How the tier list is calculated

We put every character through the same simulated boss fight, in the teams their kit was built for, with the same gear for everyone. Then we time how fast each team burns the boss down. A Main DPS (the character on the field doing most of the damage) is scored on how fast its best teams kill. A Sub-DPS or Support (the teammates) is scored on how much faster they make the carries that are built to use them. The bands (Crowned, Exalted, Excellent, Strong) are then cut by a fixed rule, never by hand. Two switches on the list show the same simulations another way: Team or Solo, and the ranked character at S0, S3 or S6. Team at S0 is the main list.

Nothing on the list comes from anywhere else. If the simulator cannot measure something that matters, we extend the simulator; we never import someone’s opinion.

The simulator is the same engine behind every character page on this site. It plays a rotation (the exact list of skills a player presses, in order) hit by hit. Every hit lands at its own moment, reads the buffs and debuffs active at that exact moment, and uses the damage formula and skill multipliers from the game’s own data. Nothing is random: the same rotation always gives the same result, and crits are counted at their average value.

One build for everyone

If one character had luckier Echoes than another, the list would partly rank luck. So every character wears the same benchmark: 5 Echoes with 5 substats each (25 lines), every line rolled at its average value (the probability-weighted average of the game’s roll tiers, snapped to a real tier), 15 lines on stats the character wants and 10 on junk. What stays personal is what makes the kit itself: the Echo set, the main Echo and its skill, and the main stats a kit requires.

Two bars decide what a team is allowed to press, and the simulator enforces both on every character, the same way the game does.

• Resonance Energy is the bar that pays for a character’s Liberation (their Ultimate). Hits fill it by the amount the game data gives each hit, multiplied by the character’s Energy Regen (ER). The Liberation spends the whole bar. Overflow above a full bar is lost, and damage from debuffs (like damage over time) and Tune Break gives no Energy.
• While someone is on the field, each teammate waiting off the field gets 50% of each hit’s Energy (times their own ER). Flat “restore X Energy” effects are never multiplied by ER.
• Concerto is the swap bar, 0 to 100. A character’s own hits fill it (ER never helps). A full bar is needed for the Outro (the skill a character fires as they swap out), which spends it. Only after a real Outro can the next character open with their Intro (the skill they fire as they swap in). Unison partners are the one exception: inside a Unison window the Outro is free.
• A Liberation without a full Energy bar, or an Outro without a full Concerto bar, is blocked: it does not happen, takes no time and deals no damage.

Rotations must pay for themselves

A real fight is long, so a rotation that only works once (because it starts with full bars) is not a real rotation. We call the repeating rotation the loop, and a loop that keeps repeating cleanly is in steady state. To check it, the simulator plays the loop 3 times back to back on one timeline, carrying everything over (bars, cooldowns, buffs, stacks, gauges, forms), and searches for the bar levels the loop leaves exactly as it found them. If any cast in the second or third repetition is blocked, the team is illegal: it is not ranked, and no number on the list is taken from it.

Energy Regen is fitted per team

Players build more ER when their team needs it, so the simulator does too. For every team, each member starts with no extra ER (healers whose own kit scales with ER keep what their kit needs) and gains whole ER substat lines of 9.2% until every Liberation in the loop is paid, then sheds any line the team can do without. Each line replaces one of the build’s useful damage lines, so ER is never free. The limit is 5 lines (one per Echo). A team still short at the limit, or short on Concerto (which ER cannot fix), is illegal.

Wuthering Waves is a team game, so a carry is only ranked in teams that actually work with its kit. Every partner is chosen from the kit text itself, for one of three reasons:

• Designed: both kits name a shared mechanic (Unison, Tune Strain or Tune Rupture, Glacio Chafe, Havoc Bane, Fusion Burst, Aero Erosion, Echo Skills, shield feeds, Hack).
• Resonant: the partner’s Outro or team buff amplifies the carry’s element, or a damage type that makes up at least a quarter of the carry’s own damage (measured in the simulator). A generic flat damage buff alone does not qualify.
• Healer: a character who heals or shields the team, whose team buff lands on anyone.

A real team is the carry plus two different partners, with at least one healer and at least one designed or resonant partner, so every team can survive and none is just two generic healers. Every real team is played in both slot orders (who acts first changes whose Outro buffs whom), and the faster order counts, the way a player would pick it. A character with two Resonance Modes is ranked on its better mode, each mode in its own teams.

