The Patient Tide
A deep dive into Shroud of the Avatar's Water magic, what the genre learned about cold and cleansing from Ultima onward, and twenty-five concrete proposals.
Water is the least-played real school in the game, and the players who do run it mostly are not ice mages at all: they are melee and hybrid characters dipping in for the healing rain, the pet that heals, and the fire extinguisher. That is not a misreading of the school. It is an accurate one. The ice half of Water sells exactly one product, a slow, and the engine erases that slow on the very next cast: every slow in the game shares a single effect slot, so a water mage's chill evicts their own previous chill, and anyone else's slow evicts both. This document measures the school against the genre that invented frost magic, and proposes twenty-five changes built on one idea: let the cold accumulate into something the whole kit can spend, and name the half the players already love — negation, the school that takes fire, poison, and harm away — as the identity it has been since Ultima spelled it An Flam.
The genre never agreed on what water magic is, and that disagreement is the whole design space. Ultima Online typed cold as a thin damage flavor and put control elsewhere. Dark Age of Camelot made ice the control school and attached every snare to a damage spell so control never cost a wasted cast. World of Warcraft's frost mage built the most admired ice kit ever shipped on one trick: "frozen" is a shared state tag, and other spells read it for a payoff. Guild Wars 2 sent water the other way entirely — the healing and cleansing attunement, whose chill slows cooldowns, not just legs. Divinity: Original Sin 2 made water the primer that turns everyone else's spells into combos. And Final Fantasy XIV made ice the phase of the rotation where your resources come back. The verdict across all of them is consistent: cold control is only interesting when it builds toward something, and water support is only distinctive when it removes what the enemy did.
Measured against that bar, our Water school does a weak version of three identities and commits to none. The 2021 specialization telemetry ranks Water 19th of 21 schools — last among the true elements, above only Chaos and Shield — and the most recent balance audit measured its sustained damage at roughly half of Fire's. The community's own school-by-school verdict reads: "Damage is low. Slow just not good enough. The shield needs more incentive to use it over the alternatives." The gaps cluster in three places:
The deal that runs through all twenty-five proposals: Tier 1 fixes what is broken or dishonest, in the open, with the defects filed as bugs the day this page went up. Tier 2 is the keystone — water's slows leave the shared slot and coexist, chill becomes a visible stacking state, three stacks earn a governed freeze inside the crowd-control resist system that PvP already trusts, and a Shatter payoff spends what the kit builds. Tier 3 is the deeper water: the school's first real area spell, a wall, a focus stance, weather, and the water mage owning the water itself. The control half gets a payoff; the negation half gets named; and neither waits on the other.
| # | Proposal | Tier | Size | Server | One-line pitch |
|---|---|---|---|---|---|
| S1 | Cone of Cold checks the right resistance Built | 1 | S | none | the big spell's debuff is typed against the wrong element; one field |
| S2 | Shield of Ice, the typo and the missing minute Built | 1 | S | none | a misspelled formula token and an avoidance buff that dies a minute early |
| S3 | Ice Field tells the truth Built | 1 | M | none | the description names the damage and the resistance strip nobody knew about |
| S4 | Soothing Rain you can see through Built | 1 | M | none | ten years of blindness complaints; players wrote an addon to hide it |
| S5 | Tooltip and doc truth pass Built | 1 | S | none | the school's newest spell is missing from its own documentation |
| S6 | Water expression pin suite Built | 1 | M | none | a school that shipped a case-typo gets pinned tests before any tuning |
| S7 | Douse from the utility bar Built | 1 | S | none | the fire extinguisher works when you are on fire outside combat |
| S8 | Rehydration ships | 1 | S | changeset | the school's dormant focus spell is one disabled flag from existing |
| S9 | Slows that land on the big ones | 1 | S | changeset | a duration floor where the formula collapses to under a second |
| S10 | Slows that stack Built | 2 | M | channel data | water slows hold three deep on their own channel; the ten-year ask, answered |
| S11 | Chill you can see | 2 | L | shared data | chill becomes visible stacks on the target; the keystone state |
| S12 | Frozen, fairly | 2 | L | shared data | three stacks earn a governed freeze, ten years after the removal |
| S13 | Shatter | 2 | M | changeset | cold that pays you back: bonus crit on the chilled, a burst on the frozen |
| S14 | Chill the clock | 2 | L | shared data | deep chill slows the enemy's cooldowns, the slow nothing else sells |
| S15 | Cone of Cold earns its cooldown Built | 2 | M | changeset | the school's one loved nuke stops costing a minute of your life |
| S16 | Ice Field worth its focus | 2 | M | changeset | the genre's only zero-damage field becomes the school's stack engine |
| S17 | Shield of Ice worth the slot | 2 | M | changeset | touch the ice mage, get chilled; the school's first absorb |
| S18 | Two elementals, two jobs | 2 | M | changeset | the healer pet healed; the ice twin becomes the chill battery |
| S19 | The rain that cleanses Built | 2 | M | changeset | the school's best spell becomes its cleanse statement |
| S20 | An Flam: the negation suite Built | 2 | M | shared data | Douse eats the next ignition, not just the last one |
| S21 | Blizzard | 3 | XL | shared data | the school's first true area damage spell; asked for by name since 2016 |
| S22 | Water Wall | 3 | L | shared data | a defensive line that chills whoever crosses it, on a name already reserved |
| S23 | Undertow, the focus stance | 3 | L | shared data | water becomes the mana school: cold casts refill what other schools spend |
| S24 | Rain-slick world Built | 3 | M | none | water magic gains 20% in rain or standing water; wading refuses no spell |
| S25 | Waterborne | 3 | M | shared data | the water mage owns the water: swim at speed, cast while swimming |
The pieces in place are real, and there are more of them than any other magic school: fourteen implemented skills — nine castables, a weapon enchant, and four passives — in a clean prerequisite tree. A ranged slow, a melee slow, a slowing field, a cone nuke, an anti-fire toolkit, a group heal-over-time, a shield, and the game's only school with two summons, one of which heals its owner. Underneath it, the same effect machinery every school shares: stacking rules, effect channels, trigger volumes, and formula-driven scaling on every number.
