The Necromancer's Due
A deep dive into Shroud of the Avatar's death magic, what Ultima Online and the genre since learned about necromancy, and eighteen concrete proposals.
Death magic is the school the community reaches for when it wants to prove a balance point, and the proof is easy to make: the school of the grave is at its weakest in a grave. Undead heavily resist it, its life drains refuse to feed on the dead, and its summons cost more experience than anything comparable while delivering less. This document measures the school against the genre that invented necromancy, reads every death rune and summon formula in the shipping data, and proposes eighteen changes built on one finding: much of what this school is missing is already in the game's files, wired to stats that nothing grants and a summon nobody can see.
Ultima Online launched in 1997 with undead everywhere and no necromancy at all. Five and a half years later, Age of Shadows retrofitted a complete death school onto the live game: sixteen spells built on four pillars. Curses that warp the fight (Corpse Skin turning resistance into vulnerability, Evil Omen priming the next hit). Drains that make health a currency. Forms with real trade-offs (Lich Form's mana engine paid in health regeneration). And corpse-fed sustain (Spirit Speak healing the caster by channeling nearby corpses). Shroud of the Avatar shipped the Ultima lineage's death school with only one of the four pillars. The drains are here: Tap Soul, Death Touch, Death Ray, the near-death damage bonus. The curses, the corpse economy, and the tools that would make an undead court worth keeping are missing, and some of them are literally in the data already, wired to stats that nothing grants.
The community's two loudest asks about this school are both on record, and both are symptoms of the missing pillars. First: Death is helpless against the undead it claims to command. The measured number is stark: zero sustained damage against tier-5 undead at grandmaster skill, and the community thread that surfaced it asks specifically for active mechanics to counter monster magic resistance, not just quiet number changes. Second: the Lich is not worth its price. Summon Lich costs six times baseline experience, and players have said plainly that the summons do not deliver. The gaps cluster in three places:
25 × deathMastery, which reads zero for every player who has not taken Death specialization. The school has no working answer to its own signature matchup. The resistance numbers themselves belong to the running combat balance program; this page proposes counterplay verbs only.Of the eighteen proposals: eleven are data-only, seven touch client code, and none require server work. During this review we filed nine defect reports (#2598–#2606) for the dead stats, wrong-token formulas, and orphaned assets the audit turned up; those are bugs regardless of any vote. The proposals are where the votes matter: they decide whether the school gets its missing pillars back.
| # | Proposal | Tier | Size | Server | One-line pitch |
|---|---|---|---|---|---|
| N1 | Ward Rend | 1 | S-M | none | the resist strip finally fires for everyone, and it tears at magic warding too |
| N2 | Bonepiercer | 2 | S | enum | a new innate: your Death spells bite undead targets harder |
| N3 | Necrotic weapon enchant | 2 | M | none | the death enchant, sibling of Frost and Shock; a lane the magic wards never tax |
| N4 | Grave Siphon | 1 | S-M | none | consume a corpse to restore health and focus; sustain that works in a crypt |
| N5 | Grave Omen | 1 | S-M | none | one short window where the target's ward falls; land your heaviest spell inside it |
| N6 | Hungry Dead | 1 | S | none | Undead Mastery teaches your summons the life drain they already know |
| N7 | The Skeleton Mage takes the field | 1 | S | none | the fully authored fourth summon joins the tree |
| N8 | A Lich worth its price | 1 | S-M | none | two scaling bugs fixed plus a drain aura; the XP question goes to the balance program with your votes |
| N9 | Feed the Dead | 1 | S-M | none | feed a corpse to a summon to wind back its decay |
| N10 | Deathknell | 1 | S-M | none | a slain or expired summon detonates; mortality becomes ammunition |
| N11 | The Last Command | 1 | M | none | order a summon to die for you; keep the army or cash it in |
| N12 | Grave Essence | 1 | M | none | corpses you spend leave stacking essence in you, capped at five |
| N13 | Exhume | 1 | M | none | call up workable remains when the field is empty |
| N14 | Blood Price | 1 | S-M | none | actives that spend health for power, in Tap Soul's grammar |
| N15 | A Death Shield that shields | 2 | M | none | the shield mitigates again and the tooltip tells the truth |
| N16 | Death Field, priced by the dead | 1 | S-M | none | the school's priciest spell gets cheaper among the dead |
| N17 | The enchant that does nothing, fixed | 1 | S | none | a shipping enchant starts doing what it says |
| N18 | Corpse Explosion chain | 1 | S-M | none | explosion kills prime new corpses, damped per hop |
Death magic ships as eight spells and five passives hung off a single root, Tap Soul, plus two capstones. The identity in the data is coherent: health as a currency (Tap Soul trades health for focus and grows stronger the closer you are to death), life drains at three ranges, a corpse detonation, a shield that converts suffering into healing, decaying skeleton and lich servants, and a specialization capstone that opposes the Life school. The trouble is how much of it does not fire. Here is the tree as it ships, with the audit's findings marked:
[ TAP SOUL ] ◂1
trade health for focus, the school root
| | |
+--------+ | +-----------+
v v v
[ CORPSE EXPLOSION ] [ DEATH TOUCH ] ◂2 ( DEATH MASTERY )
needs Tap Soul 10 needs Tap Soul 10 innate, Tap Soul 10
targets a corpse ◂4 | | |
| v v v
v [ DEATH RAY ] ◂2 ( UNDEAD ( MUMMIFIED
[ SUMMON SKELETON ] needs Death Touch 40 MASTERY ) ◂5 UNDEAD ) ◂7
needs Corpse Expl. 40 | | needs DM 40 needs DM 40
decays over time ◂7 v v |
| [ DEATH SHIELD ] [ DEATH FIELD ] |
| needs D.Ray 40 needs D.Ray 40 |
| ◂8 highest focus ◂8 |
| | cost in school |
+-------+-------+ +-----------+--------+
v v v
[ SUMMON LICH ] ◂5 ( MASTERY DEATH ) ( WARD LIFE )
needs Death Shield 40 needs UM 80, spec needs UM purchased
and Summon Skeleton 40 capstone, 2000% XP capstone
600% XP cost
not on this tree at all: SUMMON SKELETON MAGE, fully authored, school
field set to None, invisible to every player ◂6
school wide: zero sustained damage against high tier undead at GM skill ◂3
25 × deathMastery, and every deathMastery term reads zero unless the player runs Death specialization in Learning or Maintaining. For most players the school's only counterplay tool never procs. → N1 Ward Rend.Every proposal on this page plugs into the same pipeline, and the pipeline's shape is why this page can promise so much with so little code: combat here is a deterministic simulation run by the players' machines, with the server picking a random seed and relaying it. Spells, riders, stacking rules, and target rules are authored in data.