Teams played the way they are played

Most teams use the same simple shape the character pages use: each support takes their turn, then the carry takes theirs. Some kits hand off inside each other’s turns, which that shape cannot express (for example a Unison pair trading windows, or Echo Skills summoning an off-field partner’s attacks). Those teams get an authored team cycle that weaves the turns together, and the pages, the rotation editor and the tier list all play it.

Every time on the list, solo or team, is a fresh boss fight: the clock starts at the first second of a single-target boss fight. Energy and Concerto bars start full, every kit starts with its normal fight-start resources, and the team plays its opener (the one-time setup at the start of a fight) and then its loop back to back for as long as the fight lasts. Every hit counts at the moment it lands. Setup turns and ramp-up are part of the time, because a real fight pays for them too: a kit whose damage builds over the fight is slower here than one that is strong from the first second.

The clock is game time, on measured animation frames: a Liberation cutscene, a Tune Break or another time stop freezes the fight (buffs, cooldowns and statuses pause with it), and every character swap costs a quarter second. That includes time stops a kit triggers on purpose: Augusta’s Sworn Allegiance plays out entirely inside one, so her whole burst costs no fight time, which is worth about 0.13 on her Main DPS score. A kit that packs its damage into time stops gains the most from this rule.

Two starts, the faster clean one counts

When a rotation has an opener, the fight is played from two starts, the way a player picks their opener: as written (the opener first, then the loop) and loop first (no opener, the teammates’ turns from the first second). The faster start counts, as long as nothing in it is blocked.

The average bar

A fight is timed at two boss sizes: 1.5M damage rewards a team that pays out early, 3M reads the sustained fight. To weigh burst and sustain equally, every team time on the list is the average bar:

Example, live: Aemeath’s best team reaches 1.5M in 20.5s and 3M in 45.0s, so its average bar is (20.5 + 45.0 / 2) / 2 = 21.5s (stored as 21.5s). Solo views are timed to one real endgame boss instead (below).

In the Team views a carry’s score adds up three terms: team speed 60%, team depth 25% and ease of play 15%. Every time is a fresh boss fight; team times are average bars. This section walks through the main list (Team, S0); the S3 and S6 views use the same three terms with the carry at that Resonance Chain.

How each term is normalised

Times are turned into a 0 to 1 score against the best on the board, so the fastest carry on a term gets 1 and a carry twice as slow gets 0.5. Ease is already 0 to 1.

Ease of play

A mechanical index read from the repeating solo rotation at the view’s Resonance Chain (S0 on the main list). Four signals, each divided by the highest value of any character we simulate at that chain: casts per 10 seconds, distinct skills used, stance changes per 10 seconds, and resources that punish a mistake (ones that empty when you swap out, press the wrong skill or change stance). Their average is the rotation’s complexity, and ease = 1 − complexity. At 15% it is a tie-breaker, never worth a whole tier.

Worked example

Aemeath is first today. Being fastest on a term earns its full weight, and any slower time earns the fraction best / own:

Jingran is second. Every time term is now a fraction of the best:

Every Main DPS, every term

A teammate is measured only on the carries whose real teams include them, so a strong generic buffer cannot score on a kit it does not work with. Five measurements:

Each measurement becomes a 0 to 1 score as a fraction of the highest value any of the 18 teammate-capable characters posts (today: uplift +90.3%, fit 100.0%, presence 13, reach 13, own damage 19.2k per second). Uplift is measured per second on the field because a team turn cannot simply be repeated alone, so comparing damage totals would measure field time, not buff value.

Which column a character appears in is decided by what their kit does, before any score: the Support column is only for characters who heal or shield the team, and a damage dealer who also buffs goes to Sub-DPS. A flexible character appears in every column they fit, ranked for that role. Carries also get a teammate score, but are only placed as Main DPS unless their kit makes them a real teammate too.

Two switches sit above the list. Team or Solo picks what is measured: a character inside the real teams built around their kit, or a carry fighting alone. S0, S3 or S6 picks the Resonance Chain of the character being ranked, while everyone around them stays at S0. That makes six views, all from the same simulations and all cut by the same band rule. Team at S0 is the main list, and the only one written by hand, with a note on every card. The other five are computed straight from the numbers, and each card shows that view’s own numbers when you hover it.