Ice Fist ............................. [1] ├─ Ice Arrow ....................... [1] │ └─ Ice Field .................... [2] │ ├─ Summon Water Elemental .... [5] │ │ └─ Summon Ice Elemental ... [5] │ └─ Cone of Cold .............. [3] ├─ Douse ........................... [4] │ ├─ Soothing Rain ................ [4] │ │ └─ Rehydration (disabled) .... [6] │ └─ Shield of Ice ................ [7] ├─ Frost Weapon .................... [8] └─ Icy Breath ── Blue Blooded ── Water Mastery / Fire Warding
Combat here is a synchronized simulation run on the players' machines: the server relays which rune fired and a random seed, and every client plays the result forward identically. A slow becomes real on the target's machine. The stage that kills the school is the last one:
flowchart LR
A["Glyph played"] --> B["Rune data\ndamage, slow, duration\nall formula-driven"]
B --> C["Server relays\nrune id + seed\nnever sees the slow"]
C --> D["Every client\nsimulates the effect"]
D --> E["Target applies\nthe movement debuff"]
E --> F{"Effect channel check\none slow slot per target\nnewest evicts oldest"}
F -->|evicted| G["Your previous chill\nsilently removed"]
F --> H["Slow applies\nfloored at 20% speed"]
style F stroke:#3e8fa3,stroke-width:2.5px
style G stroke:#a33d2e,stroke-width:2px
Water magic is the genre's least agreed-upon school, and every disagreement produced a lesson. The through-line that matters most for us: the games whose ice kits players still praise all gave cold a state — something that visibly accumulates on the target and that other spells read — and the games whose water kits players still praise made water the school of taking things away: fire, poison, harm, momentum. Ours currently does neither, and our own history is the sharpest lesson on the page: this game shipped stack-to-freeze ice magic once, watched it lock players down in PvP, and deleted it in 2016 instead of governing it. The community has cited that removal for ten years. The fix was never "bring it back exactly"; it was "bring it back inside the crowd-control rules that every other hard control already obeys."
Ultima Online: cold as a resist-table entry, Paralyze's break-on-damage as the original freeze governor, the water elemental as the hybrid pet — and, from the single-player Ultimas, An Flam: Douse's literal ancestor, water as the syllable of negation. EverQuest: the frost line exists so you have an answer when fire is resisted — cold as the reliability element; its water pet was the fragile top-damage archetype. Dark Age of Camelot: snare-on-the-nuke, root as a hard commitment with a hard cost, and the ice wizard as the kiting solo spec. WoW: Shatter's state-tag-plus-payoff loop, Fingers of Frost's earned procs, and the diminishing-returns system that finally made hard CC governable. Guild Wars 1: the cautionary tale — sixty-six percent snares with no payoff produced the least-played line in the class. Guild Wars 2: Chilled hitting cooldowns, cleanse riding on the heal, auras shared to the party — the complete water-as-support kit. Divinity: Original Sin 2: the wet-chilled-frozen ladder behind an armor gate, and Hydrosophist scaling healing and magic armor off one investment. Path of Exile 1 and 2: chill magnitude scaling with damage dealt, freeze as a buildup meter rather than a coin flip, and Brittle as the crit-payoff on cold. Elden Ring: frostbite as a party-wide damage amp with buildup and decay. Magicka and Noita: water as the interaction hub — soaked doubles lightning, frozen triples damage, wet cannot ignite — and the wet-stain as a consumable buffer rather than a timer. FFXIV: Umbral Ice, the resource stance. Diablo 2 and Last Epoch: three exclusive cold armors with distinct triggers, and chill as readable discrete stacks with a hard cap. ESO: the frost tank, and Minor Brittle as ice's seat in the group.
| Capability | UO | DAoC | WoW | GW2 | D:OS2 | PoE | FFXIV | ESO | SotA today | SotA proposed |
|---|---|---|---|---|---|---|---|---|---|---|
| Slows from multiple sources coexist | ○ | ● | ● | ● | ◔ | ● | ◔ | ● | ○ one slot game-wide | ● S10 |
| Chill visible as discrete stacks | ○ | ○ | ◔ procs | ◔ | ● | ● PoE2 | ○ | ◔ | ○ | ● S11 |
| Hard freeze with a governor | ◔ breaks | ◔ | ● DR | ○ | ● armor gate | ● buildup | ○ | ◔ | ○ removed 2016 | ◔ S12 |
| Payoff on the controlled state | ○ | ○ | ● Shatter | ◔ | ● | ● Brittle | ○ | ● Brittle | ○ | ● S13 |
| Chill slows more than movement | ○ | ○ | ◔ cast | ● recharge | ◔ dodge | ● action | ○ | ○ | ◔ attack only | ● S14 |
| Anti-fire negation suite | ● An Flam | ○ | ○ | ● | ◔ | ○ | ○ | ○ | ◔ strips the last | ● S20 eats the next |
| Heal and cleanse seat | ○ | ○ | ◔ shaman | ● | ● | ○ | ◔ | ◔ | ◔ fire-only cleanse | ● S19 |
| Cold as resource stance | ○ | ○ | ○ | ◔ | ○ | ○ | ● Umbral | ○ | ○ spell dormant | ◔ S8/S23 |
| Defensive ice armor identity | ○ | ◔ | ● barrier | ● aura | ● | ○ | ○ | ● Warden | ◔ no absorb | ● S17 |
| Summons with distinct jobs | ● hybrid | ● snare pets | ◔ | ○ | ◔ | ○ | ○ | ○ | ◔ one visible | ● S18 |
| Field control that ticks | ◔ fields | ● | ● | ● | ● | ● | ◔ | ◔ | ○ 0–1 once | ● S16 |
| True area damage spell | ○ | ◔ | ● Blizzard | ● | ● | ● | ● | ● | ○ | ● S21 |
| Mobility and water traversal | ◔ | ○ | ◔ | ○ | ○ | ○ | ◔ move-cast | ○ | ○ | ◔ S25 |
| Weather reactivity | ○ | ○ | ○ | ○ | ◔ rain | ○ | ○ | ○ | ○ cosmetic | ◔ S24 gated |
● native · ◔ partial or conditional · ○ absent. "SotA today" = the live Water school on the migration branch, 2026-08. Row definitions and per-game evidence live in the research notes accompanying this proposal.
Tier 1 fixes what is broken or dishonest, mostly in data, with the defects filed as tracker bugs the day this page went up. Tier 2 is the keystone chain — the slows coexist, chill becomes a state, the state earns a governed freeze, and a payoff spends it — plus the negation half of the school growing into its name. Tier 3 is the deeper water: the big spell, the wall, the stance, the weather. Every tuning number is an illustrative starting point: final values ship through the combat balance program's test-shard changesets and its measurement windows.
Built 2026-08-14. The debuff is typed water now, the same as the damage half of the spell, so a sun-warded enemy no longer shrugs off the cone's slow while taking its damage in full. It shipped as a bug fix the day this page was published, before the page had a way to say so. The sweep the card asked for was run across every rune in the game and came back clean: no other water or ice spell carries a sun type, and Cone of Cold was the only one that ever did. It did turn up a different wrong element on Ice Field, which S3 fixes.
Cone of Cold's attack-speed debuff is typed against the wrong element: the damage half of the spell correctly checks water resistance, but the debuff half carries a sun damage type — a copy-paste slip in the spell data — so the school's signature debuff gets resisted by sun resistance instead of water's. Against sun-resistant enemies your cone quietly does less than the tooltip has any way of expressing. One field changes, and the fix ships through the balance changeset process with a regression pin so it stays fixed.
