Warcraft Logs pGFQH28M3dnbPjhW · Heroic

Ula'tek, 25 pulls,
one heart to open.

Damage, deaths and cooldown use, pull by pull. Piian's Mid-Pressure night, 13 September.

The Venomous Abyss · Heroic20 raiders Best pull: 34.03% remaining open the log

The night, pull by pull

Boss health remaining at the wipe. Bar colour is the number of heart windows the pull reached. Reaching a second window means getting past the 51% shelf.

one heart windowtwo heart windows
Lower is further. Bar labels show pull length in seconds.

The burn windows

Rage of the Shackled exposed the Venomous Heart 27 times. Every one ran the full 20 seconds; no window was lost early. The boss stays cleavable through an exposure, so both targets count: the heart takes double damage, Ula'tek takes normal. Across the night the split is 76% heart, 24% boss. First windows landed 208.7M on average, second windows 83.4M. The 60% drop from the first window to the second tracks lust, which went out 3 to 17 seconds before the first exposure on every pull that reached one and is not up again for the second.

window 1: heartwindow 1: Ula'tekwindow 2: heartwindow 2: Ula'tek
Damage in millions per 20-second exposure, solid onto the heart and pale onto Ula'tek. Adds are excluded.

Inside the window

A running total of burn damage, heart plus Ula'tek, second by second, each window laid end to end. The target switch isolates one or the other. Slope is output; the gap between two lines at a given second is the difference between those players. Resetting the total at each window compares a player's windows against each other, with window 1 carrying the lust.

Target
Running total
Players
Cumulative burn damage. Each divider marks the end of one window and the start of the next.

Best heart windows

Each player's highest window 1, highest window 2, and highest combined pull, with the pull each came from. Third windows are excluded, since only the deepest pulls have one and it runs short. The point is the pull number: open that pull in the chart above, or in the log, to see what the player did in it. Theoretical best adds the two best windows together even when they came from different pulls, so it is a ceiling nobody hit rather than a pull that happened. The totals row adds each column down the roster, on the same terms.

PlayerSpecBest window 1Best window 2 Best pull, both windows Theoretical best

Who burns the window

Damage per player across the 537 seconds of exposure: solid onto the heart at double damage, pale into Ula'tek, beside their share of the night's damage. A positive delta means a bigger share of the burn than of the night, which is burst rather than sustained output. Heart share across the damage dealers runs 87% down to 68%.

Bars are total burn damage, heart first then Ula'tek; the number beside each is damage per second of exposure.
PlayerSpecBurn damageHeartUla'tekOn heartBurn DPSShareNight shareΔ

Consistency in the heart windows

One dot per window, placed at what that window was worth against the player's own best window of the same number. A player whose dots sit together near the right repeats their ceiling; dots strung out to the left mean the best pull was the exception. Measuring each window against the best of its own number keeps the lust out of it, since window 1 and window 2 are not comparable on this fight. Typical is the median of a player's dots, not the average, so one bad pull does not set it.

Windows worth under a fifth of a player's own best are counted as missed rather than low and left out of the median: those are pulls they were dead or absent for, which is worth knowing but is not the same as uneven output. Healers are left out of this chart, since their window damage is small enough that ordinary noise swamps the signal.

window 1window 2tick = median
Every dot is one window on one pull. Hover a dot for the pull number.

Damage onto the boss

Ula'tek, Gore Rattle and the Venomous Heart share one health bar, so progress on the boss is the sum of the three. This counts every pull start to finish, not only the 537 seconds of exposure the burn charts measure. 11.63B went into the pool across 25 pulls: 57% Ula'tek, 32% heart, 10.8% Gore Rattle. Indris led it with 939M. The order differs from the burn windows: Alfi three places higher and Emberforge three places higher than the windows put them.

Ula'tekVenomous HeartGore Rattle
Raid damage into the pool, per pull. Click a bar to load that pull below.
Ula'tekVenomous HeartGore Rattle
Each bar is one player's whole night, split by target. Beside it: total damage, then damage per second of pull time.
PlayerSpecUla'tekHeartGore RattleTotalDPSShare

Caustic Waves

195 waves went out; 92 of them clipped somebody, and 9 of those hits killed. Tank contacts are the cost of holding Ula'tek in melee; the other 65 are the avoidable ones.