The Solo views: one real endgame boss

A Solo view times the carry alone, from the first second of a fresh fight, until it has dealt the HP of the Tower of Adversity’s Hazard Tower floor 1 boss: 5,510,473 HP in Season 40 (14 September 2026 to 12 October 2026). That boss is Lv100 with 20% RES and 1592 DEF, the same as the boss every other time on this list uses, so only the HP changes. A dual-mode carry is timed on its faster mode alone. The score is 90% that time (fastest / own) and 10% ease of play.

Each card carries the crests that time would earn on the floor’s 300-second timer: 3 crests within 120 seconds, 2 within 150, 1 within 210.

The Team views at S3 and S6

The band tables for every view are in the next section.

Bands are cut by a fixed rule in every column of every view, never by hand. Sort the column by score, then:

• T0 Crowned: the top third (rounded up).
• T0.5 Exalted: the middle third.
• T1 Excellent: the bottom third.
• T2 Strong: anyone below the largest score gap inside the bottom third, if that gap is at least 0.01.

Ties are never split. Two engine passes can move a score by about 0.01, so two characters within 0.01 of each other are a tie. A band line that would fall inside a tie moves up to the nearest real gap, so tied characters share the lower band and a band can come out smaller than its third. A tie never empties a band, though: if moving up would leave the band above the line with nobody in it (the Crowned band, or Exalted when the line would climb all the way to the crown’s line), the line moves down to the nearest real gap instead, and the tied characters share the upper band. Inside a band, characters are listed by score.

The other five views are cut by the same rule over their own scores:

A simulator is only as honest as its list of blind spots. Here is ours, in plain words.

• One target only. Damage to groups of enemies (AoE) earns no credit; every character is judged on a single boss.
• The boss never fights back. No movement, no invulnerable phases, no time lost to dodging. Skills that need a perfect dodge (Dodge Counters) are allowed at most once per 10 seconds a character spends on the field, and only where the kit rewards them.
• Healing and shields are not measured by size. The Support column checks that a character heals or shields the team; how much they heal is not scored.
• Every swap costs the same quarter second. The simulator charges 0.25s of fight time per character swap, but how hard a team is to quickswap is still invisible to the ease index.
• Energy only comes from the fight. The 50% off-field share is inferred, and no Energy comes from defeating extra enemies or anything outside the rotation.
• Everyone wears the same gear. Your account’s Echoes, weapons, refinements and Resonance Chains will differ. Only the S3 and S6 views look past S0.
• The Tower badge is a pace. The Solo views use one real boss’s HP and its timer, but the simulated boss stands still, never turns invulnerable and never makes anyone dodge, floor effects are left out, and only floor 1 (one boss) is timed.
• Field-time uplift has quirks. A carry’s lingering damage (damage over time, summons) that lands during a teammate’s turn still counts for the carry, and a carry whose team turn is far stronger than its solo loop shows a big lift beside every partner.
• Ease is relative. It is measured against the hardest rotation on the roster, so a new, harder kit can shift everyone’s ease a little.
• Beta characters are ranked by our simulations of their beta kits and can move when the release version lands.

1. 01

   One build, one boss, our engine

   Every character runs on the same benchmark Echo build (25 substat lines at average roll values, 15 of them on stats the kit wants) with their signature weapon, against the same Lv100 boss with 20% RES. Nothing on this page comes from anywhere but that simulation.

2. 02

   Two switches, six views

   Team or Solo picks what is measured: a character inside the real teams built around their kit, or a carry fighting alone. S0, S3 or S6 picks the Resonance Chain of the character being ranked, while everyone around them stays at S0: in a Team view that is the carry for the Main DPS column and the teammate for the Sub-DPS and Support columns. Team at S0 is the main list. Solo views rank Main DPS only, because a Sub-DPS or Support is measured by what they add to a team. Every view is scored and banded by the same rules from the same simulations.

3. 03

   Real teams only

   Wuthering Waves is a team game first, so every carry is ranked only in teams that actually work with its kit. A partner qualifies when both kits name a shared mechanic (Unison, Tune Strain or Tune Rupture, Glacio Chafe, Havoc Bane, Fusion Burst, Aero Erosion, Echo Skills, shield feeds, Hack), or when their buffs amplify the carry’s element or a damage type that makes up at least a quarter of the carry’s damage. Every team has a healer, and a generic flat buff alone does not earn a slot. A dual-mode kit is ranked on its better mode, each mode in its own teams.