Seen in: every live game's audit history; this is the classic wrong-school coefficient bug.
Pure data: retype the debuff's damage type to match its sibling damage effect. Filed as a tracker bug alongside this page. Before closing, sweep the other schools' multi-effect spells for the same paste — this is the second wrong-element slip found in water data this year, after the wild elementals read the wrong mastery stat.
Built 2026-08-14. The misspelled variable is gone, and every buff a single cast grants now runs the same length, damage avoidance included. One correction to the card: the missing minute was already fixed on 24 July, five weeks before this page said it was still live, and refined again on 9 August so the timers also line up at low skill. The typo was the half still standing. Both halves are now held by tests that read the formula text as well as the compiled form, which is the gap the typo lived in.
Two defects in one spell. First, one of Shield of Ice's formulas contains a misspelled variable that only evaluates because the formula parser happens to ignore letter case — a live typo in shipping spell data. Second, the shield's damage-avoidance buff expires after one minute while every other buff the same cast grants runs two — an asymmetry players reported in December 2021 that is still live today. Fix the token, and give the avoidance buff the same duration expression as its siblings.
Seen in: the community bug tracker, submitted December 2021; the wiki still documents the mismatch.
Both changes are formula edits on the shield's spell asset, shipped as one changeset so the expressions regenerate through the real tokenizer — hand-editing the raw text is a silent no-op, which is precisely how a typo like this survives for years. The pin suite (S6) adds a case-sensitivity check across every water formula.
Built 2026-08-30. Ice Field's description now says what the spell does: it names the water damage it has always ticked, and it names the water resistance that Water Specialization strips from anyone caught in the field. That second one was invisible. It only fires for a specialized caster, and the tooltip hides a line that cannot fire for you, so a player had no way to learn it existed. The two stat effects were also typed part Moon, the only spell data in the game carrying that value, which is fixed here.
Two things on the card turned out not to be what they looked like, and both are worth stating. The 37.8 seconds and the 10 seconds were never the same clock. 37.8 is how long the field stands on the ground; 10 is how long the chill lasts on one enemy, and it is reapplied on its own timer while they stay inside. Nobody was reading a wrong number, they were reading two right numbers as one, and the description now separates them. The reapply half was fixed on 14 August. The 70% against a measured 45% is not an Ice Field bug. Every percentage debuff in the game is estimated in its tooltip at the ceiling of the caster's gear rather than at what the cast will land, so every one of them reads high. That is a game-wide tooltip question and it did not ship here.
Ice Field is the most bug-reported spell in the school, and this week it earned another one: the tooltip promises a 37.8-second debuff and the field delivers 10. In 2022, players measured a promised 70% slow landing at 44.9%. And the slow applies only at the moment an enemy steps in — stand in the field all day and it is never reapplied, a known quirk of how persistent area effects process their targets. This card is the honesty pass: fix the reapply logic, make the tooltip's numbers the numbers, and document what the field actually does — including the part nobody knows, that it already ticks a small amount of damage and, under specialization, shreds water resistance.
Seen in: reported by Barugon this week, measured by Violet Ronso and Chiasma in 2022, and raised as far back as 2021.
| What the card said | What it turned out to be | |
|---|---|---|
| Debuff duration | 37.8 s promised, 10 s delivered | two different clocks: 37.8 s is the field, 10 s is the chill on one enemy |
| Slow at high investment | 70% promised, 44.9% measured | every percentage debuff in the game estimates high in its tooltip; not an Ice Field bug |
| Slow while standing inside | on entry only | real, and fixed on 14 August |
| Damage and resistance shred | nobody knew they existed | real, and now named in the description |
The reapply fix rides a known one-line defect in the shared area-effect processing code — the same fix benefits every persistent aura in the game, and this card is water's claim on it. Reapply-while-inside is a real power increase wearing a bug-fix hat, so the tuning offset ships in S16's changeset. Debuffs are target-authoritative: the standard two-client verification applies.
Built 2026-08-30. The rain emits less than half of what it did. The two layers that sit between you and the sky took the deepest cut, because they are what fills the camera in a doorway or a cellar: the cloud layer went from 25 particles a second to 8 and the rain sheet from 5 to 2, with the sheet also drawn fainter. The rain itself went from 33 a second to 20, so it still reads as rain. The splashes and ripples on the ground were left alone, since that is what tells an ally standing in it that the rain is on them. Worst-case particle counts across the whole effect dropped from five thousand to about a thousand.
Two things the card asked for that are not in this build, said plainly. The dense version is not an option, it is simply gone. A toggle would have meant a new settings row and a second copy of the effect, and the thinner rain was judged good enough that nobody should have to go looking for a switch. If players disagree, that is worth hearing. And Cleansing Rain uses the same effect, so it thinned along with Soothing Rain. That is the right answer for two spells that are both rain, but it was not a separate decision.
Soothing Rain is the best-loved spell in the school and its visual effect has been generating complaints for a decade: it blinds the caster in tight spaces, it has collapsed framerates on busy lots, and the community eventually wrote an addon whose purpose is to hide it. When your school's signature spell is something players pay to not see, the art is the bug. Rebuild the effect at a fraction of the particle density, keep the silhouette readable from the healing radius, and make the dense version an option rather than the default.
Seen in: complaint threads from 2017 through this year, and the FXStop addon the community built in self-defense.
Presentation-only, no sync constraints: retint and thin the existing emitters rather than authoring new ones, and verify the rebuilt prefab's asset-bundle membership before shipping — the fire page's audio pass hit exactly that trap. Pairs with S19, which grows the same spell mechanically; the two ship in either order.
Built 2026-08-30. Checking the four claims on this card found three of them already answered. Ice Fist's tooltip has named its unarmed-combat bonus since the game's code was first mirrored. Ice Arrow's tooltip prints no slow duration, so it never promised one. And Water Magic Specialization has always stated the fire resistance it costs you. The tooltip half of this card was delivered by the whole-school description sweep of 15 August, which corrected four Water strings. What was stale was the school reference, and that is what changed: Frost Weapon and its on-hit effect are written down for the first time since the spell shipped, Ice Field's entry now says the field damages and, under specialization, strips 20 points of water resistance rather than only slowing, and Shield of Ice's listed duration caught up with the buff timers that were aligned in August. The pass also turned up a wrong element on Ice Field's slow and resistance strip, filed as its own bug against the spell data.
The school's own documentation is missing its newest spell: Frost Weapon — the weapon enchant the community asked for on three separate streams before it shipped — appears nowhere in the school reference. Ice Fist's tooltip does not mention that it already scales with unarmed-combat skill, which quietly answers a standing wishlist request. Ice Arrow's slow duration collapses against strong enemies while the tooltip promises a flat number. And Water Mastery's fire-resistance tax appears nowhere a player can read it. Say what the school actually does, everywhere the school says anything.
