The answer
Skyborne
| Place | Race | Scenarios won | Range | Racial worth |
|---|---|---|---|---|
| #1 | Skyborne | 55% | 54–56% | +2.6% |
| #2 |  Human | 31% | 31–32% | +2.3% |
| #3 |  Dwarf | 10% | 10% | +0.4% |
| #4 |  Night elf | 4% | 3–4% | +1.5% |
| #5 |  Gnome | 0% | 0% | +0.8% |
Scenarios won: in how many of the fights each race did the most damage. Range: where that share lands if each unconfirmed game rule goes one way or the other. Racial worth: damage gained over a warrior with no racial.
Every set of shares on this page is rounded to whole percents that add up to 100.
- Why Skyborne: Wind Blessed, 3% faster attacks, always on: more swings, more rage.
- If the boss is a Beast, Dwarf wins 99% (Big Game Hunter); on an Elemental, Skyborne (Elemental Insight). On any other boss: Skyborne 59%, Human 37%. We check every raid boss at launch.
- How much it matters: about 2.6% more damage. Pick what you love.
What a scenario is
A scenario is one raid boss fight, simulated. We built 10,000 of them, and each one draws at random:
- how long the fight lasts: 3 minutes in the middle, from 45 seconds to 10 minutes;
- the raid: 10, 20 or 40 players, and the buffs that come with it;
- consumables;
- your gear, from fresh level 60 to raid-geared;
- how much of the fight you're in melee range;
- the boss: how hard and how fast it swings, how much damage the raid takes, how long the execute phase lasts, and its creature type;
- every game value nobody has confirmed yet (listed in the full method).
All five races play every scenario, with the same draws. The race that does the most damage wins it. Within a scenario the races differ only by their racials and their base stats.
The fight: one boss, one target, level 60, talents 36/15/0 (Arms / Fury / Protection). A two-hander in Battle Stance after Recklessness: Execute, Mortal Strike, Rend, Overpower, Slam. Each race tries an axe, a mace and a sword in every scenario and keeps its best: a sword in 73–79% of them.
Built from the game, not from guides. Abilities, talents and racials are read from the game's own tables for build 1.60.1.70450. Random seed 20261104, so every number here can be rerun exactly. Sampling error: ±1.0 points.
Why Skyborne wins
Each race's racials as build 1.60.1.70450 has them, and what they're worth to a Arms Warrior: the damage gained over a warrior with no racial, in the middle scenario, and the spread across 8 in 10 scenarios.
| Race | Racials | Worth |
|---|---|---|
| Skyborne | Wind Blessed: 3% faster attacks, always on. Elemental Insight: +5% damage against Elementals. | +2.6%1.5–4.1 |
| Human | Ruthless: +2% crit chance, always on, with any weapon. | +2.3%1.9–2.7 |
| Dwarf | Earth Fury: +2% critical damage. Big Game Hunter: +5% damage against Beasts. Stoneform: 10% less damage taken for 8 seconds, every 3 minutes. | +0.4%−0.2–4.4 |
| Night elf | Quickness: +1% dodge, always on. Elune's Light: +10% crit for 15 seconds, every 3 minutes. | +1.5%1.0–2.2 |
| Gnome | Eureka!: for 8 seconds every 2 minutes, abilities cost 10% less rage and hit 10% harder. Expansive Mind: +5% maximum rage. | +0.8%0.3–1.4 |
A two-hander's rage comes from its swings, and Wind Blessed's 3% speeds up every one; Mortal Strike, Slam and Overpower turn that rage into damage. It's by a hair: Skyborne's racials are worth 2.6%, Human's 2.3%.
It rests on Wind Blessed's 3%. With the racials of build 70245, an older stress-test client where Wind Blessed was 1%, Human wins Arms 66%. The stress-test client of 21 October, 70532, has the same spell data as 1.60.1.70450, 3% included.
The shortest fights go another way: at 30 s, Night elf wins 60% (Elune's Light, +10% crit for 15 seconds). From about 1 min on, Skyborne leads.
The numbers as a table
| Fight length | Skyborne | Night elf | Human | Dwarf | Gnome |
|---|---|---|---|---|---|
| 30 s | 30% | 60% | 0% | 9% | 1% |
| 45 s | 33% | 56% | 0% | 10% | 1% |
| 1 min | 46% | 42% | 0% | 11% | 1% |
| 1 min 30 s | 57% | 18% | 13% | 12% | 0% |
| 2 min | 59% | 2% | 27% | 11% | 1% |
| 3 min | 56% | 0% | 32% | 12% | 0% |
| 5 min | 50% | 0% | 39% | 11% | 0% |
| 7 min | 48% | 0% | 41% | 11% | 0% |
| 10 min | 50% | 0% | 39% | 11% | 0% |
The deciding question: what creature type is the boss?
