Voidscar Arena

The 16.9 M of healing on Hyporock's sheet is 63% self-healing. Eternal Flame (the Holy Power spender that heals) put 8.2 M on the party and 7.0 M of that on himself; Lightbearer's passive splash was 3.5 M at 42% overheal. The healing that reached other people was 6.3 M over 25 minutes -- about 4.2k HPS -- and 3.2 M of it was four Lay on Hands.

The two on the healer are the finding. At 0:59 Resto Shaman was at 50,443 of 1,121,826 (4%) after a Null Eruption and a Lava Bolt; Lay on Hands put back 1.07 M, and 0.1 s later another Lava Bolt landed for 327,436. At 4:51 he was at 66,950 of 958,826 (7%) after a Savage Leap and Insidious Aura; the heal landed, and a 151,451 Savage Leap arrived 0.1 s after it and a 285,802 Insidious Aura 0.8 s after that. Neither hit was survivable at the health he had. Two healer deaths that did not happen: 30 seconds of penalty on a key timed by 194, and the first one would have left the opening pull without a healer.

The two self casts were at 40% each. The 17:02 one prevented a death (555k landed in the next four seconds, 3.6k of other healing reached him). The 24:20 one did not need to be Lay on Hands -- 177k landed after it -- but the cooldown is about four minutes on this build (232 s between the first two), so spending it there cost nothing.

Sacrifice ate 1,527,879 of other people's damage across seven casts, three on the tank, four on the two damage dealers. Never on the healer, whose death was one hit.

The interrupt axis counts stops, so it cannot tell a press that stopped something from one that landed on a mob somebody else had already kicked. Every Rebuke press was checked against the target's cast bar from the raw log:

• 13 of 16 Rebukes stopped a cast: 5 Lava Bolts, 4 Mending Voids, 2 Violent Sands, a Mad Shriek and a Demoralizing Shout.
• 2 landed 0.1 s and 0.4 s after Arms Warrior's kick on the same mob (10:23, 17:38). Two kicks in one tenth of a second is a coordination overlap, not a wasted cooldown -- the other kicker would have missed it otherwise.
• 1 (0:39) hit a shaman that was not casting.

Hammer of Justice, 9 presses: two caught a Lava Bolt 0.1-0.3 s into its cast while Rebuke was down, three stunned a Protective Turtle and three a Voidminder between casts. Blinding Light twice. The group leaked 23 of 44 Lava Bolts; Hyporock stopped 7 of them himself, with the tank at 8 stops against 17 expected and the healer at 5 against 11.

Cooldowns were pressed on cooldown. Wake of Ashes 48 times at a median 32 s gap on a 30 s cooldown, Divine Toll 25 at 62.5 s on 60 s, Avenging Wrath and Execution Sentence 22 each at 68 s on 60 s, Divine Shield five times at the 5-minute floor, five potions, eleven trinket presses on a 2-minute cooldown. The one that was not: Divine Protection, 12 presses at a median 125 s gap on a cooldown the registry puts near 45-68 s -- roughly half of what it could have been on a key where he took 31,732 damage per second.

Six Cleanse Toxins: five removed Mind-Numbing Poison (a cast-speed slow) and one took Corrosive Essence off himself. Corrosive Essence, the 130-second Poison that carries most of this dungeon's dispel load, was removed three times -- twice by the healer's Poison Cleansing Totem at 22:54, once by Hyporock -- and two applications ran their course for 2.9 M. Freedom four times, three on the tank. Blessing of Protection twice, one on the mage at 0:19 and one on himself at 18:06, two seconds after Dreadbellow landed.

245,016 DPS is Hyporock's second-best on a completed Voidscar (253k in a +10 on 09-09) and above his 34-key Ret median of 222k. Inside the party it is 104% of the other damage dealers' median, which on a three-DPS key means he edged the mage (240k) and the warrior (230k). The damage axis is a rank, so that reads 78; it would read the same if all three had done 20% more.

The outside number is Warcraft Logs' 64th percentile against other Ret Paladins at +15 in this dungeon, where his own Voidscar record from a lower key is the 95th. Arms Warrior's 230k at item level 302 against Hyporock's 321 is the comparison that matters in the party: the warrior was 19 item levels under him and 6% behind on damage.

Holy Power overcap was 25% of 1,896 generated. The build wastes about 22% in a simulation that plays it perfectly (see the Ret rotation notes in the archive), so 3 points of that is real and the rest is the spec. Every 'Not yet recovered' refusal is the global cooldown, not a mistake, and the 21 'target needs to be in front of you' are a rounding error.

Nothing in the rotation was missed. The generators and spenders were pressed at the cadence the simulation wants, every major cooldown was used within a few seconds of ready, and the gap to a top parse in this key is the usual one: melee uptime on a dungeon whose bosses -- Taz'Rah's shades especially -- spend a third of the fight where a paladin is not.

Two of Hyporock's four weighted axes are scored against the other players in this key rather than against a scale. `damage~ 78` says 'a little above the middle of three'. `interrupts~ 97` says '20 stops where the party's share would have been 9' -- that one is a real measurement wearing a rank's clothes, because the expectation is the group's total divided by the number of players who could have kicked, and the group's total was 51 of 75 reachable casts.

The utility axis ('19 used of ~91 the kit offered') carries no weight for a damage dealer and is the sum of every press every cooldown could have produced. Its 87 is a decent number and mostly says Divine Protection was under-pressed; the Lay on Hands saves above are not in it, because a save and a routine press count the same.