Voidscar Arena

This is new since the analysis was first written, and it is new because the tool changed. Arbiter now resolves every spell in a key against Blizzard's own client tables before it grades anything. Until then a cast counted as *stoppable* -- a cast an interrupt could have cancelled -- only if some log had already shown somebody kicking that exact spell. A cast nobody has ever kicked was invisible. Six of this key's spells sat in that blind spot, and both of Raj'kess the Spellstorm's are among them.

**Pull two, 2:19 to 3:36, is Raj'kess.** The party took 14,142,232 in it. Thundering Storm was 7,748,046 of that (54.8%) and Forked Lightning 6,098,313 (43.1%) -- **97.9% of the pull from two spells.** Fifteen casts: thirteen Forked Lightning on a 2.5 second cast bar and two Thundering Storm on a 3.0 second one. None was stopped. None was *attempted*: the last interrupt anybody pressed in this key was Doryhunky's Disrupt at 2:08, ten seconds before Raj'kess's first cast, and nothing was pressed again for the remaining 3 minutes 56 seconds of the run.

It was not a capacity problem. Four of you carry an interrupt, on 15, 15, 15 and 25 second cooldowns, across a 73 second fight. Scored against those cooldowns and the cast times, **all fifteen were reachable.**

And it did not stay in that pull. Hyporock spent **Aura Mastery at 3:23** on this damage. Its cooldown is 180 seconds, so it was due back at 6:23 and the key ended at 6:04. When Chitigoth's first Insidious Aura window opened at 5:11, the section above is right that Aura Mastery was genuinely unavailable and not a fair criticism -- but it was unavailable because it was spent on damage a kick removes.

**The caveat, and it matters.** This is the game's spell data, not proof from a log. Blizzard's interrupt table sets the interruptible bit on both spells; no log on this machine shows either one kicked, because nobody has tried. Lava Bolt, Demoralizing Shout and Shadowbolt Volley in the same key are the proven kind -- they were kicked here, 19 casts stopped of 50 that carried the bit. If a kick bounces off Raj'kess, that is worth knowing and this line is wrong.

**One smaller thing from the same source.** Melt Armor, from the Sycophantic Tarasek, is a **Magic** debuff -- Cleanse takes it -- and the spell data lists it as a stacking physical resistance reduction. It was cast at Doryhunky ten times and sat on him for 49.5 seconds of the wipe pull, never removed. Melee on him averaged 36,906 per swing while it was up against 30,781 while it was not, over 119 and 109 swings. Three of the four main attackers show the same 1.24x to 1.44x gap; the Feral Saberon, the largest sample, shows none, so treat 20% as the shape of it rather than the figure. Arbiter does not count this as a missed dispel, by design: a debuff that deals its target no direct damage is excluded, so an armour shred on the tank never reaches the healer's sheet.

This analysis was re-checked on 2026-09-06 against both logs after a dungeon guide contradicted parts of it. The guide was right about one mechanic, wrong about two, and the log settles all three.

**1. Fel Steps is not unavoidable.** The first pass called it a 300-yard pulse that ignores line of sight and told you to stop thinking about it. The radius is real, but the trigger is not passive: **Brutok only pulses while he is walking.** Cut both runs into half-second slots and the pattern is absolute -- across 109 slots in which Brutok did not move, Fel Steps fired **zero** times; in the slots where he moved it fired every 0.5 seconds. That makes it damage a tank can switch off, which is the opposite of what you were told.

**2. Nobody pulled Brutok.** The guide's explanation was that the tank yanked him from range and made him run. He did not. Brutok's first contact in this key is *Brutok hitting Arcane Mage with Fel Steps* -- he walked in on his own, because the party was standing in the middle of the arena with five packs alive. On the clear, the tank opened him deliberately with Death's Caress. Same for the Raging Raptor and the Abducted Drakonid: both engaged this party unprompted, and both were opened on purpose the day before.

**3. The turtle shields nothing.** The guide says Shell Guard protects nearby enemies, so pulling the turtle risked armouring Chitigoth. It did not happen. Shell Guard is cast by the Protective Turtle **on itself** -- the only target in either log -- and the turtle and Chitigoth were never closer than 20.6 yards. Chitigoth took 228,284 per second inside the shield windows and 245,976 outside them, which is the same number. The turtle also never swung at anybody in *either* run, including the clear where it absorbed 424 separate hits. Not attacking is just what it does.