4. 04

   Played the way the team is played

   Teams whose kits hand off inside each other’s turns (Hsin and Suoming’s Unison windows, Qiuyuan’s Echo Skills summoning Hecate during Phrolova’s Maestro, Aemeath’s burst timed inside her partner’s Outro window) run an authored rotation that weaves them together; every other team runs each support’s turn and then the carry’s, the same three-slot model the character pages use. Both slot orders are played and the better one counts.

5. 05

   Time to kill, from the start of a boss fight

   Team speed is the average of the seconds to burn 1.5M HP and half the seconds to burn 3M, so the opening burst and the sustained fight count equally. Every time, solo or team, is timed from the start of a single-target boss fight: Resonance Energy and Concerto bars start full, every kit starts with its fight-start resources, and the team plays its opener and then its rotation back to back for as long as the fight lasts, every hit counted where it lands. Setup turns and ramp-up are part of the time, because a real fight pays for them too. The clock is game time: a Liberation cutscene, a Tune Break or another time stop freezes the fight (and every buff and cooldown with it), and every character swap costs a quarter second. Each fight is played from two starts and the faster clean one counts, the way a player picks an opener: as the rotation is written (its opener first), and straight into the rotation with the teammates’ turns first. Energy honesty is kept separately: repeated back to back, the rotation must pay for itself every time (each Liberation costs its full Resonance Energy, each Outro needs a full Concerto bar), with Energy Regen fitted per team (the fewest ER substats that pay every Liberation). A team that cannot pay is not ranked.

6. 06

   Main DPS score (Team views)

   60% the fastest real team, 25% depth (the middle of its three fastest teams with different partners, so one perfect pairing is not the whole story), 15% ease of play. At S3 or S6 the carry plays the same real teams at that Resonance Chain, teammates at S0, and ease reads its S3 or S6 rotation. Solo speed and the S6 team used to be part of the main score; since 5 October 2026 they are views of their own, so the main list measures one thing: the carry in its real teams at S0. Higher is better; bands never contradict the score.

7. 07

   Solo score (Solo views)

   90% solo speed: the seconds the carry alone needs to deal the HP of the Tower of Adversity’s Hazard Tower floor 1 boss (5,510,473 in Season 40), from the first second of the fight, on its faster Resonance Mode. 10% ease of play at that Resonance Chain. Main DPS only. Each card shows the crest pace that time reaches on the floor’s 300-second timer (3 crests within 120 seconds, 2 within 150, 1 within 210): a pace, not a clear, because our boss never moves, never turns invulnerable and never makes anyone dodge.

8. 08

   How the bands are cut

   By a fixed rule, never by hand, in every view. The top third of each column (rounded up, so 5 of 13 Main DPS) is Crowned, the middle third Exalted and the bottom third Excellent, with anyone below the bottom third’s largest score gap marked Strong. A boundary never separates two scores within 0.01 of each other, because a margin that small moves between engine passes: it moves up to the nearest real gap, so tied characters share the lower band, unless that would leave the band above it empty, in which case it moves down to the nearest real gap instead.

9. 09

   Sub-DPS and Support scores

   A teammate is measured only on the carries whose real teams include them. Uplift is the rise in the carry’s own damage per second on the field beside them, against the carry playing alone, over the same boss fight up to 3M damage, averaged over those carries and capped at +100% each. Fit is how close each carry’s best team with them comes to its best team overall. Presence counts the carries whose top teams (within 10% of their best) include them; reach counts the carries with a working real team that includes them. Sub-DPS: 25% uplift, 25% fit, 15% presence, 15% reach, 20% their own damage per second. Support: 30% uplift, 30% fit, 20% presence, 20% reach. At S3 or S6 the teammate is at that Resonance Chain in every one of those teams, the carries at S0. The Support column is for units that heal or shield the team.

10. 10

    Ease of play

    A mechanical index from the executed rotation: casts and stance changes per ten seconds, distinct skills, and resources that punish a swap or a wrong cast. A tie-breaker worth 15% of a Team view’s Main DPS score and 10% of a Solo score, never a whole tier.