flowchart LR H["Hotbar click"] --> V["Client validation: focus, range, target, fizzle"] V --> M["Master effect pick"] M --> S["Seed relay"] S --> C["All clients simulate: same seed, same result"] C --> D["Effect dispatch"] D --> DA["Damage: attacker's machine"] D --> ST["Stat and CC: target's machine"] DA --> K["On kill: kill effect riders fire"] ST --> K E1["Extension point: rune data"] -.-> M E1 -.-> D E2["Extension point: kill effects"] -.-> K E3["Extension point: channels"] -.-> D classDef ext fill:#c9a55a33,stroke:#8a6d33,stroke-width:2px; class E1,E2,E3 ext;
The three marked extension points are where the cards live. Rune data: formulas, riders, stack counts, and target rules are asset-authored; a change here reaches players without new code, which is why eleven of the eighteen cards are data-only. Kill effects: on-kill riders are data-authored and already network-wired; Corpse Explosion uses this slot today, and N10 and N18 extend it. Channels: rune channels enforce exclusivity (the Shield channel is why only one shield spell holds at a time); the new curse cards compete on a shared channel instead of stacking.
base + mastery bonus so it works for unspecialized players and specialization amplifies it. And every number printed on a card is an illustrative starting point: final values ship through the combat balance program's changesets and its measurement windows, never through this page.The instructive part of Ultima Online's story is not that it had necromancy. It is that it did not. UO launched with undead as its most iconic enemy family and "necromancer" as pure flavor, and ran that way for five and a half years. When Age of Shadows shipped the Necromancy skill in 2003, the designers were answering the same question this page answers: what is a death school for, in an Ultima game, years after launch? Their answer was not another damage bar. It was curses that warp the fight: Corpse Skin trades away a target's cold and energy protection to open it to fire and poison, the oldest resist-manipulation curse in the lineage. Evil Omen primes the target so the next magic hit lands harder, a one-shot amplifier that creates setup-and-payoff play. Blood Oath reflects the victim's own aggression back at it. Curse Weapon temporarily gives the caster's weapon a life drain, a necrotic weapon enchant in 2003. The forms complete it: Vampiric Embrace, Wraith Form, and Lich Form each trade something real for power, and Spirit Speak heals the caster by channeling energy from nearby corpses, corpse-fed sustain as an official skill design. Necromancy also carried a price: casting it drained karma, and the karma-loss item property added in 2005 deliberately deepened the bargain.
Every pillar of that 2003 retrofit maps onto something this page proposes, which is the argument in one sentence: restoring the lineage toolkit is not invention. The rest of the genre spent twenty years refining each pillar, and the survey below is organized around what each game proved.
EverQuest: the drain-tank and the mana engine. The Lich line converts the necromancer's health into mana, lifetaps close the loop, and the class carries undead-specific nukes and undead charm: the school that raises the dead is also the school that commands the enemy dead. Its failure mode is the lesson too: self-sufficiency that complete made necros awkward in groups. Guild Wars: corpses as a contested, finite resource. Minions are raised from actual battlefield corpses and whoever exploits a corpse first consumes it; minions decay permanently, Blood of the Master heals them by sacrificing the caster's own health, Soul Reaping pays energy on nearby deaths and had to be rate-capped, and the Death Nova "minion bomber" made a minion's death into a payload. Diablo II: the resist-strip canon. One curse per target makes cursing a decision; Amplify Damage and Lower Resist are the template for active shred, and Lower Resist breaks immunities at one fifth effect, the canonical answer to hard-immune enemies. Corpse Explosion splits its damage half physical, half fire so no single resistance walls it, and Revive raises slain monsters on a 180-second decay timer. World of Warcraft: the strip as party utility (Curse of the Elements raised the whole raid's damage), Affliction's rolling-DoT drain-tanking, Unholy's disease-and-ghoul build-and-spend loop, and Death Strike's heal-scaled-to-damage-taken as the modern drain defense. Path of Exile: the corpse as a first-class object. Detonate Dead scales to the corpse's maximum life, offerings spend corpses on timed buffs, Desecrate manufactures corpses when the field is empty, and Minion Instability turns a dying minion into a fire payload. Diablo IV: corpses spawn as a visible class resource, and the Book of the Dead lets a necromancer customize each minion line or sacrifice it entirely for permanent power, along with the launch lesson that when sacrifice beats the pet, nobody keeps the pet. Guild Wars 2: life force as a death-fed resource and, in Scourge, the death school as a group defender through barrier, with years of over-performance nerfs as the cautionary tale. ESO: the corpse loop modernized with tethers and Blastbones, sacrifice as a rotation verb, and visible necromancy flagged as a criminal act, the modern heir to UO's karma price.
