Arknights: Endfield Game Mechanics

Endfield's headline damage formula is:

Damage = ATK × Skill% × (1 + Bonus) × Crit × DefMult × ResMult

The full engine chain is longer — the complete form is Damage = Attack × BaseMultiplier × MultiplierGroup × CriticalMultiplier × AmpMultiplier × StaggerMultiplier × FinisherMultiplier × LinkMultiplier × WeakenMultiplier × SusceptibilityMultiplier × IncreasedDMGTakenMultiplier × DMGReductionMultiplier × ProtectionMultiplier × DefenseMultiplier × ResistanceMultiplier × MultiplicativeMultiplier — but for practical deck-building the short formula above is the one that matters, and every term below (Crush, Breach, Arts Burst, Arts Reaction) plugs into it as the BaseMultiplier slot.

• Attack = ((OperatorAttack + WeaponAttack) × PercentageBonuses + FixedBonus + SpecialBonus) × AttributeBonus, where AttributeBonus = 1 + 0.005×MainAttribute + 0.002×SecondaryAttribute.
• Crit: base Critical Rate 5%, base Critical DMG 50%.
• DefMult = 100/(DEF+100) when DEF ≥ 0. A full best-in-slot gear set (140 DEF) gives a multiplier of 0.4167, i.e. 58.33% damage reduction.

There are five elements, each tied to a damage type and a color-coded number: Physical (white) drives all physical damage and the Vulnerable lane; Heat (red), Cryo (cyan), Electric (yellow) and Nature (green) all deal Arts DMG and are the only four elements that generate Arts Reactions.

Every Heat/Cryo/Electric/Nature hit that carries an Infliction component applies an Arts Infliction status, shown as an icon by the enemy's health bar. Two rules govern it:

• Duration: an Arts Infliction lasts 20 seconds and the timer fully resets whenever the same element is reapplied.
• Stack cap: a single element's infliction can stack up to 4 times on one target — this is the '4-stack cap' that maps directly onto the four Status Levels (I–IV) used in every reaction's damage table below. Status Level I = 1 stack, Level IV = 4 stacks.

What happens at each stack depends on what you apply next:

• Same element reapplied → triggers an Arts Burst: a flat 160% ATK hit of that element's damage type, on top of adding a stack (until the 4-cap).
• Different element applied on top of an existing infliction → triggers an Arts Reaction, which consumes the existing infliction entirely.

This is the single most under-explained mechanic in most overviews of Endfield's combat, and it is the actual answer to 'does application order matter': yes, and the two sides of the interaction do different jobs.

• The element you apply last (the incoming hit) determines which reaction fires.
• The element that was already on the enemy determines the reaction's Status Level (strength), based on how many stacks it had accumulated (1–4) at the moment it gets consumed.

Practical read: pre-loading an enabler's infliction to 4 stacks before your reaction-trigger operator swings in is not optional flavor — it is the difference between a Level I reaction (160% ATK payload) and a Level IV one (400% ATK payload, plus longer duration and stronger secondary effects). Sequencing your rotation so the setup unit tags the target repeatedly before the trigger unit lands the final hit is the core combo loop of the entire Arts system.

Nicknames carried over from other games don't match Endfield's own terminology. Use this mapping:

Shatter is the odd one out: it isn't triggered by two Arts elements colliding, it's triggered by Physical (Vulnerable, Lift, Knock Down, Crush, or Breach) landing on a target that is currently Solidified — Physical consumes the Solidification status itself rather than an infliction stack.

All four reactions share the same underlying base formula: Initial DMG = 80% + 80% × (stacks consumed), which resolves to 160/240/320/400% ATK at Status Level I–IV. Everything past the initial hit is reaction-specific.

And the Physical-triggered payoff:

Practical notes solving common confusion points:

• Corrosion has no dedicated main-DPS payoff — it's a resistance-shred support tool. Because the shred benefits whichever element your actual carry deals damage in, Nature operators (e.g. Ardelia, Gilberta, Fluorite) function as enablers, not damage dealers.
• Reapplying the same reaction resets its duration but does not stack its magnitude beyond whatever Status Level triggered it — you refresh the clock, you don't double the DoT or resistance shred unless the new trigger consumes a higher stack count.
• Enemies can also inflict these statuses on your operators (hostile Combustion at 4 stacks, for example, deals 4% max HP initial + 2% max HP/s, ignoring Defense but reducible by Heat Resistance/Reduction/Protect) — worth knowing before you assume every status icon is something you caused.

Physical does not use Arts Infliction at all — it runs its own parallel resource called Vulnerable, which is the actual mechanic behind the 'physical vulnerable lane' question.

• Vulnerable stacks up to a maximum of 4, lasts 20 seconds, and resets its duration (not its stack count) on reapplication.
• Lift and Knock Down (base 120% ATK Physical) apply Vulnerable if the target doesn't have it yet, or add a stack if it already does.
• Crush and Breach are the payoff moves: applying either one consumes every current Vulnerable stack to scale its own damage. If the target has no Vulnerable yet, using Crush or Breach on it just applies a fresh Vulnerable stack instead of paying off.

The underlying formulas (Crush = 150% + 150% × stacks; Breach = 50% + 50% × stacks) match the per-level numbers above exactly.

