" value="100" />
    <string name="u" value="-60" />
  </dir>
</dir>

Sentinel Rise is dealing less damage than expected, even though the values in the files appear correct. There is likely a hidden interaction at play.

Test conditions:

• Level 16 Erda Shower → 690% per hit.
• Shining Sentinels expected: 690% (150 base + 160%p + 180%p + 200%p).
• As expected, Erda Shower and Shining Sentinel deal roughly equal damage.

The problem:

• Sentinel Rise should deal 8 summons × 12 hits × 940% per hit.
• In the reference test below, Erda Link “Sentinel Rise Boost” was activated, which adds +11% Final Damage, so each hit should land at 940% × 1.11.
• Maximum per-hit damage on Sentinel Rise should therefore far exceed Shining Sentinel - but it lands lower.

Reference image: BA screenshot

Status: root cause unknown. Numbers look right in the files; in-game damage is lower than the formula predicts.

Radiant Sentinel is an attack summoned when attacking with Spear of Lugh, granted by the 5th job passive Eternal Light.

The tooltip of Eternal Light states that Radiant Sentinel should attack 6 times for 250% damage at level 30, using the formula 130 + 4 * level.

A battle analysis (see image below) of the skill makes it immediately clear those numbers are wrong. Inspecting the skill data confirms it:

<dir name="400011166"> <!-- Radiant Sentinel (Eternal Light) -->
  <dir name="common">
    <int32 name="maxLevel" value="30" />
    <string name="damage" value="600+6*x" />   <!-- Actual damage the skill applies -->
    <string name="time" value="10" />
    <string name="mobCount" value="6" />       <!-- Mobs hit per cast -->
    <string name="attackCount" value="8" />    <!-- Hits per cast -->
    <vector name="lt" value="-617, -490" />
    <vector name="rb" value="721, 170" />
    <string name="attackDelay" value="2900" />
    <string name="s" value="90" />
    <string name="cooltime" value="10" />
  </dir>
</dir>

Notably, attackCount and mobCount have swapped values vs. what the Eternal Light tooltip describes (tooltip implies 6 hits, 8 mobs; data shows 8 hits, 6 mobs).

While the damage value (780% at level 30) looks correct when compared to other skills, the skill does not hit the correct number of times in-game.

I have no idea what causes this - nothing in the data that I've been able to discern explains this.

As with all 2-part skills, Fall of Melin is split across two skill IDs:

• 181141500 - Part 1: 14 hits × 70 activations, 254 + 7 * level
• 181141501 - Part 2: 14 hits × 40 activations, 219 + 7 * level

<dir name="181141500"> <!-- Tooltip shown for Fall of Melin -->
  <dir name="common">
    <int32 name="maxLevel" value="30" />
    <string name="mpCon" value="1200" />
    <string name="mobCount" value="15" />
    <string name="attackCount" value="14" />   <!-- Part 1 attack count -->
    <string name="damage" value="254+7*x" />   <!-- Part 1 damage -->
    <string name="x" value="70" />             <!-- Part 1 activations -->
    <string name="y" value="219+7*x" />        <!-- Tooltip placeholder: Part 2 damage -->
    <string name="z" value="15" />             <!-- Tooltip placeholder: Part 2 mob count -->
    <string name="v" value="14" />             <!-- Tooltip placeholder: Part 2 attack count -->
    <string name="s" value="41" />             <!-- Tooltip placeholder: Part 2 activations -->
    <string name="cooltime" value="360" />
    <string name="updatableTime" value="7000" />
    <string name="ndTime" value="8440" />
    <string name="ignoreMobpdpR" value="log10(x)*20+log30(x)*30" />
    <string name="bdR" value="log20(x)*20+log30(x)*30" />
    <vector name="lt" value="-1200, -800" />
    <vector name="rb" value="1200, 800" />
  </dir>
</dir>

Tooltip template string:

MP Cost: #mpCon, Invincible during casting
Max Enemies Hit: #mobCount, Damage: #damage%, Number of Attacks: #attackCount, Blows: #x
Afterwards, Damage: #y%, Number of Attacks: #v, Blows: #s
Cooldown: #cooltime sec

<dir name="181141501">
  <dir name="common">
    <int32 name="maxLevel" value="30" />
    <string name="damage" value="219+7*x" />
    <string name="mobCount" value="15" />
    <string name="attackCount" value="14" />
    <string name="x" value="40" />             <!-- Part 2 activations -->
    <vector name="lt" value="-1200, -800" />
    <vector name="rb" value="1200, 800" />
  </dir>
</dir>

Note: the tooltip placeholder s = 41 in 181141500 (the tooltip) does not match 181141501.x = 40. The tooltip claims Part 2 activates 41 times, but the per-skill data says 40.

