“Your opening pick” compares Tier I records in the starting pool with one weapon equipped. “Across the whole run” compares every usable family at its highest usable tier.
Loadout identity
Starting choices and grants
Choose one unlocked entry from a normalized maximum pool of 23. A save’s visible pool can be smaller because unlock state and active DLC alter it.
Only direct character deltas are shown here; selected starting entries and weapon-set bonuses are not presented as base stats.
Your opening pick · pool-limited view
Tier I starting-pool DPS
This answers the first weapon-choice comparison. The across-run ranking below answers a separate question across every usable weapon family.
The normalized maximum starting-weapon pool contains 23 families. A save can display fewer because only unlocked entries appear, and active DLC can alter the pool.
Reading this list: all eligible pool entries are compared together because the opening selection is choose-one. Each value is expected direct primary-target DPS at Tier I with one weapon equipped.
How the opening-pool ranking is computed
Tier I, this character’s starting-weapon pool, and 1 equipped weapon. Default 10 Max HP and zero other default combat stats; no bonus stats beyond Apprentice’s fixed direct base deltas. Stat-gain factors, matching character weapon-set or weapon-type bonuses, and equipment-linked effects are applied through the shared damage calculation. The equipment context is 1 equipped weapon, 1 unique weapon family, 1 Tier I weapon, 0 Tier IV weapons, and 0 duplicate items; item, material, structure, and pet counters are 0.
The target is not cursed, burning, or a boss. Every attack or projectile aimed at the primary target connects, and DPS is before enemy armor. Explosions, burning, spawned attacks, and later pierce or bounce targets are outside this metric.
DeclaredEach level-up: Melee Damage +2 Bespoke trigger, stacking, or unit behavior is not used in the ranking.
DeclaredEach level-up: Ranged Damage +1 Bespoke trigger, stacking, or unit behavior is not used in the ranking.
DeclaredEach level-up: Elemental Damage +1 Bespoke trigger, stacking, or unit behavior is not used in the ranking.
DeclaredEach level-up: Engineering +1 Bespoke trigger, stacking, or unit behavior is not used in the ranking.
DeclaredEach level-up: Max HP -2 Bespoke trigger, stacking, or unit behavior is not used in the ranking.
Wanted-tag roll
On a 5% conditional item roll, the pool is filtered to entries tagged XP Gain when that filtered pool is non-empty.
Across the whole run · all usable families
Ranked by computed DPS for this character
Unlike the opening-pool view, this uses every usable weapon family at its highest usable tier and the full stated weapon-count context. 0 weapons change position versus the generic order.
Reading the Apprentice-specific column:↑ 5 = 5 places higher in this character’s ranking than in the generic, zero-bonus ranking;
↓ 5 = 5 places lower; — = same place.
“generic #N” is the weapon’s rank with no character bonuses applied.
How this ranking is computed
Default 10 Max HP and zero other default combat stats; no bonus stats beyond Apprentice’s direct base deltas; 6 equipped weapons from distinct families; the listed weapon is at its highest tier this character can use. Other equipped weapons are assumed Tier II or III. Item, material, structure, pet, and duplicate-item counters are 0 unless a note below says otherwise.
The target is not cursed, burning, or a boss. Every attack or projectile aimed at the primary target connects. DPS is before enemy armor. Character-local weapon bonuses and the stated equipment counters are applied through the calculator’s staged truncation, rounding, and timing rules.
The lists compare expected direct damage to one primary target per second. They do not combine melee with ranged, and they do not model enemy armor, target density, positioning, closing distance, survivability, misses or spread, overkill, explosions, burning, spawned attacks, later pierce or bounce targets, or conditional target bonuses. Those omitted factors can change results observed during a run.