The groups share GUID spawn-batch suffixes (DAE2C, DAE32, DAE3F, DAE4D, DAE51), and the run that cleared this arena has the same batches in the same order with the same compositions. That much looks like a scripted gauntlet. It is not. The first damage event on each pack settles it:

  4:49.3  Chitigoth              <- Doryhunky  Throw Glaive     (27 yd)
  5:03.0  Feral Saberon x2       <- Doryhunky  Infernal Strike  (he leapt in)
  5:16.1  Protective Turtle      <- Doryhunky  Throw Glaive     (37 yd)
  5:17.1  Longtooth Tuskarr      <- Doryhunky  Throw Glaive     (37 yd)
  5:30.9  Sycophantic Tarasek x2 <- Doryhunky  Immolation Aura  (8 yd, he walked to them)

Every pack responded 0.5 to 8 seconds AFTER being hit. Nothing in this arena force-aggros: on the clear, **thirteen** distinct enemies were each opened by the tank, one at a time, across 352 seconds -- Chitigoth at 2:17, the Raptor at 2:23, the Turtle at 3:16, Brutok at 3:28, the Drakonid at 4:27, the Krolusk at 4:50, Taz'Rah at 5:40. Only one mob in that entire arena moved first, and it was an Ethereal Shade six minutes in.

That also means the "five packs" in the first pass undercounted the room. This group died forty seconds into a sequence that runs six times longer. The three mobs that opened on *you* did so because you were standing in the middle of it with everything alive -- which is a pacing consequence, not a scripted one.

Five mechanisms were tested and excluded as the cause of the pacing. Kill rate on Chitigoth was identical for thirty seconds. Nothing died in this arena at all, so nothing was kill-gated. Chitigoth's health at the third pack's arrival was 52% here against 0.3% on the clear, so no health threshold. The two parties fought in the same place -- average positions **2.8 yards apart**. And no enemy healed: the largest rise in any enemy's health across the pull was 1.6%, sampling noise.

Brutok pulses Fel Steps every half second while he moves and never while he stands still. Measured in both runs:

                        Brutok's path   time engaged   Fel Steps damage
  this key                   104.8 yd         31.4 s          2,871,582
  the clear, 09-05           557.5 yd         74.1 s          4,428,029

So the mechanic is real and the tank owns it. But be careful what you conclude from that, because **the group that cleared walked him more than five times further and ate 54% more Fel Steps damage than you did, and cleared comfortably.** Smashing Charge, the other movement the guide warns about, accounts for 3% of the Fel Steps damage here and 11% on the clear. Fel Steps was not what killed this key.

What it is worth: 2,871,582 of the roughly 16.0M in this pull, or 18%, and most of it was Brutok walking to a party that had not gone to him. Holding him still once he arrives is free, and worth doing. Rebuilding the pull around it is not.

Taken from the spell data, not inferred from the log:

**Insidious Aura** (1250695, Chitigoth) -- 100 yard radius, *Ignore Line of Sight*, dispel type none, instant, no cast. It cannot be dodged, outranged, line-of-sighted, dispelled, stolen, purged or interrupted. It is a **buff Chitigoth puts on himself for exactly 10.0 seconds, every 20.4 seconds**, ticking every 2 seconds for about 290,000 a tick. Same clock in both runs, first application ~21 s after he is engaged. 7,728,440 here, and genuinely unavoidable.

**Savage Leap** (1267894, Feral Saberon) -- 100 yard leap, one target, flagged no-dodge and no-parry. Roughly 70% of it is a 5 second bleed, and **Blessing of Protection strips that bleed** -- proven in this very key at 0:31.5, where the bleed left Arcane Mage in the same millisecond as the cast. 3,975,848.

**Fel Steps** (1269284, Brutok) -- 2,871,582, and per above, it only fires while he walks.

**Ravenous Swarm** (1234833, Chitigoth) -- a swarm placed at a player's location. Walk out of it. 1,406,870.

The first pass said 80% of this was unpreventable by anyone. With Fel Steps reclassified the honest figure is about **73%** -- Insidious Aura plus the leap itself. The rest is the tank's positioning and one person moving.