Testing in-game with a battle analysis reveals the actual behavior:

• Part 1 deals 966 hits (= 14 × 69 activations)
• Part 2 deals 574 hits (= 14 × 41 activations)
• Total: 1540 hits

The cause is the multiAttackInfo array, which is what the game actually uses to schedule hits. Each entry triggers one cast of the referenced skill ID (which then deals attackCount = 14 hits). The array is presumably generated automatically - and it generated 110 entries instead of 111 since part 2 has 1 fewer hit than intended. As a result, the auto-generation appears to have messed up the sequencing.

Truncated for clarity:

<dir name="multiAttackInfo">
  <dir name="0"><int32 name="attackTime" value="30"/><int32 name="x" value="181141500"/></dir>
  <dir name="1"><int32 name="attackTime" value="30"/><int32 name="x" value="181141500"/></dir>
  ... <!-- entries 2 through 67 all reference 181141500 -->
  <dir name="68">
    <int32 name="attackTime" value="30" />
    <int32 name="x" value="181141500" />
  </dir>
  <dir name="69">                                  <!-- SHOULD still be 181141500, since this is 0-indexed -->
    <int32 name="attackTime" value="2300" />
    <int32 name="x" value="181141501" />            <!-- but switches to part 2 here -->
  </dir>
  <dir name="70"><int32 name="attackTime" value="10"/><int32 name="x" value="181141501"/></dir>
  ... <!-- entries 71 through 108 all reference 181141501 -->
  <dir name="109"><int32 name="attackTime" value="10"/><int32 name="x" value="181141501"/></dir>
</dir>

What this tells us:

• 0–68 (69 entries) trigger Part 1 → 69 × 14 = 966 hits
• 69–109 (41 entries) trigger Part 2 → 41 × 14 = 574 hits
• Total: 110 entries, when there should be 111 (70 + 41).

The generator appears to have honored the tooltip placeholder s = 41 for Part 2's entry count, but came up one entry short overall. Part 2 got its intended 41 activations; Part 1 silently absorbed the missing one, ending up with 69 instead of 70.

Damage breakdown at level 30:

• Part 1: 254 + 7 × 30 = 464% per hit
• Part 2: 219 + 7 × 30 = 429% per hit

Origin is dealing roughly ~0.9% less damage than intended. Obviously the origin is very weak in general, but this bug does not have a significant impact.

Pretty self-explanatory. Female-character animations for Origin and Ascent exist in the game files but aren't being used - female Erel characters play the male animations for both skills.

Destruction of Roan hits one extra time when the Eternal Guardian buff is active.

<dir name="400011176"> <!-- Destruction of Roan -->
  <dir name="common">
    <int32 name="maxLevel" value="30" />
    <string name="mpCon" value="450" />
    <string name="damage" value="680+40*x" />
    <string name="attackCount" value="8" />     <!-- Hits per activation -->
    <string name="updatableTime" value="1800" /><!-- Hold duration in ms -->
    <string name="mobCount" value="15" />
    <string name="attackDelay" value="120" />   <!-- Delay between activations -->
    <string name="cooltime" value="120" />
    <vector name="lt" value="-540, -600" />
    <vector name="rb" value="99, 25" />
    <string name="s" value="40" />
    <string name="t" value="1.8" />
    <string name="u" value="300" />
  </dir>
</dir>

Math: with a 1800ms hold duration and 120ms delay between activations, the skill should activate
1800 / 120 = 15 times, each dealing 8 hits → 120 hits total. In normal use, that's exactly what we see.

However, when Eternal Guardian is active, Destruction of Roan instead hits 128 times (16 activations) - alongside the Sentinel Captain doing its own 128 hits. It's not clear from the visible data why the extra activation occurs.

Erel Light has five Rush Stones that should grant Skill Cooldown Reduction. Three are wired up correctly, but two are linked to the wrong stat. The name, description, and icon all say cooldown - but the actual stat being granted is something else.