The boss throws them in two phases. Phase one is two volleys of about four waves a pull, 192 casts across the night, catching non-tanks 63 times, 0.33 a cast. Phase two is 3 casts on the one pull that reached it, 2 hits, 0.67 a cast. Deaths split 8 to 1. Phase one accounts for 63 of the 65 non-tank hits and 8 of the 9 deaths.

hit in phase onehit in phase twohits that killedhatched = tank
One hit = one wave contact, ticks within two seconds grouped. Bars run phase one then phase two; the pair in brackets is the same split as numbers. Damage is after absorbs and mitigation.
PullLengthBoss leftP1 castP2 castP1 hitsP2 hitsHits (tank)Non‑tank per waveVipersBurn dmg

Left and right

The raid splits ten seconds after the first heart closes, into two groups either side of the platform, thousands of coordinate units apart. The phase opens at 160 to 175s on every pull and runs 30 to 105 seconds; the boss is untargetable until about 110s in, which is the deadline for getting back. Across the 21 pulls that reached it the sides were west 10 (1 tank, 2 healers, 7 damage) and east 10 (1 tank, 2 healers, 7 damage). One player moved to the other side partway through, at pull 4, so each pull is read on its own. Four pulls had somebody on the far side of their own assignment (pull 3, pull 5, pull 11 and pull 24), which is what a wrong turn looks like. Deaths inside the phase are listed per player.

The sides are not comparable on damage this night. West is handed 227 adds across the night against east's 113, 11 a pull against 5. Most of the difference is one add: west takes 120 Blightscale Rawlings to east's 62. The 56.9/43.1 phase damage split measures that workload, not throughput: night-wide the two sides are 51.2/48.8. Per add the order reverses, 6.4M west against 9.7M east, because east's are the larger targets. Compare them on deaths and on getting back to the middle instead.

west: Fleshcrusher's sideeast: Vesemirr's side
Damage per second inside the phase. The sides are handed different amounts of work, 11 adds a pull west against 5 east, so most of any gap here is workload rather than throughput.
west, first to last backeast, first to last backboss takes damage again
How to read a bar first one back median: this is the one judged last one back boss can be damaged again
Each bar runs from the first player of that side back in the middle to the last, notched at the median. Everything left of the red mark is free, and being back early buys nothing. A side counts as late when its median is past the mark: one healer trailing behind to res somebody is not the side being late. Only the five pulls where both sides walked back at all; on the other sixteen the raid died out there.
PullPhaseWest DPSEast DPSWest HPSEast HPSWest back (median)East back (median)Boss backDeaths W/E

Reading the sides: a player's side is their median x while the two groups are apart, so a few samples on the far side in the opening seconds are the walk out, not a wrong turn. Side counts per pull ran from 7 to 11.

Sides are read from position, not assignment: every cast event carries the caster's coordinates, so each player's median x while the two clusters are apart says which side they stood on. West and east are the log's axis, not the screen: west is whichever side Fleshcrusher tanks.

Damage taken, second by second

Incoming damage on the raid across a pull, tanks excluded, since their intake is the job and would bury everyone else. The axis is fixed at 454 seconds, the longest attempt, so every pull reads against the same clock. Drag the handles to a stretch of the fight and everything below recalculates for it: damage taken, how much was mitigated, and every cooldown, stone and potion spent inside it.

Window
PlayerDamage takenReducedAbsorbedMajorsMinorExternalsStonesPotionsDeaths
Fleshcrusher and Vesemirr are left out of this section entirely, chart and table alike; the 18 non-tanks are what you see. Damage taken is health actually lost. Every hit in the log carries what it was worth before mitigation, and the difference splits two ways. Reduced is armour, versatility, Stagger, class passives and any damage-reduction cooldown that happened to be up; absorbed is shields, from healers or the player's own. The log does not separate passive reduction from a pressed cooldown, so read these as "how much of the swing never reached them", not as a defensive-usage score: for the 18 non-tanks across the night it is 26% reduced and 10% absorbed, and most of that 24% is passive. The curve is a seven-second centred average so the shape reads; the tooltip gives the raw second. Per-player figures bucket in five seconds, which is why the handles move in fives; the chart itself is per second. Rows are fixed, damage dealers first and then healers, alphabetically, so a name stays in the same place as you move the window.