Two racials only count against one kind of boss: Dwarf's Big Game Hunter, +5% damage against Beasts, and Skyborne's Elemental Insight, +5% against Elementals. That's more than any other racial is worth, so on those bosses they win almost every fight. The game files we read don't give the raid bosses' creature types, so the model makes 1,009 of the 10,000 bosses Beasts and 494 Elementals: about 1 in 10, and 1 in 20.
On a Beast (1,009 fights)
- 99% Dwarf
- 1% Skyborne
On an Elemental (494)
- 100% Skyborne
On any other boss (8,497)
- 59% Skyborne
- 37% Human
- 4% Night elf
So the honest answer is per boss: Dwarf on a Beast, and Skyborne on every other boss, which is most of them.
No unconfirmed game rule changes the winner. The biggest, how WoW Forever gives rage, we forced each way over all 10,000 fights: Skyborne wins 55.4% with rage per swing and 54.5% with Classic's rage.
How we test it
At launch we read every raid boss's creature type in its tooltip. The first raid tier (9 December) has 22 named bosses, and only Onyxia's is known so far: Dragonkin. We also check Wind Blessed's value on the launch build: at 1%, Human leads Arms. This page gets the answers, dated, in its changes.
How sure we are
The stress test moved every assumption we could, one at a time: fight lengths, gear, raid size, consumables, other random seeds, the rage rule at any odds, and the racials of the older stress-test build 70245. Skyborne stays first in 63 of 68 cases. Besides a Beast or an Elemental boss, it loses only in these cases: with build 70245's racials, from an older stress-test client (Human 66%); Wind Blessed at 1% alone (Human 70%).
How much it matters
Skyborne's racials are worth about 2.6% more damage than none. The others' are worth 2.3% (Human), 0.4% (Dwarf), 1.5% (Night elf), 0.8% (Gnome).
In a single pull, it's close. With the same raid and the same gear, a Skyborne out-damages a Human in 51% of pulls (41–63% across the scenarios). Every roll counts there, the boss's included, so one night's luck is bigger than the racial.
Pick what you love. The gap is real, but small.
The human factor
Simulators press every button the moment it's ready. People don't. Skyborne's Wind Blessed is always on; Elune's Light and Eureka! have to be pressed. Here are the shares when they all get pressed less:
| How often they get pressed | Skyborne | Human | Dwarf | Night elf | Gnome |
|---|---|---|---|---|---|
| Every time | 55.0% | 31.4% | 10.0% | 3.5% | 0.1% |
| 3 times in 4 | 55.5% | 32.9% | 10.1% | 1.6% | 0.0% |
| Half the time | 55.5% | 34.0% | 10.1% | 0.4% | 0.0% |
| 1 time in 4 | 55.5% | 34.4% | 10.1% | 0.0% | 0.0% |
| Never | 55.5% | 34.4% | 10.1% | 0.0% | 0.0% |
Elune's Light is the button that counts: pressed every time, Night elf wins 3.5% of the scenarios; never pressed, 0.0%. If nobody presses theirs: Skyborne 55.5%, Human 34.4%, Dwarf 10.1%. In the middle scenario Night elf wins, it has to press Elune's Light 70% of the time to stay ahead. After launch we count how often our own raiders press theirs.
Why other guides disagree
The open simulator's Arms table, gunba/wow-forever-sim (a fork of wowsims/classic), ranks Human, Night elf, Skyborne, Dwarf, Gnome. So we gave our engine the simulator's own character sheets and its fight (5 minutes, raid-geared, always in range) and ran it two ways: once with Skyborne's 1% faster attacks, the racial before build 70425, and once with the game's 3%. Its gear and world buffs aren't ours, so this checks the gaps between races, not the damage.
| Against Human | Simulator | Ours, Skyborne at 1% | Ours, at the game's 3% |
|---|---|---|---|
| Human | 848.1 DPS | base | base |
| Skyborne | −0.50% | −1.22% | +0.12% |
| Night elf | −0.38% | −0.82% | −0.82% |
| Dwarf | −1.17% | −0.85% | −0.85% |
| Gnome | −1.17% | −1.41% | −1.41% |
| Order | Human, Night elf, Skyborne, Dwarf, Gnome | Human, Night elf, Dwarf, Skyborne, Gnome | Skyborne, Human, Night elf, Dwarf, Gnome |
The simulator puts Human first by 0.50%; with the game's 3% we put Skyborne ahead by 0.12%. Both say it's close between them. Gnome and Dwarf trail further in ours: the simulator searches gear for each race separately and picks different weapons, while in our scenarios every race wears the same gear, so only the racials and base stats differ.