Example of an incorrect stone - name, description, and icon all reference cooldown, but the applied stat is abnormalDamR (Abnormal Status Damage Rate):

<dir name="37">
  <int32 name="maxLevel" value="1" />
  <dir name="passive">
    <dir name="0">
      <dir name="0">  <!-- Value when inactive -->
        <int32 name="abnormalDamR" value="0" />  <!-- WRONG STAT -->
      </dir>
      <dir name="1">  <!-- Value when active -->
        <int32 name="abnormalDamR" value="1" />  <!-- WRONG STAT -->
      </dir>
    </dir>
  </dir>
  <string name="name" value="Skill Cooldown" />
  <string name="desc" value="Decreases skill cooldown by 1%." />
</dir>

Example of a correct stone - uses coolTimeR as expected:

<dir name="36">
  <int32 name="maxLevel" value="1" />
  <dir name="passive">
    <dir name="0">
      <dir name="0">
        <int32 name="coolTimeR" value="0" />
      </dir>
      <dir name="1">
        <int32 name="coolTimeR" value="1" />
      </dir>
    </dir>
  </dir>
  <string name="name" value="Skill Cooldown" />
  <string name="desc" value="Decreases skill cooldown by 1%." />
</dir>

The other broken stone is Rush Stone ID 41 - same issue, both should be granting coolTimeR instead of abnormalDamR.

SHINE Spear of Lugh's tooltip has a couple of issues. The big one: Stellar Strike - the follow-up attack it summons — does noticeably less damage in-game than the tooltip claims.

How the skills are wired: when you cast SHINE Spear of Lugh (181141000), its extraSkillInfo summons Eternal Light and Stellar Strike (181141001):

<dir name="181141000"> <!-- SHINE Spear of Lugh -->
  <dir name="common">
    <int32 name="maxLevel" value="30" />
    <string name="mpCon" value="20+2*d(x/5)" />
    <string name="damage" value="470+x*8+4" />  <!-- SHINE Spear of Lugh's own damage -->
    <string name="mobCount" value="8" />
    <string name="attackCount" value="8" />
    <vector name="lt" value="-482, -280" />
    <vector name="rb" value="20, 60" />
    <string name="y" value="9" />
    <string name="x" value="15" />
    <string name="s" value="565+x*2+10" />      <!-- Tooltip placeholder: Stellar Strike damage. 635% at lvl 30. -->
    <string name="u" value="10" />              <!-- Tooltip placeholder: Stellar Strike attack count -->
    <string name="a" value="6" />               <!-- Tooltip placeholder: Stellar Strike chase cut count -->
  </dir>
  <dir name="extraSkillInfo"> <!-- Skills cast when SHINE Spear of Lugh is used -->
    <dir name="0">
      <int32 name="skill" value="400011165" /> <!-- Eternal Light -->
      <int32 name="delay" value="0" />
    </dir>
    <dir name="1">
      <int32 name="skill" value="181141001" /> <!-- Stellar Strike -->
      <int32 name="delay" value="0" />
    </dir>
  </dir>
</dir>

11.1Tooltip Damage Formula vs Actual Skill DataMajor

The tooltip displays Stellar Strike's damage using placeholder s on SHINE Spear of Lugh (above), which evaluates to 565 + x*2 + 10 → 635% at level 30.

But the actual Stellar Strike skill (181141001) defines its damage as 320 + 8*x → 560% at level 30:

<dir name="181141001"> <!-- Stellar Strike -->
  <dir name="common">
    <int32 name="maxLevel" value="30" />
    <string name="attackCount" value="10" />
    <string name="mobCount" value="15" />
    <string name="damage" value="320+8*x" />    <!-- Actual damage. 560% at lvl 30. -->
    <vector name="lt" value="-784, -437" />
    <vector name="rb" value="84, 32" />
    <string name="updatableTime" value="2800" />
  </dir>
</dir>

Verification - compare Stellar Strike against Erda Shower (690% per hit at level 16).

Scenario	Stellar Strike damage	Expected gap vs Erda Shower (690%)
If tooltip is correct	635%	~8% (690 / 635)
If skill data is correct	560%	~23% (690 / 560)

Test conditions: 198 attacks each side (3 Stellar Strike procs × 66 hits and 33 Erda Shower casts × 6 hits). +20% Boss / +20% Damage Hyper Skills for Spear of Lugh were removed to make the math simple.

The observed gap (~21.5%) lines up almost exactly with Stellar Strike actually dealing 560%, not the tooltip's claimed 635%. The skill is under-tuned by roughly 12% relative to its tooltip.

MP Cost: #mpCon, Max Enemies Hit: #mobCount, Damage: #damage%, Number of Attacks: #attackCount
Upon using Spear of Lugh #y times, the Sentinel Captain is summoned and strikes.
#c[Blown]#: Max Enemies Hit: #x, Damage: #s%, Number of Attacks: #u, Number of Chase Cuts: #dummyStr