The genre's answer to "this enemy resists your school" is remarkably consistent, and it is never "make the resist smaller and move on." It is counterplay verbs. The vulnerability curse: a castable debuff that visibly flips the matchup, from Corpse Skin through Amplify Damage to Curse of the Elements, balanced in Diablo II by one-curse-per-target exclusivity. Immunity broken at reduced effect: Lower Resist pierces hard immunities at one fifth strength, so the immune monster becomes workable but stays tough; undead keep their identity, the wall stops being absolute. The one-hit amplifier: Evil Omen's single primed strike, self-limiting and legible in PvP. Minions as resist-agnostic damage: a skeleton's sword does not care about magic resistance, which is the classic answer to immune monsters and the reason the summon line can be the school's own built-in counterplay. Commanding the resisting family: EverQuest's undead charm and Diablo II's Revive make the school that undead resist into the school that turns undead against each other. And the anti-undead item lane: UO's Silver weapons doubled damage against undead, counterplay that lives in gear. Group A of this catalog is these patterns, translated into this game's data.
| Capability | UO | EQ | GW1 | D2 | WoW | PoE | D4 | GW2 | ESO | SotA today | SotA proposed |
|---|---|---|---|---|---|---|---|---|---|---|---|
| Active resist shred | ● | ○ | ○ | ● | ● | ○ | ○ | ○ | ○ | ◐ gated | ● N1 |
| Immunity break at reduced effect | ○ | ○ | ○ | ● | ○ | ○ | ○ | ○ | ○ | ○ | ● N1 |
| Corpse consumption spells | ◐ | ○ | ● | ● | ○ | ● | ● | ○ | ● | ◐ one verb | ● N4/N9/N12 |
| Corpse creation and reliability | ○ | ○ | ○ | ○ | ○ | ● | ◐ | ○ | ● | ○ | ● N13 |
| Minion death detonation | ○ | ○ | ● | ○ | ○ | ● | ○ | ○ | ● | ○ | ● N10 |
| Corpse or blood fed minion upkeep | ○ | ○ | ● | ○ | ○ | ◐ | ○ | ○ | ● | ○ | ● N9 |
| Minion sacrifice choice | ○ | ○ | ◐ | ○ | ○ | ◐ | ● | ○ | ● | ○ | ● N11 |
| Death weapon enchant | ● | ○ | ○ | ○ | ◐ | ○ | ○ | ○ | ○ | ○ | ● N3 |
| Blood magic actives, health as cost | ● | ● | ● | ○ | ◐ | ○ | ○ | ○ | ○ | ◐ Tap Soul | ● N14 |
| Drain to defense conversion | ○ | ○ | ○ | ○ | ● | ○ | ○ | ● | ○ | ◐ off in code | ● N15 |
| Death fed resource | ○ | ○ | ● | ○ | ○ | ◐ | ● | ● | ○ | ○ | ● N12 |
| Anti-undead identity | ◐ items | ● | ○ | ○ | ○ | ○ | ○ | ○ | ○ | ○ | ● N2 |
● native · ◐ partial or era-limited · ○ absent. "SotA today" = the unity-6.5 branch, 2026-07. "Gated" = exists but reads zero without Death specialization; "off in code" = the conversion half of Death Shield is disabled while its tooltip still claims it.
Group A is the counterplay the community asked for by name. Group B makes the undead court worth its price. Group C is the corpse economy, and it is labeled honestly: no forum thread asks for it, but every card in it is the genre's proven answer to the problems groups A and B keep exposing. Group D is the housekeeping the audit found. Every tuning number is an illustrative starting point; final values ship through the combat balance program's changesets and measurement windows.
Death Touch and Death Ray gain a real curse. Each hit can now tear a piece of the target's Death and magic warding away, stacking up to five times, and the tearing works for every Death mage instead of only specialists. Heavily warded undead stay tough, but they stop being untouchable.
Seen in: Diablo II Amplify Damage and Lower Resist, WoW Classic Curse of the Elements, Ultima Online Corpse Skin.
Two rules are imported from the genre on purpose. The Diablo II immunity rule: the total strip is capped at roughly one fifth of a high-tier undead's warding, so a shredded target becomes workable, never naked. And the one-curse rule: the strip shares a curse channel with Grave Omen (N5), so shred-over-time and omen-burst are mutually exclusive per target and cursing is a decision, not a rotation.
Redesigns the existing second effect on the Death Touch and Death Ray rune assets: today a 25×deathMastery chance to apply -5 DeathResistance for 30 s, 3 stacks, which is 0% for unspecialized players. Proposed grammar: chance (35+(25×deathMastery)), stack cap 5 via the stack-count expression, and a second leg biting MagicResistance, counted against the rune's fixed master-effect slot budget before it is added. The exclusivity rides the existing rune channel machinery (the same system that makes shield spells exclusive), so no new code. All expressions regenerate through the formula tokenizer; the unspecced-functionality test pins that every term is nonzero at mastery 0. Related report: #1391 (bleed riders affecting ancient undead that should be immune).
