Warcraft Logs 37LT6jAbWXFKqv8x · Heroic

The Coiled Altar, 20 pulls,
and a very long drop.

Two bosses, three stages and an intermission, on the Piian progression night of 12 September. This is what the orbs cost, what the shield demanded, and where the floor ran out.

The Venomous Abyss · Heroic 20 pulls, 14 reached stage three Killed on pull 20 in 7:57 open the log

The night, pull by pull

Bar height is how much boss was left when the pull ended, so shorter is better. Colour is the furthest stage reached. The stage timings barely move: stage one runs 123s, stage two 126s and the intermission a fixed 38s in every pull that got there. The only variable is stage three, and how much of it you survive.

Best wipe of the night was 4.82% on pull 18; the three closest were 4.82%, 8.21% and 8.23%.

The orbs, and what they cost

Poison orbs are carried and dropped into the tank frontal, which detonates everything sitting in it. The carry is a flat 5 seconds: pick it up, get it placed, it goes off. Each orb caught by the cleave stamps one stack of the burst DoT on the whole raid, so the stack count on a frontal is the orb count. 545 orbs were picked up across the night and 491 went off in a cleave, a median of 4 per frontal and a high of 8. Only four of 140 cleaves went off empty.

One mark per cleave, left to right in pull order. The pulsing per-orb aura dealt 1.63bn across the night and the detonation DoT 1.95bn. They are the two largest damage sources in the log, and both are self-inflicted by design.

Who carried

18 different people picked up at least one orb, but Liimage, Piiaan and Vesemirr handled 67% of them. Liimage alone carried 150.

Stage one against stage three

The job concentrates in stage one and spreads out in stage three, where several people who never touched an orb early pick one up late.

The shield, and fifteen seconds

Malacrass wraps himself in an absorb and starts a 15-second channel. Break the shield and the channel dies with it; miss and it resolves on the raid. The shield is a flat 18.6M every single time, not a percentage, so it is a fixed gate of about 1.2M raid damage per second held for 15 seconds.

29 of 36 shields broke, at a median of 9.7s and a best of 7.9s. The seven that did not killed 56 people between them.

Who broke the shield

Every hit that lands on the absorb counts towards breaking it, so this is damage onto Malacrass inside the shield windows only, not damage for the night. 18.6M has to come from somewhere, and with 25% of it coming from Indris, Vokster and Frdrschott the gate is more or less three people wide.

Show
 
624.7M onto the shield across 36 windows. Across all pulls the number on the right is the average per window, which is the useful figure since not everyone was alive for all of them. On a single pull it is that pull's total.

The intermission is a resurrection race

Stage one burns the serpent from full to nothing on a 274.9M health pool. In the intermission Malacrass shields himself to near-immunity, binds what is left, and hauls it back up on a new and far larger 700.9M pool: 14.0M a second for 34s. Whatever he has climbed to when the window closes is the handicap you carry into stage three. The kill held him to 35%; the worst pull let him reach 100%.

The ghosts decide most of that gap, rather than the damage. About 41 of them walk at the serpent across the window, and each one ends one of two ways. Body-block it and it dies, at the cost of a raid-wide blast worth about 1.3M: 530 went that way across the night. Let it reach him and it heals 21.0M instead. 69 went that way, putting 1.37bn back into the resurrection. One ghost through is worth 3% of the stage-three pool, so blocking costs about a twentieth of what letting one past does. This is also where the lust goes, and on 15 of the pulls that reached it the haste landed between -1.5s and +9s of the window opening.

Each pull's ghosts, blocked against let through. The two add up to about the same number every time, which is what makes it a straight trade rather than a damage race.
Damage into the serpent during the intermission, as a multiple of that pull's own stage one rate. Median 2.17x; the kill managed 3.19x, the best of the night.

The three strongest burn windows of the night are the kill and the two closest wipes.

When stage two handed over

Stage two ends on a damage threshold, not a clock: every pull crosses after 167.5M into Malacrass, identical to three figures across the night. So the handover time is entirely a choice about how fast you feed it. 7 pulls spent cooldowns and crossed early at a median 4:01 (1.4M/s into Malacrass through stage two); 9 held damage back and crossed at 4:18 (1.2M/s). Crossing early buys more stage three before the enrage, but those cooldowns are then not up for the burn window; holding back banks them for the burn and costs stage three time. On this night the held pulls burned harder, a median 2.28x against 1.88x, and both the kill and the best wipe sit in that group. That is one night with the two strategies unevenly sampled, so treat it as a lead rather than a verdict.

Either way the fight is being finished at the edge of the raid's own ceiling. The kill took 7:57, and only two of 20 pulls lasted longer than eight minutes at all. The longest reached 8:48 with no sign of a damage ramp, so this log cannot say where the enrage actually sits.

Across the bottom, when stage two handed over. Up the side, how much of the serpent had come back when the burn window closed. That is the handicap carried into stage three, so lower is better. The dashed line is the 4:10 split this page uses to call a handover early or late: a label, not a mechanic, since the real gate is the 167.5M damage threshold.

Who burns the window

Damage into the serpent during the intermissions only, totalled across the 577.8s of burn window the night contained. The bar is the damage; the multiplier is that player's own intermission rate against their own stage one rate, which is the honest way to ask who actually lifts when the amp and the haste are up. Tuari put in the most at 248.1M, but Liimage has the steepest lift at 3.77x.

