Annihilator Quest

Introduction

This page explores the mechanics, statistics and strategy of the Annihilator quest.

Set up

The quest must be completed by four level 100+ players of any class. Once the quest reward has been taken the player cannot repeat the quest.

There are 6 demons, players must defeat two to escape the room.

Demon spellcasting

• Fireballs: 110-200 fire damage, cast with a chance of 1/3. A fireball damages every player.
• Energy beams: 300-420 energy damage, cast with a chance of 1/10. For any player there are 3 demons that can cast energy beams.
• Mana drain: average of 70 mana drained, cast with a chance of 1/8, an average of 12 mana per second per player. This mostly affects mana shield users. This damage isn’t considered in the calculator.

The demon in melee range rolls the chance to cast spells in time with its melee hits, and only retargets on 10% of these two second turns. The other 5 demons are unable to find a path to their target, so they search (retarget) every second. Each retarget triggers a spell roll, causing these 5 demons to attack twice as fast as the adjacent demon.

If every beam hit every player, the average damage from fire and energy on turn 1 and 2 would be 310 + 108 = 418, and 258 + 72 = 330, for an overall average of 374 damage per second, from 1.83 fire attacks and 0.25 energy attacks per second.

Demon targeting

Looking more closely at original videos, you can spot the demons occasionally change direction. This is because they are targeting something other than the player in front of them. Whenever a demon retargets it uses the following logic, to select a target among all creatures within 10 tiles (a 9 tile gap):

• 70% Closest creature, by Manhattan distance (dX + dY).
• 10% Lowest HP creature.
• 10% Highest Damage, the player that has dealt it the most damage.
• 10% Random creature.

Ties in the first three are picked at random.

A demon faces its target, along the x direction if dX and dY are equal, and fires its beam in that direction whether or not the demon can walk or throw to the target. A demon that has turned to target another player or summon can miss its energy beam completely.

A demon can only cast a fireball if its target is in throw range, i.e. only to players in the quest. If the demon targets a player or summon outside the quest room, it will not cast a fireball.

In contemporary 7.4, people would often wait outside the Anni room to cheer on their brave friends. It is nice to think these supporters were actually helping their friends' chances of survival.

Manipulating Demon Targeting

We can try to increase how often these distractions happen:

Target closest

This will almost always be one of the four team members, meaning the demons will fire as normal. There is one exception: if there is a summon behind the quest door, then demon 6 will be equally close to it as to player 4.

Target lowest HP

This targets the creature with the lowest “max HP”. By summoning skeletons outside of the demon room and in range of every demon, each demon will cancel a fireball and miss an energy beam 10% of the time; every time. With no other low HP targets outside the demon room, the order of the team can make a small difference, with the weakest player on the left being preferred, to make demons 2 and 4 turn.

Target highest damage

This causes demons to target the player who has dealt them the most damage. If a demon heals off damage it still counts, similar to how XP is calculated. Players 1 and 4 can shoot explosions at the two demons opposite each of them, causing those demons to turn and miss their beams (fireballs will still hit).

If no damage has been done to demons 1 to 4, then all creatures in range, having all dealt 0 damage, will be valid for the tie-breaker and chosen at random. This means if a large number of skeletons are summoned outside, then the demons will target them as “highest damage” most of the time.

Target random

This targets a creature at random out of the available targets. Without any summons the demon will sometimes target a player to its side and turn, missing a beam.

By maximising the number of surrounding creatures using Animate Dead or Undead Legion, the chance of the demons targeting the four Anni team members can be significantly reduced. Each demon can have from 8 to 74 skeletons in range (88 skeletons total), without using the reward room.

Reward room

There is a way to leave summons in the reward room; this requires a previous team to have completed the quest. I will leave the method unexplained for now.

Annihilator calculator

The below map and settings allow you to calculate how demons behave and damage distributions for various annihilator scenarios, for one of the four players. Rather than just using the average damage, a damage distribution helps to tell how often high combos occur.

To use the calculator click the six tiles on the player row to set up to 4 players and 2 demons. To include the effect of skeletons click on valid tiles on the map, or use the preset toggles. Click a demon to show its target range. Additional toggles are provided to enter which player has dealt each demon the most damage, which rings or amulets are equipped, and player HP.

Analysis and Discussion

1. A lv 100 knight will very rarely be killed in one hit, particularly with resistance gear, though this is much more likely if they are blocking. Knights are more likely to be killed by a misclick or a 2-turn combo (2 consecutive high damage seconds), although even these are quite rare.
2. Fire damage makes up 50% or more of the total damage, for most damage values. This is due to it being more likely that two demons cast fireballs (1/9) than 1 energy beam (1/10).
3. Player damage can be calculated on the other calculators on this site. Knight melee damage does virtually nothing, and ice rapiers are also not very effective due to the demons' huge melee defences.
4. At level 100, power bolts do slightly more damage than burst arrows for paladins, but they only deal damage to one demon.

Strategy

Examples use level 100 characters unless stated otherwise. Many players waited until higher levels for more HP, damage, healing power and mana, which help a great deal, so a level 100 example is not a statement of viability in itself.