A new innate for necromancers who hunt their own kind. Ranks of Bonepiercer add bonus damage to your Death spells against undead targets. The school that studies death becomes the school that ends the walking dead, instead of the worst tool for the job.
Seen in: EverQuest's undead-specific necromancer lines, Ultima Online Silver and Undead Slayer weapons.
One new skill-sheet innate row, reusing the dead, already Death-school-mapped skill id that #2604 flagged. Damage formulas gain a term through the existing creature-type token, so the bonus applies only when the target is undead. The granted stat is the one place this page mints a new character stat (the four dead stats the audit found are all consumed by other cards); the fallback, if that cost is refused, is a data-only version baked into the formulas without a skill row. Mid-tree placement and a moderate magnitude keep it from becoming an always-take: outside crypts it is dead weight by design.
Death gets its weapon enchant, the analog of the Frost and Shock enchants that shipped this month. A necrotic gem gives a weapon a chance to proc a life-draining strike, with a rider that bites harder against undead. Melee hybrids gain a lane the magic wards never tax.
Seen in: Ultima Online Curse Weapon and the Silver weapon line, WoW Death Strike as the sustain analog.
The Frost/Shock pattern, applied to the one school family the weapon-shell art never covered: the shared weapon shell prefab carries fire, ice, lightning, poison, and energy variants and no death shell, so this card adds the missing shell plus a weapon-hit proc rune (drain damage, small heal-back; the heal-back is excluded against undead targets by the existing engine gate and the card says so). The orphaned Death Ray impact effect (#2603) is the candidate donor FX. Which gem and which enchant pool carry it is decided jointly with the weapons and enchanting pages, one shared itemization rail, no duplicate design here. Related asks: #2349 (per-school weapon enchants), #1908 (the old Drain Health gem proc underperforms; this is its successor).
A new drain that feeds on the fallen instead of the living. Target a corpse and consume it to restore your health and focus. In a crypt full of the dead, the necromancer finally sustains like one.
Seen in: Ultima Online Spirit Speak, the official 2003 corpse-channeling design; Guild Wars Soul Reaping.
Uses the existing corpse-targeting primitives (the require-dead flag and dead-target types Corpse Explosion already uses). The restore is a direct heal plus focus effect authored off a corpse target, which structurally sidesteps the drain heal-back exclusion: that gate fires on living undead targets, and a corpse never enters the path. Loot safety is a design rule on the card: a siphoned corpse keeps its loot, or siphoning is blocked until the corpse is looted, decided at implementation. Targeting must not inherit #1549 (Corpse Explosion consumes the corpse nearest the player instead of the one aimed at); a generous range and area-corpse targeting dodge the fiddliness ESO had to patch around.
A curse of moments. Mark a target and for a few seconds its warding against your next Death spell falls away sharply. One window, one payoff. Land your heaviest spell inside it or waste the omen.
Seen in: Ultima Online Evil Omen, the canon one-hit amplifier.
A short-window, single-stack curse: a large Death-warding strip for roughly six seconds, on the same curse channel as Ward Rend (N1), so the sustained shred and the burst omen are mutually exclusive per target, reproducing Diablo II's one-curse decision with existing channel machinery. A true consume-on-next-hit refinement would need client effect code and is deferred; the timed window ships first. Burst numbers are illustrative and flagged for PvP review.
The framing for the whole group: minions are resist-agnostic damage. Making the court worth fielding answers both community asks at once. Decay stays; decay becomes a game. And a boundary honored throughout: summons are not pets. Commands, stances, gear, and bonding belong to the taming page and stay there; these cards touch only summon-specific machinery.
Your skeletons and liches already know how to drain life. The skill tree just never taught them. Undead Mastery now feeds your summons' hunger directly, so every rank makes their life-drain strikes stronger. The Zombie Skin Belt keeps its bonus on top.
Seen in: EverQuest pet sustain, Guild Wars minion upkeep economics.
The drain strike is authored and live in the summon weapon runes, gated on a stat only the Zombie Skin Belt artifact line grants (#2599). This card adds the missing skill path: Undead Mastery's skill-sheet row additionally grants the drain stat per level, belts stack additively on top and stay best-in-slot for drain builds. Honest constraint stated on the card: the drain's heal-back is still excluded when the summon strikes undead targets, so the drain sustains the court in living content while its physical damage carries crypts. Related: #1709.
A fourth summon already exists in the game's data, fully built. The Skeleton Mage has a spell, a staff attack, a model, and a name, and has been invisible for years because of a single wrong field. It joins the tree as the court's ranged caster.
Seen in: Diablo II's fighter and mage skeleton split, Diablo IV Book of the Dead minion variants.
The skill-sheet row for Summon Skeleton Mage carries school None, so the finished summon appears nowhere. The card flips the school to Death, places it at Summon Skeleton 40 to mirror the fighter-to-lich ladder, and adds the power-stat grant the fighter row has and the mage row lacks (reusing the existing skeleton power stat rather than minting one). The structural audit test this page promises exists precisely to catch this class of defect. Related: #1387 (the Death tree rendering empty in the trainer window). Its numbers have never seen live play, so it lands together with N8's scaling fixes and the shared before-and-after summon sheet.
The Lich costs six times normal experience and the community has receipts that it fights like it cost half that. Two data bugs are part of why: a stat that was clearly meant to power undead summons has never been granted by anything, and the summon damage formula reads the wrong character's mastery. Both get fixed, and the lich gains a drain aura worthy of its name.