Defensives in the heavy seconds

111 spans across the night ran at twice the pull's median damage taken or more. They are 13% of the time on the boss and 27% of the damage. This is who pressed something in them.

Share of the 111 heavy spans in which the player used at least one major personal cooldown. Labels give the raw count of majors used in heavy spans, out of everything they pressed all night.

Two bars that don't mean what they look like. The Frost bar sits low because that spec mitigates with a short cooldown rather than a saved one: 184 casts of Ice Barrier, which is not a major by this count. The Blood bar is the opposite, where 224 majors is a rotation rather than 224 saves.

Healthstones & health potions

195 healthstones and 136 health potions went down across 25 pulls, and they went down late rather than early: 186 of the 331 at under 40% health, 56 of those under 20%, 13 above 80%. Average 39%, median 37%. Two of the roster used no healthstone at all; the chart marks them. The potion count is every health potion on the shared cooldown: 71 Concentrated Silvermoon Health Potion, 48 Silvermoon Health Potion and 17 Potent Healing Potion; the stones are 134 Healthstone and 61 Demonic Healthstone.

healthstoneshealth potionsaverage health when usedunder 25%, cutting it fine
Bars count uses across the night; the marker on the right is the average health the player was sitting at when they pressed it, measured before the heal landed.
PlayerMajorsIn heavyCoverageMinor / rotationalExternals givenStonesPotionsHP at useSolo deaths…with nothing pressed

Solo deaths are the 71 deaths where the raid was still standing, at least 70% of it alive, and fewer than three people went down within three seconds. Everything else is the raid falling over, whether that is a mechanic killing half the room at once or the last tank dropping a minute after everyone else. 58 of the solo deaths had no major defensive in the ten seconds before, and 300 deaths happened in pulls where that player never used a stone or a potion.

Heavy spans are computed, not hand-picked: raid damage taken binned per second, and every run of three or more seconds at twice the pull's median rate or above. Necrotic Vapors, Blight Vein and Acidic Expulsion dominate them. Majors are personal cooldowns (Anti-Magic Shell, Astral Shift, Barkskin, Blur, Dark Pact, Desperate Prayer, Die by the Sword, Dispersion, Divine Protection, Divine Shield, Exhilaration, Fortifying Brew, Greater Invisibility, Ice Cold, Icebound Fortitude, Lichborne, Metamorphosis, Rallying Cry, Shield Wall, Shield of Vengeance, Survival of the Fittest, Unending Resolve, Vampiric Blood). Minor covers the spammables: Alter Time, Blazing Barrier, Dancing Rune Weapon, Frenzied Regeneration, Ice Barrier, Prismatic Barrier. Stones and potions are counted from the cast, not the heal, so an overheal still counts as a use. Health at use is the value before the heal landed: WCL stamps the health on both the cast and the heal event as the figure after healing (casts that read 100% health while reporting real, non-overhealed healing give it away), so the effective heal is subtracted back off. Externals are the ones cast on somebody else: Anti-Magic Zone, Blessing of Protection, Darkness, Ironbark, Lay on Hands, Life Cocoon, Spirit Link Totem.

How this was counted

Burn damage is every damage event inside an exposure window landing on either the Venomous Heart or Ula'tek herself, absorbs included, with pet and guardian damage credited to the owner. The heart takes 100% increased damage, so a hit parked there is worth two of the same hit on the boss. The “on heart” column is how much of a player's burn went to the target that pays double. Exposure windows were rebuilt by clustering those events and breaking at gaps over five seconds; all 27 came out at 20 seconds, against 27 Rage of the Shackled casts minus the four pulls that died before the first exposure.

Wave hits are damage events from Caustic Waves (spell 1292211, 1292403, 1298570, 1311807 and 1319282), grouped per player when they land within two seconds of each other, so one wave passing through a player counts once. Wave casts count the boss's Caustic Waves casts. A wave death is a death whose killing blow is the wave itself. Damage taken and then dying to something else a second later is not counted here. Phase one is everything before the split phase and phase two everything after it, the split phase being dated from the coordinates as described below; casts within ten seconds of each other count as one volley.

Viper counts are distinct Blightscale Viper instances that took damage in a pull, a proxy for eggs hatched wrong; on this night they track the early pulls more than the deep ones.