Show
 
Healers sit near 0.9x by definition. They are here so the shares add up rather than because the chart is about them.

The edge

Going off the platform logs like nothing else in the game: no killing ability, no killer, no damage event. 58 of them happened while the pull was still winnable, which is 22.7% of the 256 deaths that cost anything, comfortably the largest single cause on the night. A further 132 came once the pull was already going down, which is a different thing entirely and is counted separately below.

Worst was pull 17, which lost 7 people over the side.

What put them there

Two mechanics own almost all of it. The knockback fires the instant Malacrass reaches zero, the same moment stage two hands over, and kills 4.5s later every time: 32 of these. A ghost had caught 22 more. Malacrass's own unavoidable Dreadmarch accounts for 2, and 2 are left unexplained.

Answering a mind control

Both mind controls log as the same debuff, which hides the difference that matters. 299 applications follow one of Malacrass's 67 casts and land on several people at once: that one is unavoidable. The other 107 land on a single person with no cast anywhere near, always as their fixation ends. That is a ghost catching somebody, and it is the one you can prevent. All 20 raiders were controlled at some point. Of the 107 caught, 43 were dead within thirty seconds and 39 went over the edge. The answer was 70 targeted stops, 43 of them grips, and 148 ground-targeted ones.

The ghost picks again. 622 fixations arrive in a spawn wave, but 127 land off-wave, in ones and twos, a median of 24s after the last wave. Those are a ghost choosing a new target, and they follow somebody dying 38% of the time against 15% for a wave fixation. That is suggestive rather than settled, since most off-wave picks follow no death at all.

Every possession makes two more ghosts. Dreadmarch is an absorb shield: it does not expire and it cannot be stunned away. The raid shoots it off, and crowd control is what keeps the victim away from the edge while that happens. Whichever way it ends, two Manifestations of Dread emerge from the possessed player, at 2 new fixations each, arriving a median 3.3s later. A four-person cast becomes eight ghosts; one person caught becomes two, who can catch two more. The pulls that lost the intermission took a chain of mind controls rather than one bad one.

The shield holds 494,352, and the raid broke 345 of 363 of them. The median possession absorbed 494,351, the shield consumed to the last point. Most of that damage is incidental, whatever was already cleaving; Liimage, Ørnin and Thila put the most in. The boss that applied the possession does not help break it.

On the counting. Three different things log identically to a fall: a genuine mistake, the wipe reset despawning the raid, and somebody jumping on purpose to reset a lost pull. They are separated by how much of the raid was already dead and by whether the pull survived. 132 falls fell the wrong side of that line, 108 of them with no mechanic on the victim at all, and none of those are counted against anybody. A mind control counts as the cause if it was on the victim when they died or broke within 5 seconds of it, because the shield coming down does not save somebody already over the edge. The ghosts themselves are destroyed by the boss frontal rather than killed, so they never register a death and cannot be counted directly. The builder's README has the thresholds and why they sit where they do.

Gloombomb, and the three souls

Malacrass bombs exactly three players a cast, 78 times across the night, and the blast takes a median 34% of a health bar from everyone it catches, which is more people than it targets. Everyone it damages is left Gravebound at three stacks, and so is whoever is standing in the frontal, which is why the tanks are the most-bound players on the roster without ever being selected. Each stack is a soul to collect and each collection costs about 10% of health: 965 of those ticks across the night.

Leaving one uncollected is the expensive mistake. The punishment hit lands at a median 75% of a health bar. There were 22 of them all night, and the damage split is clean enough to tell the two apart with nothing in between. Of the 28 deaths Gravebound owns, 19 were somebody leaving a soul and 9 were somebody collecting theirs while already too low.

Bar is how often each player was left Gravebound; the solid part is how much of that came from being picked for the bomb. Numbers on the right are souls collected and deaths.

Defensives and consumables

The fight names its own pressure: each shield channel and each intermission, 52 of them across the night. 902 major cooldowns were pressed and 278 of those landed inside one. The rest went out on the ordinary chip damage, which on this boss is most of it. 225 healthstones and 103 potions went down, at a median 36% health. The two tanks (Fleshcrusher and Vesemirr) are left out: they press mitigation on cooldown as rotation, which buries the question this section is actually asking.

Bar is major cooldowns, darker where they landed in a shield window or the intermission. Numbers on the right are healthstones and potions.

Health when the stone went down

Median health at the moment of use, lowest first. A stone pressed in the twenties is a stone pressed a beat late, since the fight kills from roughly a third of a health bar on a bad Nightfall.

Externals given

82 cooldowns handed to a named teammate. Cast on yourself it is a personal save rather than an external, and it is counted as one, which is why the mistweaver's total is smaller here than the raw Life Cocoon count.

Raid-wide cooldowns

185 of them, a separate thing from an external: nobody in particular receives one. They log against the environment rather than a player, which is how the two are told apart.

On the spell list. Defensives are counted from a fixed roster of spell names, so a spec whose button is missing from it reads as zero rather than as missing data. Throughput and utility casts are left out, and so are buttons that only mitigate for the other specialisation, such as Metamorphosis for a havoc demon hunter. If a count looks wrong, check the list before concluding anything about the player.