Seen in: Diablo II Skeleton Mastery, EverQuest's pet economy, and Diablo IV's seasons of minion-scaling buffs.
Three legs, all data. One: fix the summon innate package's health expression, which reads the summon's own always-zero mastery instead of the owner's, and normalize it to the wrapped form its sibling line already uses (#2600). Two: grant the never-granted summon boost stat as Undead Mastery's payout; the stat's display name is literally "Undead Mastery," which is as close as data gets to stating intent. Three: a lich drain aura as a periodic authored effect, heal-back subject to the undead exclusion and stated as such. The 600% experience price itself is a number, and numbers belong to the balance program: this card hands the question over with the page's vote tally as evidence. Related: #1709 (the symptom thread), #2478 (tester-filed summon math).
Your summons rot by design. Now you can do something about it. Feed a corpse to your skeleton or lich to restore its decayed health and keep a well-fed court on its feet through a long fight. Decay stops being a countdown and becomes an economy.
Seen in: Guild Wars Blood of the Master, ESO's corpse-consuming summons, Path of Exile offerings.
A corpse-targeted rune whose payoff heals the caster's summon, reusing Grave Siphon's corpse-consume plumbing; the corpse-to-summon retargeting is the one piece that may need a small client assist. The "just resummon" math is confronted here rather than hidden: at high skill Summon Skeleton approaches zero cooldown and focus, so feeding must preserve what recasting destroys. A fed summon keeps its accumulated buffs and its Deathknell prime (N10), and feeding works mid-fight where a resummon cast does not. Boundary: this is summon upkeep keyed to summon-only decay stats, not pet feeding, which the taming page deliberately rejected; any owner-triggered summon button rides the same delivery rail the taming page proposes for pet abilities, one rail, cross-linked.
When one of your summons falls, it goes out with a bell toll. A slain or fully decayed summon detonates in a burst of grave energy, damaging enemies around it. The court's mortality becomes ammunition.
Seen in: Guild Wars Death Nova and the minion bomber archetype, Path of Exile Minion Instability.
Data-authored kill-effect riders on the summon prefabs, the same already-network-wired slot Corpse Explosion uses; no new protocol. Detonation damage in the base-plus-mastery grammar with a target cap. Hard visual dependency: the corpse-burst effect has been disabled in the engine for years with no tracker entry until this review filed #2606; that re-enable ships before this card does, because a silent detonation reads as a bug. Tuning rule stated up front: the detonation stays below the summon's lifetime damage contribution, so popping the court is a finisher, never the plan.
Order a summon to die for you. The Last Command consumes one of your undead, triggering its death burst and granting you a short surge of death power. Keep the army or cash it in. The choice is the point.
Seen in: Diablo IV Book of the Dead sacrifices, ESO Sacrificial Bones.
A rune targeting the caster's own summon (summon runes already carry self-plus-pet disposition targeting), applying a lethal negative heal to the summon plus a timed single-stack surge on the caster using existing stats; no new character stat. The kill triggers Deathknell's riders and leaves a corpse for the grave economy, so one press feeds three systems. Diablo IV's launch lesson is printed on the card: if sacrifice outperforms the standing army, the pet fantasy dies. The dodge is a real cooldown and a surge scaled so a living summon out-earns its own death over the buff window.
Labeled as design reach. No community thread asks for this cluster. It is on the page because every card in it is the genre's proven answer to the gaps the first two groups keep exposing, all of it stays in data and client code, and a proposal page is exactly the place to find out whether players want it. Vote accordingly.
Every corpse you consume leaves something behind in you. Grave Essence stacks a short-lived charge each time a corpse feeds one of your spells, and each charge sharpens your death magic. The necromancer in a corpse-rich fight visibly ramps.
Seen in: Path of Exile offerings, Diablo IV's corpse resource, Guild Wars Soul Reaping with its rate-cap lesson.
Every corpse-consuming verb on this page (Grave Siphon, Feed the Dead, Exhume's product being spent, and Corpse Explosion itself) carries an additional self-buff effect granting one essence stack: stack-count expression, hard cap five, short duration, granting existing stats per stack. The visible buff icon is the resource UI; no new interface. The rate cap Guild Wars had to retrofit is tuned here in data from day one via stack duration and the one-verb-per-corpse rule.
No corpses, no necromancer, says the old genre problem. Exhume answers it. Call up unearthed remains from the ground, a fragile husk that collapses in moments and leaves a workable corpse behind. Your corpse spells keep functioning in any fight.
Seen in: Path of Exile Desecrate, the genre's answer to Guild Wars' corpse drought; ESO Blastbones, a summon born to die.
No corpse-spawning primitive exists and inventing a server operation is off the table, so Exhume rides the existing summon machinery: a summon rune producing inert remains (a reused skeleton prefab with no attacks, minimal health, near-instant decay), and summon corpses are already legal corpse-spell targets in the shipping corpse-chaining rules. Exploit guards printed on the card: manufactured corpses carry no loot and award no experience when detonated, and the rune respects combat flagging. The Path of Exile downside is designed against: a real cast time and focus cost keep found corpses strictly better, so Exhume is the floor of the economy, never the loop.
Tap Soul already trades your life for focus. Blood Price extends that bargain into a small family of actives that spend health for power, rewarding the necromancer who lives at the edge. Your health bar is a resource. Spend it on purpose.
Seen in: Guild Wars Blood Magic's sacrifice costs, EverQuest's Lich line, Ultima Online's form trade-offs.