We read the simulator's published table and didn't run it. It's open source under the MIT license.
The full method
Every unconfirmed game value, and the odds we gave it
Each scenario draws every one of these. The ranges on this page are how far a share moves when each of them goes one way or the other. None moves a winner; the boss's creature type does, and it has its own section above.
| Value | In the scenarios | Why it's unknown |
|---|---|---|
| WoW Forever gives rage per swing | yes in 65%, no in 35% | each landed swing gives rage by the weapon's speed, whatever it hits for; the other way is Classic's, rage from the damage dealt. The open simulator plays the first but marks it provisional, and the game files don't hold rage formulas |
| The off hand gives half the rage | yes in 70%, no in 30% | the open simulator's rule; untested |
| A glancing blow loses | 25 to 35% of the hit | Classic's average is 35%; the WoW Forever beta measured about 25% |
| Earth Fury's +2% critical damage scales the whole crit | yes in 60%, no in 40% | the same effect did on Burning Crusade’s meta gems; the other reading adds the 2% to the crit bonus only |
| Earth Fury needs a mace | yes in 40%, no in 60% | the game files keep the mace row of Mace Specialization, the Dwarf racial it replaced; the tooltip names no weapon. Only Arms changes: its Dwarf then needs a mace |
| The raid boss is a Beast, or an Elemental | a Beast in 1,009 fights, an Elemental in 494 | Dwarf's Big Game Hunter and Skyborne's Elemental Insight add 5% damage against them; the raid bosses' creature types aren't in the game files we read |
How the model reads the game
The model reads the game's own tables (DB2) for build 1.60.1.70450, exported from wago.tools: every level-60 warrior ability at its highest rank (damage, rage cost, cooldown), WoW Forever's talent tree with each talent's value from its own curve, the racials, and the level-60 creature row the rage from damage taken depends on. The spells' data decides what modifies what: Eureka! touches every rage ability, Heroic Strike included, but not Rend's bleed or Sunder Armor. Nothing is a number copied from a guide.
Every roll is real and drives what happens next: rage, Flurry, Overpower, Enrage, Shield Block. But each attack is scored by its expected damage, not the rolled number, which keeps the score fair and takes out most of the luck. Every scenario is the average of 48 pulls, with the same dice for every race. The head-to-head pulls score every roll as rolled.
10,000 scenarios with the random seed 20261104, so every number here can be rerun exactly. The fight-length chart reruns 600 scenarios at each of 9 fixed lengths.
What we typed in by hand
- The talent build (51 points): Arms 36 / Fury 15, 380 DPS in the middle scenario, ahead of Arms 34 / Fury 17 (366). We compared the two over the same 300 scenarios with no racial.
- The rotation, a raider's usual priority. Fury: Recklessness at the pull, Death Wish on cooldown, Execute, Bloodthirst, Whirlwind, Overpower when it lights up, Heroic Strike with rage to spare. Arms: Execute, Mortal Strike, Rend, Overpower, Slam. Protection: Shield Block whenever it's ready, Thunder Clap kept up, Shield Slam, Revenge, Sunder Armor. Elune's Light waits until Recklessness ends; Eureka! and the tank's Stoneform go on cooldown.
- The warrior's level-60 base stats and the races' stats against them: the game keeps them outside the tables wago.tools exports, so the values are wowsims'.
- The attack table against a level-63 boss (misses, dodges, parries, glancing blows) and the boss's armor and crushing blows: Classic's rules.
What isn't in the model
- Set bonuses, trinkets and procs from gear: they help every race alike.
- World buffs, fights with more than one target, Sweeping Strikes, and movement beyond how much of the fight you're in melee range.
- Threat. The tank is scored by the damage it takes, not by its damage or its threat.
- Healing, magic damage, and the tank's own survival cooldowns beyond Stoneform.
- PvP. The PvP racials don't count here.
The data. The rankings, ranges, racials' worth and the fight-length sweep on this page as files, free to reuse with credit (CC BY 4.0): CSV and JSON.
Our prediction, before launch
Simulated on 20 October 2026, on build 1.60.1.70450, before WoW Forever launches on 4 November. After 9 December we grade it against our own raid logs, here on this page.
Changes
- Simulated and written, on build 1.60.1.70450.