Seen in: the pattern every school page in this series has needed; fire's was S8.
Strings and documentation only. The localization export pipeline still runs through the legacy setup, so the standard export-path verification comes first — the same caveat every school page has carried. The duration-collapse half of the story gets its mechanical fix in S9; this card makes the tooltips stop lying about it in the meantime.
Built 2026-08-30. Six automated checks now cover every formula on the school's eleven spells, the five combo glyphs water takes part in, and everything those spells fire underneath them. Each formula is re-read through the game's own parser and compared against the compiled version the game actually runs, which is the gap the Shield of Ice typo lived in for years. The checks also reject any water formula that reads another school's specialization, and evaluate the whole school both untrained and at grandmaster, so a spell cannot quietly stop working for anyone who has not specialized. One correction to the card: three water spells already had tests of their own before this. They stay. They pin specific numbers, where this suite pins the shape of every formula in the school.
A school that shipped a case-typo in one formula and a wrong-element type in another needs its formulas pinned before anyone starts tuning them. This card adds an automated test for every water spell expression — evaluated untrained and at grandmaster, with a guard that re-parses each formula through the real tokenizer — plus a structural audit of the school's skill-sheet rows and its five cross-school combo definitions. Every changeset on this page lands on top of these pins.
Seen in: the fire page's S9; the one water formula test that exists today covers a single spell.
Editor-mode tests in the project's usual style. The pin evaluates each expression at specialization zero and grandmaster so the "works untrained" invariant holds numerically, and the tokenizer round-trip catches exactly the class of defect S1 and S2 fix.
Built 2026-08-30. Douse now sits on the utility hotbar. Drag it there and it can be cast with your weapon away, without starting a fight, so catching fire while fishing or stepping into lava no longer means drawing a sword to put yourself out. A correction to an earlier version of this note, which said Douse had stopped being party-only and now worked on any friendly player: that was wrong. Douse lands on yourself and on people in your party, the way it always has, and so does the rest of the healing line. Two things stayed as they were on purpose. Once you are in a fight the utility-bar copy greys out and Douse is cast from your deck exactly as before, which keeps its cooldown and focus cost honest; and the spell itself is unchanged, same cleanse, same fire resistance, same 3 focus. The flags it now carries are the ones Soothing Rain, Purify and the healing line have had all along.
Douse is the cheapest spell in the school and the one with the clearest job: put the fire out. But it can only be cast in combat — fall in lava while fishing, catch fire outside a fight, and the fire extinguisher is grayed out. Let Douse sit on the non-combat bar and work anywhere, the way the school's healing rain already refuses to start fights. Requested by Lained on the forums, alongside the memorable observation that a crafted bucket of water already does this job better than the spell.
Seen in: utility casting in every era of the genre; requested by Lained.
One usability flag on the spell's data, using the same no-combat-trigger convention Soothing Rain already carries. Verify the cast cannot flag you into combat from the utility bar.
There is a finished Water spell hidden in the game's data: Rehydration, a focus-transfer spell — you spend your own focus and a slice of it arrives on an ally, with a small heal under specialization. The spell asset exists, the localization exists, the tree seat exists; a single disabled flag keeps it out of the game. Every school page in this series has found one of these — Earth's cut wall, Life's dormant ankh — and water's is the one that answers a real wishlist ask: more focus tools in the school whose rain already restores health. Flip the flag, tune the transfer ratio, ship the school's twelfth active.
Seen in: EverQuest's mana-transfer lines; the wishlist ask for more focus-regen skills.
The flag flip is one cell in the skill sheet; the tuning is a changeset on the authored formulas. The transfer ratio is the design risk — group play can loop focus through a dedicated battery character — so the ratio starts lossy and tunes against the balance bands.
Water's slow duration is a formula that subtracts the target's strength, and against anything big the subtraction wins: players measured slows lasting half a second against tooltips promising eight and a half. A GM water mage discovers that trolls, dragons, and bosses — the fights where control matters — are precisely where the school's identity stops working, while an archer's Disabling Shot keeps its listed duration everywhere. Add a floor to the duration formula so a landed slow always does something visible, and leave the deeper resistance-wall question to the balance program that already owns it.
Seen in: reported by Morph in 2017 and Lord Baldrith in 2018; the boss-immunity question is a committed balance-program item, credited here rather than re-proposed.
One term added to the duration expressions on the two arrow spells via changeset, within the standard delta cap. The floor value is a boss-control policy decision the changeset review makes explicitly; this card's job is only that "resisted" never silently means "nothing happened while the tooltip said otherwise".
Built 2026-08-31. Water's slows now sit on their own channel, WaterChill, which holds three at once. Ice Fist and Ice Arrow hold together, an archer's Disabling Shot no longer erases either, and the field's chill stopped being something the engine could take away. Because three slows now add up, each one was trimmed by a quarter in the same change: one slow leaves a target at 78% speed, two at 56% (about as strong as the single slow you got before), three at 26%. The 20% floor is still there and is now what it was meant to be, the emergency stop rather than the normal result.
Two things the card said turned out to be wrong, and one of them was hiding a bug. Ice Field's slow was never on the shared channel at all, so it was neither erasing other slows nor being erased. What the field did carry was a channel entry on the spell that places it, and that entry worked against the caster: a slow landing on you could take the field with it, and casting the field could strip a slow you were carrying. Clearing it fixes a fault nobody had reported. Frost Weapon's on-hit chill also stayed where it was; it already had its own channel and never needed moving.
The oldest and most-requested water fix in the game's history, and it is one line of effect-channel data. Today every slow on a target — yours, your party's, an enemy bard's — occupies a single shared slot, newest wins. Cast Ice Fist then Ice Arrow and your fist's chill vanishes; an archer's shot erases both. Proposed: water's slow effects move to their own chill channel with three slots, so a water mage's slows coexist with each other and with the rest of the world's. The eviction that has silently deleted the school's identity since 2016 simply stops. Requested by Garfunkel Humperdinck, originated by Lord Baldrith in 2016, championed by Violet Ronso, and reposted to the wishlist as recently as July — and conceded on a livestream years ago as something that should likely change.
Seen in: DAoC and WoW, where slows from different sources have always coexisted; Last Epoch's capped chill stacks are the modern form.
A new channel entry in the effect-channel registry (the multi-slot pattern is proven: the food channel runs six slots today) and a channel reassignment on the water slow sources. Archery, chaos and thrown slows deliberately stay behind on the old channel so neither side evicts the other. Three simultaneous slows is a real power change: the per-slow magnitudes rebase in the same changeset, and the built-in minimum-speed floor (nothing slows below 20% speed) is the standing anti-permaslow governor. A cross-source matrix test pins who evicts whom.