New actives authored in Tap Soul's exact grammar: a negative self-heal as the cost, the payoff as the product, for example a health-paid instant nuke and a health-paid ward. The near-death bonus's hard cap stays untouched, so the low-health identity deepens without a runaway, and Tap Soul's own stacking attunement penalty is reused as the anti-degeneracy guard. The known genre failure, health-as-fuel collapsing when healing is cheap, is countered by the school's own built-in taxes: Death Shield's healing penalty and the attunement debt. Related: #2306 (the live tester pass on Tap Soul's attunement interaction; these actives cite it rather than refiling it).
Death Shield's tooltip promises to convert incoming physical damage. The conversion is currently switched off, so the spell heals a little and protects nothing while claiming otherwise. This card makes the shield honest and gives it a real defensive identity drawn from the genre's best death-school defenses.
Seen in: Guild Wars 2 Scourge barrier, WoW Death Strike.
The conversion half of the death-transfer handler is commented out in the client, so the absorb only heals and damage passes unreduced while the localized description still claims conversion. The card restores mitigation with a retune, or respecs data-only toward a heal-scaled-to-damage-taken form if the code path is judged too hot; either way it writes the shield's active-state stat (#2602, following the working chaotic-shield precedent) so other formulas can key off shield state, and fixes the tooltip. Shield-channel exclusivity already exists. Group-shareable barrier is explicitly out of scope for v1: the genre's history says group absorbs are chronically overpowered, and a shareable version would be its own card and its own vote.
Death Field is the school's most expensive spell. Standing among the dead should make it cheaper. When corpses lie near you, Death Field's focus cost falls, turning the school's biggest bomb into the natural finisher of a fight it is already winning.
Seen in: Guild Wars 2 life force, Diablo IV's corpse-fueled economy.
A corpse-adjacency discount or partial refund on the spell's focus cost, the highest in the school. One binding gate, stated plainly: combat formulas may only read values that agree on every machine, so the corpse-proximity input must be verified as cleanly replicated before this ships; if it is not, the card downgrades to a flat cost-and-shape rework and says so. The discount is floored, never free.
The geometry bug this card used to carry (#2605) has already shipped on its own. The reading behind it was wrong: Death Field has no ring and no hole in the middle, and the number in question was never a radius. It was a minimum casting distance, which the spell should not have at all, because the blast goes off where the caster is standing. That number is now zero.
The cards that exist because the audit found shipped, player-purchasable things that do nothing. Cheap, visible, and the page's proof that the school's problems are specific and fixable.
A shipping armor enchant grants Corpse Explosion Damage. No formula in the game reads that stat, so the enchant has never done anything. This card wires it in, and normalizes a sister stat on Death Ray that follows the wrong math convention, so gear that claims to boost these spells actually does.
Seen in: no game; this is the housekeeping the genre never gets credit for.
The enchant grants a stat with no reader (#2598); the fix appends the school-convention term to Corpse Explosion's damage expression and flips the orphaned initializer. The same pass normalizes Death Ray's power stat from a bare multiplier to the wrapped convention every sibling stat uses (#2601). Both expressions regenerate through the tokenizer, and the promised pin test asserting the stat actually moves the number is the card's precondition, not its follow-up: an unrecognized token evaluates to zero silently, which is exactly how this bug shipped the first time.
The genre's most famous necromancer moment is the chain: a corpse detonates, kills, and the new corpses are primed to follow. Corpse Explosion gains a chance for its kills to prime the corpses they create, with a hard cap on cascade length. Density becomes drama.
Seen in: Diablo II Corpse Explosion, Path of Exile Detonate Dead, Diablo IV Corpse Explosion.
A kill-effect rider on the explosion: kills scored by the blast prime resulting corpses for a follow-on burst, chain chance in the base-plus-mastery grammar, cascade hard-capped and damped per hop. The damping is a design requirement, not a tuning nicety; every game since Diablo II re-tuned this spell because density farming degenerates. Diablo II's half-physical damage split is deliberately not imported, because it would need a new damage type bit, which the rejected list explains; the counterplay lives in Group A instead. Two dependencies: the corpse-burst visual re-enable (#2606) and the targeting fix (#1549) so chains aim where the player intends.
The catalog is eighteen cards, but it composes into two pictures. The first is the tree with its missing pillars restored: the same root, the same spine, and every dead stat, invisible summon, and silent rider brought into play. New nodes are marked NEW; changed behavior on shipping skills is marked CHG.