Team set-ups

• Knight blocker, knights or paladins shooting: the strongest knight blocks, ideally at a higher level, with or without burst arrows. The remaining knights and paladins shoot burst arrows or power bolts. With one shield blocker and three lv 100 knight shooters it takes approximately 70 turns (140 seconds) to kill the demons.
  • Power bolts average 80.71 damage at 90 distance skill and 91 at 100 skill.
  • Burst arrows average 50 damage at lv 100 and 83 at lv 150.
• 4 sorcerers with UE: can kill all 6 demons in moments, but if this fails they are quickly wiped out.
  • On PvP worlds the sorcerers all receive friendly fire and should wear stone skin amulets. Killing a demon with 8200 HP applies the same damage to the other players, but PvP damage is halved and a stone skin amulet reduces it by 80%, so killing the demons with UE costs each sorcerer 8200 / 2 / 5 = 820 mana in friendly fire.
  • If every sorcerer can cast 2 UEs and 2 burst arrows this can be enough to kill the demons, but it is not guaranteed, so extra levels are recommended, particularly for the blocker (TBC).
  • On non-PvP worlds sorcerers get the double benefit of no friendly fire and a spare amulet slot for an elven amulet. This is easy to do at level 100, perhaps with a stronger blocker.
• 3 sorcerers with UE, 1 druid blocker: the sorcerers would need to be a higher level.
• 4 druids: druids struggle the most with this quest due to their low HP and lack of UE. An all-druid team would want to kill all the demons with SD.
• Mass healing druid (unusual): one druid casts mass healing on the other players. This only works if the druid survives, and the druid has no mana shield. On PvP worlds the blocker must also survive, otherwise the healing will heal the demons. A lv 100 druid’s mana would last 23 seconds.

Reducing damage

• Killing demons: once the first demon dies, average damage to the shooters drops by around 24%.
• Skeletons: adding skeletons to all normally accessible areas reduces average damage by around 23%. This is very significant and can make otherwise suicidal strategies possible, such as no-mana-shield mages or an all lv 100 druid team.
  • 88 skeletons (44 animate dead runes) cost the same mana as 3 BP of SD.
  • Undead legion can do more but may be difficult to set up. Skeletons can be managed with parcels.
• Example, no mana shield: a lv 100 mage with a might ring, elven amulet, one demon killed and skeletons has average damage reduced by 60%, and only a 1/400 chance per second of being killed, compared to 1/7 for a vanilla set-up.
• Average damage vs combos: every reduction in average damage causes a larger reduction in the chance of combos of a given size. For example, a 10% reduction from an elven amulet reduces the chance to one-hit kill a lv 100 mage by 37%.
• Example, mana shield: mages with mana shield care about average damage more than combos. A lv 100 mage with 2795 mana, elven amulet, might ring and skeletons takes 190 damage per second on average. Mana shield would last around 14.7 seconds, with death at around 18 seconds. This is enough time to kill both demons, after which average damage drops to 156 (without the might ring) and UHs can be used somewhat safely.

Amulets, rings and supplies

• Elven amulet: the best choice overall, giving the highest damage reduction over a medium duration. For a knight with a shield it is strictly better than a stone skin amulet, with a slightly higher reduction that lasts longer, and other amulets may be better still.
• Stone skin amulet: good for UE teams and for blockers that are not using shields.
• Dragon necklace and strange talisman: block 8% of fire and 10% of energy respectively. Useful for players who expect to finish slowly but do not want the risk of messing up healing while refreshing amulets, or of forgetting. The necklace has the lower percentage, but average fire damage is 3x that from energy beams, which makes it the clear favourite. For higher damage combos the balance shifts towards a more even split of fire and energy, so the talisman performs the same while using fewer charges.
• Protection amulet: no real use case. The reduction is low and fire is far more threatening than physical damage, even for a weak blocker.
• Amulet of loss: useful for knights who bring a very large or expensive amount of supplies (lots of UH, might rings and ice rapiers), including for the scenario where the teammates all die and the knight has to solo the quest. Many players choose an AoL because the value of their gear and supplies makes it worthwhile. Players shouldn’t bring valuable gear without one, though other than a shield it is not very impactful. Definitely unsuitable for mages, who carry no gear and would not expect to survive more than a minute.
• Durations: a might ring lasts 9 seconds on average, or 12 with skeletons. An elven amulet lasts 24 seconds, or 31 with skeletons. Dragon necklaces and strange talismans are unlikely to run out.

Healing and timing

• Knight healing: UH heals a fixed amount, which is not much more than the average damage from the demons. A lv 100 EK heals for 550, and with no damage mitigation takes an average of 410 damage per second as a blocker and 370 as a shooter.
  • After a combo they can only build HP back at 140 (blocker) or 180 (shooter) HP per second, leaving them vulnerable to a second combo.
  • Realistically a lv 100 knight blocker with no amulet or summons would die, so a higher level is needed. Along with mistakes, this is one of the main causes of failure for knights.
  • Damage mitigation (skeletons and a dragon necklace) lets the blocker catch up much faster, at 240 to 284 HP per second.
• UH queueing: UH runes can be queued. If you use a rune 0.8 seconds after the last UH you will be healed at exactly the 1 second mark, so a consistent sub-one-second interval matters more than precise timing. Attack runes are different: an attack attempted before the two second interval fails. Queueing does not apply to instant healing spells.
• Attack rune timing: individual animations are very difficult to see, so watch the burst arrow or power bolt count to time attack rune shots.
• Equipping items: equipping an item causes a 1 second delay to your next action, so it may be counterproductive when you need to maintain HP (TBC).
• Screen set-up: some players change window size or screen resolution, narrowing the grey column to make the character and backpack bigger and shorten the distance between them, for faster and more accurate UH targeting.

Table: Annihilator damage statistics (Player 1)

1. "1/Chance" represents how many seconds between hits of that damage, on average. A value of 600 would mean you typically see that damage (or more) once every 600 seconds.

Special thanks to Kay for a few hints and pointers to mechanics I was unaware of that made all the difference to this analysis and strategy.