Race tendencies and class weights answer a narrow but useful question: which generated attributes are being nudged in the same direction? They do not tell you which hero will be best, which party will clear a dungeon, or whether a contract is worth its wage. This guide turns the two tables into a repeatable shortlist and shows where that method must stop.
Checked against v3.0.0
Begin with the vacancy. A party normally needs someone to hold pressure, someone to finish encounters, and someone to sustain the group. The nine classes are already divided into tank, damage, and healer roles. If the roster needs a tank, comparing an excellent Ranger pairing with a merely good Guardian pairing solves the wrong problem.
Within a role, decide what kind of job the party needs. Guardian concentrates on Threat, Armor Use, Parry, and Mitigation. Warden, rebuilt in 2.12 as a heavier tank, puts Vitality, Mitigation and Armor Use at the top and backs them with Strength, Bravery, Parry and Crowd Control; its Aggression weight fell to 1.0. Paladin keeps a defensive core but adds Resolve, Leadership, Bravery, and some Healing. Those are three different answers to “we need a tank.” Make the same distinction between melee, ranged, and spell damage, or between direct healing and broader support.
A high class weight means generation strongly favours that attribute. It does not mean a 1.8 weight produces 80% more combat output, and it does not guarantee every recruit will be good. Use weights to identify signature attributes and compare the relative importance of a race tendency.
Positive and negative race values point attributes up or down. Their impact depends on the class. Gnome's −3 Strength is expensive for Blademaster but much less important for Cleric, where Strength has only a 0.5 weight. Counting plus and minus signs without that context is the fastest way to produce a bad race tier list.
The race pages calculate a simple score: for every attribute in a race profile, multiply its tendency by the matching class weight, then add the results. Positive overlap raises the score and a penalty on a weighted class attribute lowers it. Attributes absent from a class's extracted weight table contribute zero. The three highest positive results appear as best-fit classes.
This model is useful because it is explainable. Elf scores strongly for Ranger through Marksmanship, Agility, and Reflexes, while its Strength penalty has no listed Ranger weight. Dwarf now scores highest for Warden through Vitality, Armor Use, Resolve, and Parry, with Paladin close behind. Half-Orc scores strongly for Warden through Strength, Power, Aggression, and Vitality while its Spellcraft penalty lands on a class weight of only 0.3.
The score is not a combat simulator. It does not know the final hero roll, wage, traits, age, party synergy, dungeon, or the precise internal effect of every attribute. It also cannot give value to an attribute omitted from the class weight table, even though since 2.12 every one of the thirty attributes does something in a fight and the Ability rating reads all of them. Use it to decide which candidates deserve inspection, never to auto-sign the highest card.
Dwarf contributes Vitality +3, Armor Use +2, Resolve +2, and Parry +1, with Speed −2 and Agility −1. Since the 2.12 rebuild Warden earns the highest fit: it values Vitality and Armor Use at 1.7 and now counts Resolve and Parry too. Paladin follows through Vitality, Armor Use and Resolve, while Guardian gains Vitality, Armor Use, and Parry. The correct choice still comes from the vacancy: Warden for a heavy physical front that holds a crowd; Paladin for a resolve-heavy, leadership-capable tank; Guardian for the most concentrated conventional defence. Then inspect Threat and Mitigation on the real candidate, because Dwarf does not automatically supply either one.
Class health multipliers moved for the three classes that trailed: Arcanist from 0.66 to 0.8, Blademaster from 1.0 to 1.3, and Warden from 1.6 to 1.7. The Arcanist's Fireball also got cheaper and faster, Warden's Whirling Cleave costs half the rage, and Druid's Regrowth heals 1.8× instead of 1.5×. A fragile damage dealer is less of a liability than older rankings suggest, so check the version on any tier list before trusting it.
Check the class's top one or two weights. A poor signature roll is not repaired by a thematic race match.
Look for balance across the attributes that make the role reliable: tank control and mitigation, damage execution, or healer operation.
Identify every negative race tendency that hits a meaningful class weight. Do not hide it behind a larger positive number.
Compare growth modifier, current age, starting quality, and the seasons in which the roster needs the hero.
Finish with traits, wage, Hall capacity, and party synergy. A fit score cannot tell you whether the complete contract is good.
Human's ×1.08 modifier is the fastest current value. Elf, Dwarf, and Gnome are at ×0.95; the other races are neutral at ×1.0. A faster modifier is attractive for a development slot, but an established older recruit may solve today's party problem better. A slower race with an excellent starting sheet can also need less development in absolute terms.
Race age ranges are game parameters, not one shared real-world lifespan scale. Compare heroes within the game's roster horizon. Ask when value arrives and how many seasons you expect to keep paying for it.
A top-scoring damage pairing is not useful when the guild lacks a tank. Two excellent specialists can also produce redundant strengths or leave control, recovery, or Threat uncovered. Build the three-role skeleton, then check named bonuses in thesynergy table.
Traits can change morale, condition, target selection, training value, and economics. Read the relevant trait analysis and price the recurring wage. The best recruitment decision is the hero who improves the whole roster, not the cleanest multiplication exercise.
A +3 on an irrelevant row can matter less than +1 on a class signature. Weight the overlap.
Halfling helps Blademaster Agility but penalises its high-weight Power and Strength. Read both sides.
A race tendency describes a direction. The candidate in front of you may land far from that average.
Use the page's verified version and the changelog before applying an old ranking.