TapSoul (root) | +- CorpseExplosion (Tap Soul 10) | | CHG N17: the shipping enchant stat now raises explosion damage | | CHG N18: kills can prime new corpses that chain, damped per hop | +- GraveSiphon (Corpse Explosion 20) NEW N4 | | consume a corpse to restore health and focus | +- Exhume (Corpse Explosion 40) NEW N13 | | raise workable remains when the field is empty | +- SummonSkeleton (Corpse Explosion 40) | | CHG N10: a slain or expired summon detonates | +- SummonSkeletonMage (Summon Skeleton 40) NEW N7 | | already authored, placed on the tree | +- FeedTheDead (Summon Skeleton 40) NEW N9 | | feed a corpse to a summon to wind back decay | +- TheLastCommand (Summon Skeleton 40) NEW N11 | | consume a summon for its burst and a surge | +- SummonLich (Death Shield 40 + Summon Skeleton 40) | CHG N8: scaling reads the caster's mastery; drain aura | +- DeathTouch (Tap Soul 10) | | CHG N1: the ward rend rider fires for everyone, tears magic warding | +- GraveOmen (Death Touch 20) NEW N5 | | one short window where the ward falls | +- NecroticWeapon (Death Touch 40) NEW N3 | | the death weapon enchant | +- DeathRay (Death Touch 40) | | CHG N1 rider; CHG N17: power stat normalized | +- DeathShield (Death Ray 40) | | CHG N15: mitigation restored, tooltip tells the truth | +- DeathField (Death Ray 40) | CHG N16: cheaper among the dead | +- BloodPrice (Tap Soul 40) NEW N14 | actives that spend health for power | +- DeathMastery (Tap Soul 10) +- Bonepiercer (Death Mastery 40) NEW N2 | passive bonus against undead targets +- UndeadMastery (Death Mastery 40) | | CHG N6: grants the summon life drain stat | | CHG N8: grants the summon boost stat | +- MasteryDeath (Undead Mastery 80) capstone, unchanged | +- WardLife (Undead Mastery bought) capstone, unchanged +- MummifiedUndead (Death Mastery 40) unchanged, decay stays by design school wide, CHG N12: consuming a corpse grants a Grave Essence stack, cap 5
Honesty notes carried with the figure: new stats reuse the four dead stat slots the audit found before minting anything; the two dead skill ids are the reuse pool for new nodes; every mastery term follows the base-plus-mastery grammar so the tree works unspecialized; and nothing on it assumes drain heal-back off undead or the caster's own minions.
The second picture is the school's new resource loop. Today a corpse is a targeting requirement for one spell. Composed, it becomes the school's economy, with one rule doing the balance work everywhere: one corpse feeds one verb.
blood price rail the loop +---------------------+ | CASTER HEALTH | +---------------- KILL <---------------+ | spent by: | | | | | Tap Soul (ships) | | v | | Blood Price N14 | | CORPSE <---- EXHUME N13 | +----------+----------+ | one corpse feeds one verb; | | | first verb to take it wins | lower health raises | (manufactured remains: no | the near-death bonus | loot, no experience) | | | / | | \ | | v v v v v | | EXPLODE SIPHON FEED CHARGE | | CorpseExpl. N4 N9 N12 | | + N17 N18 | | | | | | | +--> more v v v v | damage --> DAMAGE SUSTAIN UPKEEP ESSENCE | aoe plus health summon stacks, | damped + focus decay cap 5 | chain wound | | | back | | | +---------------+----------+-------+------+ payoffs produce the next kill summon rail: a slain or expired summon detonates (N10), so a falling minion re-enters the loop as damage; The Last Command (N11) spends one deliberately; decay is the upkeep clock and Feed the Dead winds it back. rate caps, printed as commitments: one verb per corpse; essence caps at five stacks; chains damp per hop; manufactured corpses pay nothing.
Five phases, ordered so the school visibly improves while the larger pieces are still in design. Phase 1 is deliberately all small data-only cards that fix filed defects: four cards, six defects redeemed, roughly a week of data work with tests.
| Phase | Cards | Why this order |
|---|---|---|
| 1 | N6, N7, N8, N17 | Wake the dead data: the drain stat gets its skill path (#2599), the authored Skeleton Mage takes the field, the lich scaling root cause is fixed (#2600), the dead enchant works (#2598, #2601). Every one lands on an on-record complaint; one shared before-and-after summon sheet publishes with the changesets. |
| 2 | N1, N2, N4 | Crack the wall: the shred that fires for everyone, the anti-undead innate, and the sustain that works in a crypt. This ships the counterplay verbs the community thread names. |
| 3 | N10, N9, N11 | The court's toys, gated on the corpse-burst visual returning (#2606). Detonation first as the primitive, then feeding, which preserves primes, then sacrifice, which spends them. |
| 4 | N12, N13, N16, N18 | The grave economy: essence needs consumers, which phases 2 and 3 provide; Exhume floors the resource; Death Field's discount ships only after its replication check; the chain needs #2606 and #1549. |
| 5 | N3, N5, N15 | The coordination-heavy tail: the necrotic enchant waits on the shared itemization rail with the weapons pages, Grave Omen joins N1's curse channel after PvP review, and the Death Shield rebuild is the page's one real code card and deserves a quiet window. |
flowchart TB
subgraph P1["Phase 1: wake the dead data"]
N6["N6 Hungry Dead"]
N7["N7 Skeleton Mage"]
N8["N8 Lich worth its price"]
N17["N17 Dead enchant fixed"]
end
subgraph P2["Phase 2: cracking the wall"]
N1["N1 Ward Rend"]
N2["N2 Bonepiercer"]
N4["N4 Grave Siphon"]
end
subgraph P3["Phase 3: the restless court"]
N10["N10 Deathknell"]
N9["N9 Feed the Dead"]
N11["N11 The Last Command"]
end
subgraph P4["Phase 4: the grave economy"]
N12["N12 Grave Essence"]
N13["N13 Exhume"]
N16["N16 Death Field pricing"]
N18["N18 Explosion chain"]
end
subgraph P5["Phase 5: the tail"]
N3["N3 Necrotic enchant"]
N5["N5 Grave Omen"]
N15["N15 Death Shield rebuild"]
end
N17 --> N18
N4 --> N9
N4 --> N12
N9 --> N12
N10 --> N9
N10 --> N11
N1 --> N5
xFX["#2606 corpse burst visual"] -.-> N10
xFX -.-> N18
xWP["weapons pages: shared gem rail"] -.-> N3
N8 -.-> xBAL["balance program: lich XP handoff"]
xVER["replication check"] -.-> N16
classDef ext fill:#c9a55a33,stroke:#8a6d33,stroke-width:2px;
class xFX,xWP,xBAL,xVER ext;
Twelve findings from this review. Eleven verified against shipping data and code; one, the wall itself, was already measured by the balance program. Nine are filed as tracker reports independent of any vote: they are bugs whether or not a single card passes.