As built: seven spells moved, not the eight this card first counted. Frost Weapon's on-hit chill was already on its own channel, and the bard's slow was never on the shared one, so neither needed touching. The three spells that place an ice field had their channel entry cleared instead of moved, because the slow belongs to the field itself and the entry on the placement spell was doing harm.
Fire's page proposed making the burn a visible, stacking state; this is water's mirror, and the schools rhyme on purpose. Every water slow — the arrows, the field's ticks, Frost Weapon's strikes, the Ice Elemental's claws — also adds one pip of Chilled to the target, visible on the nameplate, capped at three (five under specialization). Chill pips are the state the rest of the kit reads: Shatter (S13) spends them, the deep freeze (S12) triggers at the cap, and the field (S16) becomes the fastest way to build them. For the first time, a water mage can look at a target and see their work accumulating instead of watching it evaporate.
Seen in: Path of Exile 2's freeze buildup, Last Epoch's capped chill stacks, WoW's Frozen as a state every frost spell reads.
The same engineering recipe as fire's burn state, deliberately: one new replicated character stat appended at the enum tail (mirrored to both code trees under the shared-source discipline; the enum is append-only because formula tokens store ordinals in roughly 1,664 serialized spell assets), a counter sub-effect on each water slow rune whose stack cap is itself a formula — so specialization raises it in data — and pip UI on the nameplate and buff bar. That UI also cures a standing complaint for free: the Ice Elemental's slow, invisible today, becomes pips like everything else. Counterplay ships in the same card: chill joins the existing slow-removal cleanse list, so every skill that removes slows removes it. Stat effects are target-authoritative; this is the desync-prone class of change, so the mandatory live two-client verification applies — same checklist fire's card carries.
Ten years ago you could stack water's slows to a standstill and freeze a target solid; it was removed in 2016 because unlimited player-lockdown was miserable, and nothing ever replaced it. The community has never stopped asking — Lord Baldrith sketched "Ice Ball" and "Ice Storm" freeze spells in the original 2016 thread and offered to trade damage for the identity; majoria70 asked for a freeze moment on Cone of Cold. This card brings the freeze back inside the rules: at full chill pips, the next water hit has a chance to Congeal the target — a short, hard freeze that enters the same crowd-control resist system every stun and root already obeys, with the governor the 2016 version never had: reduced duration against players, it breaks on meaningful damage, and it grants a freeze-immunity window that consumes all pips, so the state machine resets rather than chain-locking.
Seen in: Divinity's chilled-to-frozen ladder behind a gate, Path of Exile 2's buildup freeze, WoW's diminishing returns as the governor, Ultima Online's paralyze breaking on damage.
| The 2016 freeze | Congeal |
|---|---|
| Stack slows to 100%, indefinitely renewable | Triggers only at full pips, then consumes them |
| No crowd-control resist participation | Rides the same resist system as stuns and roots |
| Same rules against players and monsters | Reduced duration in PvP; breaks on damage |
| No immunity window; chain-lock possible | Freeze immunity follows every freeze |
The freeze is a root-family effect, which means it enters the existing crowd-control resist path with zero new governor code — water's first hard control simply inherits the rules. The PvP halving, damage-break threshold, and immunity duration are all changeset-tunable data shipped in the same changeset as the state itself, never after it. Boss policy rides the same resist stat bosses already use against stuns.
Control that pays you back. Once chill is a readable state, water's damage spells learn to read it: bonus critical chance against chilled targets scaling with pips, and against a frozen target, a burst that consumes the freeze for the school's big moment. This is the missing half of every complaint about water's damage — the school does not need a flat buff, it needs its damage to arrive through its identity, the way the genre's most admired ice kit has worked for twenty years. Deliberately tuned to roughly +30 to 50% crit chance at full pips, not the doubling some games shipped: the payoff should reward the build-up, not delete the health bar.
Seen in: WoW's Shatter, Path of Exile's Brittle, Elden Ring's frostbite burst, ESO's Minor Brittle as the group seat.
Pure formula data once S11 exists: crit and damage terms on the existing water damage spells read the target's replicated pip count, and the frozen-consume rider hangs off S12's state machine. Expected value is sized so shattering early roughly equals letting the chill ride — a timing choice, not a mandatory rotation. PvP burst gets the standard duel-window review before the changeset locks.
A movement slow is worthless against a mage who stands still, a turret, or anything that fights where it stands — and after the bard rework moved the unresistable Psalm of Stagnation behind bard requirements, Barugon asked this month for a working water alternative. This is it, and it is a slow nothing else in the game sells: at full chill pips, the target's ability cooldowns recover more slowly and its casts take longer — the enemy's clock chills, not just its legs. Guild Wars 2 proved the design at 66%; we propose 20 to 35%, because that number needs to be respected, not feared.
Seen in: Guild Wars 2's Chilled, the genre's only major cooldown-slow, universally rated its most distinctive control.
The largest code surface in Tier 2: a second appended stat read by the cooldown-recovery and cast-pacing math, applied only at pip cap so it is an earned state rather than an on-hit tax. Enemy cast-loop behavior needs an AI tolerance pass — a monster whose cast never comes off cooldown should degrade gracefully. Same two-client discipline as S11.
Built 2026-08-31. Specialization pays its face value now. The wait falls from sixty seconds untrained to forty at Water Magic Specialization 40, thirty at 80 and twenty-four at 120, where the old formula never got below thirty no matter how deep you went. The base minute is kept, because the spell is meant to be a moment. Three smaller things came with it. The spell needed four times the experience per level of every other water spell, which was never a design decision, and now needs what its siblings need. The rotate keys did nothing during the two-second cast, because turning and walking were one permission and the spell roots you: it can now be re-aimed while it casts, though strafing or walking still cancels it, as before. And a sapphire-socketed chest piece can roll Cone of Cold Power, on the same terms as the Solar Flare enchant sun mages have had for years. The chill contribution the card also promised waits on S11, which is not built yet.
Cone of Cold is the only offensive water spell players write love letters about, and every letter ends the same way: the cooldown. Sixty seconds base; where other schools' specializations push their big spells under thirty seconds, water's can barely reach it. Violet Ronso's feedback thread did the math and a developer acknowledged it; the stated intent was a spell used "sparingly during key moments" — so this card keeps the moment and returns the minute: cooldown rebased so deep specialization lands near 20–25 seconds, the cast no longer cancels when you turn, the spell's slowest-in-school experience curve gets normalized, and a Tempestry chest enchant scales its damage the way Solar Flare's already does for sun mages.
Seen in: requested by Violet Ronso and Ysold; the enchant ask mirrors sun's existing Solar Flare chest piece.
All changeset: the cooldown expression, the turn-cancel lock mode, the experience factor, and one enchant row feeding the spell's existing power stat. S1 lands first — a shorter cooldown multiplies the mistyped debuff's uptime, so the retype precedes the buff. Per-target chill contribution (it grants pips like everything else) makes "sparingly at key moments" true in play: you save it for the freeze window.
