Build the party around jobs, not labels

These are three repeatable starting frameworks built from current class weights and chassis data. They are not universal best-in-slot lists. A real party can outperform the cleaner template when its attributes, traits, equipment, and named synergies are better. Use each build to create a recruitment checklist, then replace any member who fails the actual job.

Build 01 · Safe caster

Guardian · Arcanist · Cleric

This is the clearest protection-and-output framework. Guardian brings the strongest concentration of Threat, Armor Use, Parry, and Mitigation weights, plus the highest health modifier. Arcanist brings the highest damage modifier and a 1.9 Spellcraft bias, but also the lowest health modifier. Cleric supplies the highest Healing weight with Composure, Decisions, Concentration, and Anticipation around it. Every slot has one narrow promise.

Recruit the Guardian for an even defensive core rather than one star. Recruit the Arcanist only when Spellcraft and Concentration are convincing and the tank can reliably keep attention away. Recruit the Cleric for Healing plus operational stability, not damage. Dwarf or Goliath can reinforce a Guardian's durable rows; Tiefling or Gnome can support an Arcanist's spell profile; Aasimar or Gnome can reinforce parts of Cleric. Those race ideas are shortlists, not required pieces.

The weakness is brittleness when the defensive plan breaks. Arcanist's ×0.66 health chassis leaves little room for repeated attention, and Cleric's own health modifier is below neutral. Before committing, check dungeon conditions, the tank's Threat, and whether traits make target selection or morale less predictable. If the Guardian roll is mediocre, the whole framework loses the assumption that makes it safe.

Build 02 · Controlled pressure

Warden · Ranger · Druid

This framework spreads control and anticipation across the party. Warden has a durable Vitality–Threat–Mitigation core plus Aggression, Strength, Power, and Crowd Control. Ranger contributes ranged damage through Marksmanship, Anticipation, Critical Strike, Agility, and Positioning. Druid balances Healing and Support at 1.7 each while adding Anticipation, Teamwork, and Crowd Control.

The shared supporting rows are useful only if the real recruits roll them. Start with the Warden's tank core, the Ranger's Marksmanship and Anticipation, and the Druid's Healing and Support. Then use Crowd Control, Teamwork, Positioning, and Composure to choose between finalists. Half-Orc or Goliath can reinforce the Warden's physical side. Elf or Halfling is a clean Ranger shortlist. Human or Aasimar can help parts of the Druid profile, although no race directly supplies Healing or Support tendencies.

This party trades Guardian's extreme defensive concentration and Cleric's peak Healing bias for a broader group. That makes candidate quality especially important. A weak Warden cannot be excused by shared Crowd Control, and a Druid who rolls only Support still leaves recovery exposed. Use this build when the actual heroes show a complete core plus useful overlap, not when you merely like the three class descriptions.

Build 03 · Stable melee team

Paladin · Blademaster · Bard

Paladin supplies a conventional defensive core with extra Resolve, Bravery, Leadership, Discipline, and a modest Healing lean. Blademaster provides physical melee damage through Weapon Skill, Power, Critical Strike, Agility, Speed, and Positioning. Bard has the strongest Support weight and a broad Teamwork, Healing, Leadership, Anticipation, and Decisions profile. The concept is a resilient front supporting a specialised finisher with party-oriented help.

Dwarf and Dragonborn are strong Paladin shortlists because they overlap defensive or resolve rows without directly penalising the core. Half-Orc and Goliath offer Strength and Power for Blademaster, but Goliath's Agility penalty must be checked on the real sheet. Human aligns naturally with Bard through Leadership, Decisions, and Teamwork; Aasimar reinforces Composure and Leadership. Race fit cannot replace Threat on Paladin, Weapon Skill on Blademaster, or Support plus Healing on Bard.

The risk is overvaluing the party's broad social and physical shape while accepting weak fundamentals. Paladin's Healing is not enough to make Bard's recovery optional, and Bard's ×0.7 damage modifier cannot cover a poor Blademaster. Melee positioning and dungeon conditions may also change which damage profile is safest. Inspect each slot independently before rewarding the way the trio looks together.

Before you lock a build

Run the six-point party audit

Role: one hero can reliably hold pressure, one can finish, and one can sustain.
Core rolls: each recruit passes the signature checks on the individual class page.
Tradeoffs: negative race tendencies have been priced against important class weights.
Behaviour: traits do not undermine target choice, morale, condition, or guild economics.
Environment: the party has an answer to the destination's possible dungeon conditions.
Bonus: any named synergy is confirmed in the current table, not inferred from theme.

Adaptation rule

Replace the weakest job, not the least fashionable class

When a build fails, identify the first broken promise. If enemies reach the damage slot, compare Threat, mitigation, and control before replacing the healer. If delves stall safely, inspect the damage core before making the tank more aggressive. If a strong party collapses over time, check Healing, Support, conditions, and traits before changing the entire class lineup. This diagnosis preserves good heroes and makes recruitment specific.

Use the race and class guide to evaluate replacements, the recruitment guide to price and time the signing, and the synergy table to verify named bonuses. Finally, confirm the page version and changelog; a sound party from an older release may be answering weights that no longer exist.