| # | Finding | Filed | Answered by |
|---|---|---|---|
| 1 | The Corpse Explosion Damage enchant grants a stat no formula reads | #2598 | N17 |
| 2 | The summons' life-drain strike has no skill path; only one artifact belt line grants its stat | #2599 | N6 |
| 3 | The undead summon boost stat has never been granted by anything | #2600 | N8 |
| 4 | The summon health formula reads the summon's own mastery, which is always zero, instead of the caster's | #2600 | N8 |
| 5 | Death Ray's power stat uses a bare multiplier against the school's wrapped convention | #2601 | N17 |
| 6 | Death Shield's active-state stat is never written | #2602 | N15 |
| 7 | An authored Death Ray impact effect is referenced by nothing | #2603 | N3 (donor FX) |
| 8 | A dead skill id is still mapped to the Death school | #2604 | N2 (reuses it) |
| 9 | Summon Skeleton Mage is fully authored and invisible: its school field says None | the fix is N7 itself | N7 |
| 10 | Death Field advertised a minimum range it does not have | #2605, fixed | shipped on its own |
| 11 | Death Shield's damage conversion is disabled in code while the tooltip claims it | folded into N15 | N15 |
| 12 | The undead wall: zero sustained Death damage against tier-5 undead at grandmaster | balance program, wave 1 | group A works around it |
Also filed during the review: #2606, the corpse-burst visual that has been disabled for years with no tracker entry; it gates N10 and N18. Pre-existing reports this page references rather than duplicates: #1709 (summon scaling symptom), #2478 (summon math), #1549 (explosion targeting), #2306 (Tap Soul attunement pass), #1387 (empty tree render), #1391 (bleeds vs immune undead), #1908, #2349, #1106.
This page is part of the proposal series, and one of several magic-school dives running this week alongside the Fire, Life, and Earth pages. Boundaries honored here: the taming page owns pet commands, stances, gear, bonding, and pathing; the summon cards touch only summon-specific stats, and any owner-triggered summon button rides the taming page's proposed ability-bar rail rather than duplicating it. The weapons and enchanting pages own itemization pools; the necrotic enchant's gem placement is decided on that shared rail, next to Earth's weapon-enchant card. The Life page owns healing, cleansing, and prevention mechanics, and proposes the first support band for the published balance targets; the resistance-adjusted band this page's wall argument wants belongs to the same follow-up family. The Fire page carries that school's own anti-undead angle (undead are fire-vulnerable in the data); each school gets its own answer to the wall, and none of them retunes it. The combat balance program owns every number: the undead wall is already on its first tuning waves, the lich experience price is handed to it with vote evidence, and each card's figures are illustrative until its changesets and measurement windows say otherwise.
| Suite the page promises | Covers |
|---|---|
| DeathShieldDurationTests, on the shield-spell pin pattern | N15, plus a baseline pin of today's duration formula |
| TapSoulFocusGainTests, pinning the near-death bonus cap | N14, and every formula reading the bonus |
| CorpseExplosionDamageStatTests, killing the silent-zero risk | N17 precondition, N18 rider extension |
| DeathSummonDrainGrantTests plus the parent-token pin | N6, N8 |
| SkillSet structural audit extension | N7 (the motivating case), N2, N6, N8 |
| NecroticWeaponEnchantTests, on the elemental-enchant template | N3 |
| Unspecced-functionality invariant: every new or edited Death formula is nonzero at mastery zero | N1, N2, N5, N10, N14, N16, N18 |
| Stack pins: shred cap 5, essence cap 5, omen single stack | N1, N12, N5 |
| Live two-player passes, routed to the tester board rather than promised as unit tests | corpse-consume sync (N4, N13), summon heal (N9), summon kill effects (N10, N11), drain exclusion behavior |
Genre claims were verified against primary or well-established community documentation during this review; claims that could not be confirmed from a primary source were cut rather than softened. Principal sources: the UOGuide and official UO wiki Necromancy and Spirit Speak pages and the Age of Shadows publish history; the Guild Wars wiki on minion mastery, Death Nova, and Soul Reaping; the Diablo II Arreat Summit and DiabloWiki curse and Corpse Explosion pages; the Warcraft Wiki on Curse of the Elements and Curse of Shadow and current Unholy guides; Project 1999 and EQ community necromancer references; the Path of Exile wiki on corpses, Desecrate, and offerings; Maxroll's Diablo IV Book of the Dead reference; the Guild Wars 2 wiki on Scourge and Shade; and UESP's ESO Blastbones and Sacrificial Bones pages. In-game findings come from the shipping unity-6.5 data and code, verified line by line during this review; the measured undead-wall numbers come from the combat balance program's published audit.
Every number on this page is an illustrative starting point, not a patch note. Final values ship through the combat balance program. These proposals are the start of the conversation, not the plan of record: comment threads and vote arrows on every card are read before anything is scheduled.
No account needed. Every proposal above also has its own Discussion box for talking about that one alone; this thread is for the plan as a whole. Comments are plain text, held to basic decency, and may be trimmed.