Ice Field costs 70 focus at skill level 100, climbing past 100 focus at high levels — Barugon's wishlist request this month asked why the school's field costs more than most schools' finishers. What it buys today: 0–1 damage, once, and a slow applied only on entry. The genre has no other zero-damage field. Proposed, on top of S3's honesty fixes: the focus curve rebased to match what the spell does, small damage ticking every second instead of once per lifetime, and — once chill exists — one pip per tick to anyone standing in it. Violet Ronso once wrote that if Ice Field worked like the throwable Caltrops, the water kit would be perfect. This card is that, plus the stack engine the keystone needs.
Seen in: requested by Barugon and Cairo Hayden; DAoC and WoW fields have damaged and chilled per tick for two decades.
Focus and tick cadence are changeset edits on the field's sub-effect; the pip-per-tick rider is one more sub-effect once S11 exists. A ticking field that feeds freeze is the strongest zone-control loop on this page — the target cap and tick rate are the balance levers, and the area-effect polling cost stays unchanged by design (same instance budget, same poll rate).
Only one shield spell can be active at a time, and nobody picks water's: it grants stat buffs — no absorb, no reaction, nothing cold about it — while Shield of Air holds the slot across the whole player base. Give the ice shield an ice identity: melee attackers who strike you gain a chill pip and a brief slow — the classic cold-armor reaction, sixty years old and still the best expression of "you touched the ice mage" — plus the school's first small damage absorb. The community audit's exact words were that the shield "needs more incentive to use it over the alternatives"; this is the incentive.
Seen in: Diablo 2's Shiver Armor, ESO's Warden frost tanking, Guild Wars 2's Frost Aura.
Keeps the exclusive shield channel exactly as is — the card competes for the slot, it does not escape it. The retaliation trigger needs one verification: the school's weapon-enchant proc vocabulary is outgoing-only today, so an incoming-hit trigger may be the card's one piece of engine work. Retaliation resolves on the attacker's client; two-client verification. The known anti-synergy with defensive melee stances gets documented rather than silently kept.
Water is the only school with two summons, and it shows in opposite ways. The Water Elemental — the pet that heals its summoner — is the school's best-loved asset, loved enough that Barugon asked this month for the Air Elemental to copy its design; it is also still recovering its reputation from a 2019 nerf that hit new players hardest. The Ice Elemental is the neglected twin: its slow exists only under specialization and is invisible on the target. Proposed: restore the healer's cadence to where the fights players remember actually needed it, make the ice twin's slow baseline and visible, and — once chill exists — its claws feed pips, making it the school's control battery while its sibling stays the sustain pet. Two summons, two legible jobs.
Seen in: EverQuest's per-element pet archetypes; DAoC's ice pets that snared with every cast; reported by Rinaldi and Silmaris Dragonbane.
Strictly the water spell-data lane: the heal cadence and amounts on the water elemental's heal spell, the de-gating and pip rider on the ice elemental's attacks. Pet power, AI, and behavior belong to the taming rebalance and stay untouched here — the fence both pages respect.
Soothing Rain already has a secret: under specialization, each tick has a chance to strip a fire effect from whoever stands in it. Grow that seed into the school's cleanse statement: each tick may remove one hostile effect — any school's, not just fire's — with an internal cooldown per target so cleansing is relief, not immunity. At grandmaster, the rain grants the modest fire resistance players proposed a decade ago, and under specialization allies standing in the rain enjoy a continuous douse. Water becomes what Guild Wars 2 proved a water school can be: the place your party stands when the debuffs are flying.
Seen in: Guild Wars 2's Healing Rain, the genre's benchmark area cleanse; proposed by Waxillium and Curt in 2017.
Built 2026-08-31. All three pieces shipped. The douse is continuous now: with Water Specialization trained, fire damage over time comes off everyone in the rain on every tick, where before it was a roll that landed about one tick in eight. The general cleanse is deliberately a low roll — about 10% a tick at specialization skill 100, about 20% at skill 200 — because what it takes is any harmful effect from any school rather than a fire one, and it takes only one per tick. After it removes something from you, it will not remove another for 12 seconds, and that wait is shared by every rain falling on you, so standing in three rains clears you no faster than standing in one. That wait is a live tunable and moves without a build. At Soothing Rain skill 100 the rain grants 10 points of Fire Resistance while it falls, which is the piece Waxillium asked for in 2017.
The cross-school question this card opened is answered, and the fire and life pages inherit the answer rather than writing their own: a cleanse may take any damage over time, any stat pushed the wrong way for the person carrying it, and any resistance debuff, and it may never take something cast on you by a friend.
Changeset work on the rain's existing tick sub-effects. The general cleanse crosses school fences — what it may strip from bard songs and death magic is the same cross-school cleanse-rules decision the fire and life pages opened, and one coherent rule ships for all three pages, not three rules. The internal cooldown is the balance lever.
Built 2026-08-31. Douse now leaves a charge behind, and the next fire effect that tries to take hold on you is swallowed by the wet instead of landing. Two decisions worth stating plainly. It eats any hostile fire effect, not only the burn — a fireball's direct hit counts, so a Douse cast before the fight can take a fire mage's opening spell off the board entirely. That is the bigger swing of the two readings this card allowed, and it is the one that matches what the card promised. And you get one charge, or two if you have taken Water Specialization, lasting as long as the fire resistance the spell already grants. The cure Douse always did is untouched: it still strips the fire that has already caught, and it still gives the resistance buff.
One thing the card called its own review point, said plainly: this is a real change to how a fire mage opens a fight in PvP, and it wants watching on the test shard before anyone calls the numbers settled. Casting Douse on someone who is already doused refreshes the charge rather than stacking a second one, and a dispel can still strip it.
Douse is the oldest spell in the school's lineage — it is Ultima VI's An Flam, "negate flame", and it has been underwhelming since 1992, when a torch double-click did the same job. Today it mops up one fire effect that already landed. Proposed: Douse leaves behind a Doused charge — the next fire effect that tries to take hold is consumed by the wet instead of landing. Cure becomes prevention: cast it on yourself before the dragon, on an ally as the fire mage winds up, and the school's fight-fire identity finally works the way a bucket of water actually does — it is spent stopping something. The existing resist buff and strip stay; the charge is the new verb.
Seen in: Noita's wet stain that fire must burn through, Magicka's soak rules, Ultima's own douse lineage.
The one genuinely new mechanism in Tier 2: a consumable charge buff that intercepts the next matching effect application — likely one new effect type, appended at the enum tail like everything else. It must decrement fire's proposed burn-stack state coherently; the ordering contract between the two pages ships as one decision. PvP counterplay swing (a fire mage's opener can be eaten) is the review point before the changeset locks.