None of that changes the shape of the problem: incoming damage is close to a linear function of how many packs are alive, and the two packs that were pure liability added **4,057,860 of damage (16%)** and **55,533,824 of health** this party was never going to remove.

Running damage taken minus effective healing, against the party's combined 4,914,935 health:

  5:10.9  Insidious Aura window 1 opens      net    +9,538   (dead even)
  5:16.1  tank glaives pack 3 from 37 yd     net  -378,869   ( -8%)
  5:20.9  window 1 closes                    net  -930,964   (-19%)
  5:21.0  Brutok walks in on his own         net -1,866,329  (-38%)
  5:22.5  Arms Warrior dies
  5:23.3  Hyporock dies
  5:27.5  pack 5 pulled                      net -3,096,048  (-63%)
  5:33.2  Arcane Mage dies
  5:33.4  Unholy DK dies                     net -3,482,015  (-71%)

The party walked into the first Insidious Aura window at **exactly break-even and full health**. Twelve seconds later two of them were dead. Across the arena the group took 429,763 per second and healed 274,810 -- 64% coverage. The clear took 383,519 and healed 317,490, which is 83%, and its ratio rose above 1.0 in three of six ten-second bins because it had lulls between packs. This run's ratio went 1.07, 0.81, 0.50, 0.47, 0.55 and never recovered.

And no party-wide mitigation was pressed in either aura window. Available and unused: **Rallying Cry, Anti-Magic Zone, Divine Protection, Lay on Hands, Ice Block, Greater Invisibility, Alter Time, Die by the Sword, Spell Reflection, Storm Bolt, Icebound Fortitude**. Genuinely on cooldown and not a fair criticism: Aura Mastery, Darkness, Blessing of Sacrifice, and Time Warp, which came back up at 5:24.4 -- four seconds after the first death.

C- is 61. throughput 54 means you covered 67% of the 40,963,170 that needed an outside heal, and this key sat inside the corpus's normal demand range, so unlike the +15 an hour earlier that number is a measurement rather than a floor. In the arena itself you healed **156,594 per second at 22% overheal**, against 241,167 at 31% overheal in the run that cleared. The lower overheal is the tell: you were not wasting output, there was simply more damage than one healer covers.

Your death is four and a bit seconds long and worth reading tick by tick. Two Insidious Aura pulses took you 100% to 40%. A Savage Leap took you to 24%. Then four ticks of Fel Steps -- 66,099, 60,703, 48,563, 60,703 -- and a second Savage Leap for 125,937 finished you. You healed yourself 199,838 during it and were still casting Eternal Flame and Holy Shock on other people the whole way down.

Those four Fel Steps ticks in 1.5 seconds were Brutok walking across the room at you. Not your call to make, but worth knowing they had a source that could have been switched off.

**Divine Protection was up, and had been for 37 seconds.** An earlier draft said 96 -- that was time since you last pressed it, at 3:46, and the spell data now on file puts the cooldown at 60 seconds, so it came back at 4:46. Twenty percent off that window is about 242,000, which is nearly twice the killing blow. It is the same finding as the +15 and it is the one thing on your sheet that is yours.

The first pass closed by asking Warcraft Logs for Voidscar Arena's par time, which Arbiter had as "unknown, bounded 29:19-35:20". It did not need Warcraft Logs.

`CHALLENGE_MODE_END` field 6 is the run's Mythic+ score, and at every key level the timed and untimed runs sit in two clusters tens of points apart -- at +14, 395.9-399.4 against 316.1-319.4. That split reproduces both depletions independently confirmed on 2026-08-30, so it can be trusted to label the rest. Applied to all 61 scored runs in the archive it narrows Voidscar to **between 29:19.8 and 30:30.1**, and 30:00 is the only whole minute in that window.

Two consequences. Voidscar upgrades can be reported from now on. And the Voidscar +14 on 2026-09-05 -- the run this analysis keeps calling "the clear" -- finished in 30:30.1 and was therefore **depleted, not timed**. It is still the better run by a wide margin; it just did not beat the timer either.