Strategy guide
Use the current race tiers to decide what to keep, what to reroll, and when to stop spending limited rerolls without treating new or legacy races as settled data.
Best for: Players choosing whether their current race is worth keeping or rerolling
Quick Answer
Keep Demon or Angel by default. Fairy and Sorcerer are strong specialist keeps when their crit, skill, or ultimate profile fits the build. Dragonkin, Undead, Goblin, and Orc are reasonable holds when rerolls are scarce; Human is the first known race to replace. Do not force a tier onto Archdruid, Night Elf, or Lucifer yet: their published bonuses are useful reference data, but Luck and repeatable live performance still need stronger evidence.
Decision 1
A race tier answers one practical question: if a limited reroll lands here, is it safer to stop or keep spending? It is not proof of a hidden total-power formula, a guaranteed clear, or a universal best weapon. The stable S-D grouping compares published bonus packages and broad build utility. Before committing a reroll, open the live race panel and check the name, star level, listed bonuses, and available pool. The in-game panel is the final authority whenever it conflicts with a community record or an older screenshot.
The current published race record is a community-maintained 1-Star snapshot, not developer documentation. Two independent community tier pages support the established Demon-through-Human grouping, while also showing why newer data needs a more cautious standard. This guide therefore uses tiers for the established set and labels uncertain entries as pending instead of filling the gap with a confident guess.
Use the Tier List page as the comparison screen. Use this guide when the choice is whether to preserve the race you rolled, spend the next reroll, or wait for better information.
Decision 2
Demon and Angel are the safest established stops because their published stat packages help several parts of a build instead of solving only one narrow problem. Fairy and Sorcerer are still strong outcomes, but their value is more tied to crit, skill damage, ultimate charge, and the player's ability to stay active. These are keep recommendations for the current known pool, not a reason to ignore a live panel change after an update.
Keep by default for broad attack, movement, mobility, and build-support value.
Demon is the broadest established all-round package; Angel adds a balanced mix of damage, health, speed, dash distance, and ultimate support. Stop here unless you have a precise, evidence-backed reason to chase a different profile.Keep when the build benefits from critical pressure, skill damage, or faster ultimate cycling.
Fairy is the stronger established crit-and-ultimate specialist. Sorcerer remains a high-value offensive choice, but neither removes the need to solve survival, positioning, or weapon execution.A race that already helps you clear the current route consistently has real account value.
Compare the same encounter after a change. A higher tier has little immediate value when the current bottleneck is gear, positioning, cooldown use, or a route that is already too difficult.Decision 3
The best stopping point changes with the number of rerolls you can actually replace. A player holding only a few Code rewards should protect a playable outcome; a player with a larger reserve can pursue a narrower target. Pick the threshold before rolling so that a disappointing result does not turn into repeated spending without a plan.
Keep Dragonkin, Undead, Goblin, or Orc when they improve the current account's practical needs.
Dragonkin offers a mobile damage profile, Undead provides health and cooldown utility, Goblin supports movement, and Orc is a serviceable early hold. Replace Human first when a reroll becomes available.Use B Tier as a temporary stop and target Fairy or Sorcerer only after the build direction is clear.
Decide whether the route needs broad safety, crit burst, skill damage, mobility, or a more forgiving health package before making a specialist chase.Treat Demon and Angel as established end-state targets, then stop when the rolled package fits the build.
More rerolls do not make a close functional upgrade mandatory. Preserve resources for codes, gear, pets, and Forge decisions once the race no longer limits the route.Decision 4
B Tier does not mean unusable. Dragonkin, Undead, and Goblin each contribute a clear profile while an account is still learning routes and building gear. Orc is a lower-ceiling but practical hold when rerolls are scarce. Human has the thinnest published established package, so it is the clearest known reroll candidate when a free or planned reroll exists. The correct move is based on opportunity cost: do not burn the last reroll replacing a playable race if the next expected improvement is uncertain.
For a build that is failing from damage taken or poor skill availability, Undead can be more useful in the short term than a paper-damage chase. For a player who is missing safe movement windows, Goblin's mobility can support a more repeatable clear. Dragonkin remains the middle option for players who want crit damage, skill damage, and dash distance without waiting for a premium outcome.
Do not label Orc a permanent destination simply because it is comfortable early, and do not treat Human as proof that the account is ruined. Both are ordinary decision states. The meaningful difference is whether keeping the race lets you spend the next reward on the real bottleneck instead of on another unplanned roll.
Decision 5
Public community records now show 1-Star bonus snapshots for Archdruid, Night Elf, and Lucifer. That is enough to compare their visible stat themes, but not enough to publish a durable S-D placement. Luck has a listed value but no well-verified, repeatable in-game outcome in the available sources; independent current performance testing is also limited. A large listed package can be promising without proving the race beats the established pool across real dungeon runs.
+20% Luck, +25% Skill Damage, and -10% Cooldown Reduction are published in the community 1-Star record.
This points toward a skill-cycle profile, but the impact of Luck and full build performance must be tested before assigning a tier.+100% Luck, +10% Attack, +10% Crit Damage, +10% Skill Damage, +20% Movement Speed, and +15% Health are published in the community 1-Star record.
The broad visible package should be checked in the live panel, but it is not ranked above the established S Tier until repeated current comparisons support it.+50% Luck, +20% Attack, +15% Crit Rate, +30% Skill Damage, +30% Movement Speed, and +20% Health are published in the community 1-Star record.
The offensive numbers look substantial, but an apparent stat ceiling is not the same as verified dungeon performance or a known Luck formula.The published values are community-source snapshots dated 2026-07-21. They are displayed as reference data, not used to calculate a hidden score or final tier.
Decision 6
Dragoon belongs to the Season 1 context rather than the current standard race pool. Its documented attack, crit, and movement package can still explain why an existing owner keeps it, but it should not be placed beside normally obtainable outcomes when planning a current reroll path. A new player cannot make a useful budget decision by chasing a historical option that the live pool does not offer.
Decision 7
A responsible tier list changes when the available pool, visible bonuses, star behavior, or actual combat evidence changes. Return to the live panel after race-related updates and treat old screenshots, fixed roll rates, and one-off showcase damage as stale until they are cross-checked. When a pending race receives repeatable testing and the Luck effect is documented, it can move into the established order with a reason that readers can inspect.
Source notes
Official records establish public experience data and destination names. Community sources support visible mechanics and decision categories. Unstable rates, formulas, gates, and reward totals are excluded or clearly limited to an observation.
FAQ
Keep Demon or Angel by default. Fairy and Sorcerer are strong specialist keeps when their crit, skill, or ultimate focus fits the build. With few rerolls, Dragonkin, Undead, Goblin, and Orc are playable holds; Human is the clearest established race to replace.
Not yet. Community records publish their 1-Star bonus snapshots, but the available sources do not establish a verified Luck effect or enough repeatable current dungeon performance for a defensible S-D placement.
No. Keep a B Tier race when rerolls are scarce and it supports the current route. Reroll once the budget is planned and another race would solve a real build limitation rather than only improve a label.
This guide treats Dragoon as a Season 1 legacy race rather than part of the current standard pool. Confirm the live race panel before acting on any future availability change.
The sources used here do not provide a sufficiently current, independently verified rate table. The guide avoids turning old or incomplete odds into spending advice; use the live game panel for the available pool and any displayed rates.
Continue
More guides