"Why isn't there a blizzard spell for the water magic tree?" — asked on a 2020 stream, echoed by Vallo Frostbane in 2016 ("give us some kind of blizzard, something more in line with the other schools") and by EveMidnight ("can Water Magic get an AoE like Meteor Shower?"), designed by Lord Baldrith in the original 2016 thread as "Ice Storm". The school's first true area damage spell: a storm that falls in waves over several seconds, each wave dealing water damage and adding a chill pip to everyone beneath it, with a target cap. It is the largest power addition on this page, which is why it ships last — behind the full keystone and its governor — and why every number in it goes through the balance bands before it goes anywhere near a shard.
Seen in: Diablo 2's Blizzard, WoW's Blizzard chilling per tick, and the community's own designs, credited above.
The full new-spell recipe the school's own Frost Weapon shipped through: a new skill id (appended, both code trees), a skill-sheet row seated after Cone of Cold, the spell asset, an icon, and localization. Multi-wave delivery reuses the existing timed sub-effect machinery; no new engine work. The focus cost and wave cap are the levers that keep it a moment rather than a mow.
A defensive line of standing water: enemies who cross it are chilled and slowed; allies who stand behind it shed fire effects, as if under a thin rain. To be plain about what it is not: the wall does not physically block anyone — this game's combat runs on the players' machines and the server holds no terrain, so a true barrier is the one wall design we cannot build. What we can build is a line it costs something to cross. The name has been sitting reserved in the game's data for years, unused; this card finally spends it.
Seen in: the wall-spell fantasy from Ultima Online's field spells to D&D's Wall of Ice, translated to a positionless engine.
The cheapest new spell in the series: the skill id already exists in both code trees and is already claimed by the water school's own code — no enum work at all. What remains is a skill-sheet row, the spell asset on the existing persistent-area template, an icon, and localization. Crossing detection is the same trigger-volume machinery Ice Field uses today.
Final Fantasy XIV's black mage burns mana in fire and refills it in ice, and that single loop is the genre's cleanest answer to "why cast the weak element": because it is the phase where everything comes back. Water's version: Undertow, a stance in which water casts build a focus-regeneration current that any school's spells then spend. Pair it with Rehydration (S8) and water becomes the game's mana school — the tree a hybrid caster trains not as a tax but as the half of their rotation that pays for the other half. It also finally gives Water Attunement a job beyond scaling slows nobody could keep on a target.
Seen in: FFXIV's Umbral Ice; Guild Wars 2's attunement passives.
Built on the innate-ability package system (formula-driven, equipment- and state-gated) that only two schools use today, seated on a water passive's reward slot — or a stance skill if the package route proves too narrow. The abuse case is building the current in water and dumping it into another school's biggest spell; combat-only gating and a stack decay window are the levers, tuned against the focus economy bands.
Built 2026-08-31, and the design changed on the way — votes on this card were cast on the version below it. Water spells and water heals gain 20% while it is raining where you stand, and also while you are standing in water. The second trigger is the new half, and it is what let the card ship now instead of waiting: a river is weather you can walk into, so the bonus is reachable today in scenes that never rain. The weather-frequency pass is still worth doing and is still the world team's to schedule; this no longer waits on it.
Standing in water turned out to be unreachable when we went to build it. The game treats shallow water as swimming, and swimming refused every spell, so a mage standing ankle deep in a stream was refused everything. Water spells can now be cast while wading. Every other school is still refused there, and deep water where you are actually swimming still refuses everything, water included: down there your character has their weapons away and their arms belong to the swim animation. Lifting that is S25's question.
Amended 2026-09-13. Keeping the other schools refused in shallow water was a mistake. Once the game learned to see ankle deep water on 2026-09-08, that refusal fired at every river shoreline, and a tester found that standing at the edge of a stream blocked every spell and skill that was not Water. Wading now refuses nothing: every school casts in shallow water, and Water magic keeps its 20% there. Deep water where you are actually swimming still refuses everything except /stuck.
Two honest notes. Snow counts, on purpose — precipitation falls as snow in the cold regions and this is the school that throws ice. And a player town with a weather machine set to rain gets the bonus permanently, which we are leaving alone for now and watching.
Jack Knyfe proposed it on the forums in 2021: water magic should be more potent when it rains, the way sun and moon magic already answer to the sky. The engine can do it honestly — weather here derives from the server clock, so every player's client agrees on the rain. The dependency is the world itself, and the thread said it best: "I can't remember the last time I saw rain in an adventuring area." So this card is two changes in strict order: first a weather-frequency pass over adventure scenes (and a ruling for underground ones — Alley Oop's question), and only then the bonus, because a weather bonus in a world without weather is a dead stat.
Seen in: proposed by Jack Knyfe; the constraint named by Turk Key and Alley Oop; sun and moon worship are the in-game precedent.
The bonus is one conditional term on the water damage and heal formulas, reading scene weather state and the surface underfoot. It follows the same shape sun magic already uses for its daylight bonus, so it costs the formula language one new term rather than any new machinery, and the whole thing sits behind a live switch that can take it back to zero without touching spell data.
One claim in the original card did not survive the build. The weather-linked stat we cited as precedent turned out to be a dead hook: it biases what weather a player sees rather than reading it, nothing in the game's data grants it, and nothing else reads it. It is not a foundation, and the bonus does not use it.
The water mage cannot swim. Like everyone, they slow to three-quarters speed in the water, their spells locked, their weapons sheathed. Air has Dash and Blink; water — the school named for the element — has no mobility verb at all. Waterborne, an innate on the Blue Blooded line: swim at full speed, cast water spells while swimming, and come ashore with a few seconds of slow-immunity, the tide letting go of you. It is honestly niche — that is why it is Tier 3 and small — but every ranger who has fallen in a river mid-fight knows exactly the moment it exists for, and no other school can have it.
Seen in: enduring-breath and waterwalk lines from EverQuest and D&D onward, translated to self-buffs.
An innate reward row plus two client checks: the swim-state speed penalty waived for the owner, and the swimming cast-lock lifted for water-school spells only. The shore immunity is a short self-buff; in PvP its duration stays in single-digit seconds so it is an exit, not an escape kit.
Two versions of the same thirty seconds against the same wolf. The left window is today, and its two quiet lines — a chill fading the moment its sibling lands — are the whole diagnosis. The right window is this page composed: four sources of cold coexisting and counting, the sustain half running underneath untouched, a payoff spent at the top of the stack, and the freeze failing politely against the resist system, exactly as governed control should.
Obsidian Wolf Slowed Health [############ ] 55% You cast Ice Fist. The wolf slows. You cast Ice Arrow. The wolf slows. Ice Fist's chill fades. Kara's Disabling Shot slows the wolf. Your Ice Arrow chill fades. You cast Cone of Cold. 4 water damage.
Obsidian Wolf Chilled oooo Health [########## ] 47% You cast Ice Fist. Cold clings. o You cast Ice Arrow. It deepens. oo Frost Weapon strikes. ooo Your Ice Elemental rakes the wolf. Your Water Elemental soothes you for 18. You cast Shatter. Deep cold breaks for 61 water damage. oooo The wolf resists the deep freeze.
| Wave | Ships | Rule |
|---|---|---|
| W0 · Truth | S6 pins, S5 strings; the audit bugs behind S1, S2, and S3 filed the day this page goes up. | Every formula edit lands on pins; the localization export path gets verified first. |
| W1 · The thaw | S1, S2, S9 as one changeset batch; S3, S7, S8 alongside; S4 on the parallel art track. | Fix-first before any new power; each fix published with what was broken. |
| W2 · The state | S10, then S11, then S12, then S13 — strictly in order. | The channel before the stacks, the stacks before the freeze, the governor inside the freeze's own changeset, the payoff last. |
| W3 · The kit | S14 through S20. The stack-readers (S14, S16's pips, S17's retaliation, S18's battery) follow S11; S15, S19, S20 are independent. | S20's ordering against fire's burn state ships as one cross-page decision. |
| W4 · Deep water | S22, S23, S25; S24 behind the weather-frequency content pass; S21 last of all. | The biggest power addition ships behind the full keystone and its governor, never before. |
flowchart LR W0["W0 pins + truth"] --> S10["S10 slows coexist"] S10 --> S11["S11 chill stacks"] S11 --> S12["S12 Congeal, governed"] S12 --> S13["S13 Shatter"] S11 --> S14["S14 chill the clock"] S11 --> S16["S16 field = stack engine"] S11 --> S17["S17 shield retaliation"] S11 --> S18["S18 elemental battery"] S3["S3 Ice Field truth"] --> S16 S1["S1 CoC retype"] --> S15["S15 CoC cooldown"] S12 --> S21["S21 Blizzard"] S13 --> S21 FIRE["fire page burn state"] -.ordering.-> S20["S20 An Flam charge"] S8["S8 Rehydration"] --> S23["S23 Undertow"] WEATHER["weather content pass"] --> S24["S24 rain bonus"] style S11 stroke:#3e8fa3,stroke-width:2.5px style S12 stroke:#3e8fa3,stroke-width:2px style S21 stroke:#a33d2e,stroke-width:2px
Combat in this game is a deterministic simulation run on every player's machine; the server relays which spell fired and a random seed, and holds no spell logic at all. A magic school is therefore mostly authored data — formulas, effect entries, channel rows, skill-sheet cells — and this page is shaped accordingly:
| Change class | Cards | What it touches |
|---|---|---|
| Formula and tuning changesets | S1, S2, S8, S9, S13, S15, S16, S18, S19 | Spell expressions, regenerated through the real tokenizer with pins |
| Effect-channel data | S10 | One new channel row; seven slows reassigned, three placement spells cleared |
| Shared enum appends (both code trees) | S11, S14, S20, S21 | Append-only stats and ids under the shared-source mirror discipline |
| New spells on existing machinery | S21, S22 (id already reserved), S8 (already authored) | Skill-sheet rows, spell assets, icons, localization |
| Client fixes and presentation | S3, S4, S5, S7, S25 | Area-effect processing, FX rebuild, strings, state checks |
| Tests | S6 and every card's pins | Editor-mode formula and state-machine pins |
S6 is the foundation: every water formula pinned untrained and at grandmaster with a tokenizer round-trip. On top of it, per-card pins in the project's usual style — S1's damage-type pin, S2's duration parity, S3's reapply integration test, S10's cross-source eviction matrix, S11's stack caps, S12's full state machine (trigger, break, immunity, reset), S13's crit terms at every pip count, S15's cooldown curve across specialization, S16's tick cadence, S23's build-and-decay cycle. The two-client checklist rides every target-authoritative change, as it did for the fire page's mirror cards.
| Ask | Asked by | Answered |
|---|---|---|
| Make water debuffs stack instead of overwriting | Garfunkel Humperdinck, Lord Baldrith, Violet Ronso; reposted Jul 2026 | S10, S11 |
| Fix Ice Field: focus cost, reapply, duration honesty, damage | Barugon (three asks, Aug 2026), Cairo Hayden, Violet Ronso, Chiasma | S3, S16 |
| A real water AoE; a blizzard spell | echoboxtv, EveMidnight, Vallo Frostbane, Lord Baldrith | S21 |
| Bring back a freeze, fairly | Lord Baldrith ("Ice Ball", "Ice Storm"), majoria70 | S12 |
| Cut Cone of Cold's cooldown; a Tempestry chest enchant; no cancel-on-turn | Violet Ronso, Ysold | S1, S15 |
| A water control alternative after the bard rework | Barugon | S14 |
| Slows that work on big targets | Morph, Lord Baldrith | S9 |
| A reason to use Shield of Ice; fix its duration asymmetry | Syee Kco; tracker submission 2021 | S2, S17 |
| Restore the Water Elemental; make the Ice Elemental's slow visible | Silmaris Dragonbane, Rinaldi, Barugon | S18 |
| Fix Soothing Rain's visuals; give it a small fire-resist | a decade of threads; Waxillium, Curt | S4, S19 |
| Douse castable outside combat | Lained | S7 |
| Water magic stronger in the rain | Jack Knyfe; constraints by Turk Key and Alley Oop | S24 |
| More focus tools like Soothing Rain | wishlist thread, 2021 | S8, S23 |
| Frost Weapon, the water enchant | EveMidnight, asked on three streams | Shipped — delivered before this page |
Shipping game data (the fifteen water skill-sheet rows, the spell and sub-effect assets, the effect-channel registry, the summon prefabs and their innate formulas); the combat engine's effect, channel, and resist code; the balance program's audit, targets, and change log; the school reference documentation; roughly forty community threads from 2014 through this week, including the 2021 specialization-popularity data pull, the 2025 school-by-school balance audit, the original 2016 "Water Magic needs help" thread and its 2026 wishlist reposts; and the genre record — Ultima's An Flam lineage and UO's cold typing, EverQuest, Dark Age of Camelot's ice speccing, WoW's Shatter era through its modern redesigns, Guild Wars 1 and 2, Divinity: Original Sin 2, Baldur's Gate 3, Path of Exile 1 and 2, Elden Ring, Magicka and Noita, Final Fantasy XIV's Umbral Ice, Diablo 2's cold armors, Grim Dawn, Last Epoch, and ESO's Warden.
Prepared 2026-08-14 against the unity-6.5 migration branch. Companion markdown: docs/plans/2026-08-14-water-magic-improvements-proposal.md. Twenty-five proposals: nine fix-first, eleven core, five deeper water. Fifth page in the magic-school series, after Fire, Earth, Life, and Death.
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.