• Crush deals roughly 3x the raw damage of Breach at every stack count, and gets an additional flat +30% damage if the target is currently Staggered. Use Crush as your Vulnerable-stack payoff when you just want the biggest single number.
• Breach trades raw damage for a follow-up window: after it lands, the target takes 12–24% more Physical DMG from everything else for up to 30 seconds. Use Breach first if your team has more Physical hits queued up afterward, then let those hits benefit from the debuff; save Crush for the actual finishing blow.
• Standard rotation for a Physical-focused team: build to 4 Vulnerable stacks with Lift/Knock Down hits, apply Breach to shred Physical resistance and set up the DMG Taken window, then close with Crush (ideally while the target is Staggered) for the maximum 750% ATK + 30% payoff hit.

Stagger is a separate meter from both Arts Infliction and Vulnerable. It fills from Final Strikes, certain Skills/Combos/Ultimates, and from landing Arts Reactions. Once full, the target becomes Staggered: it is immobilized, the controlled operator's next Basic Attack auto-converts into a Finisher, and Staggered targets take increased damage from follow-up hits (the exact bonus percentage is gear/kit-dependent rather than a fixed base value).

• Physical is the element best positioned to build Stagger, since Lift/Knock Down/Crush/Breach hits and Final Strikes are core to the kit design, and Crush explicitly rewards a Staggered target with +30% damage — this is the mechanical link between the two systems.
• Exact Stagger meter thresholds and the base Staggered-state damage bonus remain unpublished — treat those two specific figures as not yet confirmed rather than a fixed value.

Endfield's skills don't use cooldowns — they're gated by a shared team resource, and misreading this is one of the most common early-game mistakes.

• Capacity: the whole 3-operator squad shares one SP pool, capped at 300 and displayed as three 100-SP bars.
• In-combat regen: 8 SP per second naturally (roughly one full 100-SP bar every 12.5 seconds).
• Out-of-combat behavior: SP normalizes toward 200 outside of fights — it climbs to 200 in about 3 seconds if below that, and decays at 40 SP/second if above 200. In practice this means you cannot walk into a fight banking a full 300; you effectively start every engagement around 200 unless you were already fighting.
• Skill costs: most Battle Skills cost 100 SP (one full bar); a handful of cheaper skills cost only 50 SP.
• Extra SP sources: Final Strikes (15–30 SP depending on the operator — e.g. Fluorite 15, Last Rite 30), Perfect Dodges (~7.5 SP, but this pauses natural regeneration for 0.5s), Finishers (amount varies by enemy type), and select operator skills (notably some Vanguard-class kits) that refund SP directly.
• Important asterisk: SP gained from these 'returned SP' sources is mechanically distinct from natural regeneration — it does not trigger on-SP-recovery passive effects and does not grant Ultimate Energy, so a kit built around 'gain SP → do X' won't proc off a Perfect Dodge refund the same way it does off passive regen.

Rotation strategy this implies: because SP is squad-wide, not per-character, your rotation is a team-level budgeting problem, not three independent skill timers. The efficient pattern is to spend your cheaper 50-SP setup skills (usually on your Arts-infliction enabler) to build stacks toward Status Level IV, hold your 100-SP trigger/finisher skill until the reaction or Vulnerable payoff is fully loaded, and lean on Final Strikes and Perfect Dodges to top the pool back up between engagements rather than waiting on passive regen alone.

Answering the actual question players show up with — 'what do I press and in what order' — the mechanics above collapse into a short checklist:

• Building an Arts team (Heat/Cryo/Electric/Nature): pick one operator whose kit stacks infliction fast and cheaply (your setup unit), and land 3–4 stacks of their element before your trigger operator applies a different element on top. This maximizes Status Level (and therefore the 160→400% ATK scaling, plus every reaction's secondary payoff) instead of wasting a big trigger hit on a Level I reaction.
• Building a Physical team: the loop is Lift/Knock Down to stack Vulnerable to 4, Breach to shred resistance and open the DMG-taken window, then Crush (ideally on a Staggered target) as the finisher. Never lead with Crush or Breach on a target with 0 Vulnerable stacks — you'll just waste the cast applying a single Vulnerable stack instead of cashing in.
• Mixed teams (Physical trigger into an Arts-inflicted target): remember Shatter is the one reaction Physical can trigger directly, by hitting a Solidified (Cryo-reacted) target with any Physical Status — this is the highest-ceiling single reaction in the game (up to 600% ATK) and is worth building a rotation around if you have both a Cryo setup unit and a Physical finisher.
• Don't confuse Arts Burst with Arts Reaction: reapplying the same element is a flat, unconditional 160% ATK poke with no setup requirement — useful as filler DPS, but it will never scale the way a properly-leveled Reaction does. Save your SP-expensive trigger skills for Reactions, not Bursts, once your team composition supports it.
• SP discipline: since the pool is shared, sequence cheap 50-SP setup skills first, and don't let a second operator's 100-SP skill sit unused while SP caps out unused above 200 outside combat — that ceiling is a soft punishment for over-hoarding before engaging.

All figures in this breakdown are confirmed values, except where explicitly marked as not yet confirmed; the core damage formula, crit baseline (5%/50%), and DefMult formula are accurate as presented.

The single biggest source of confusion is that Gear, Weapon, and Essence are three independent progression tracks that all sit under the same "equipment" umbrella, but they are built, upgraded, and sourced completely differently.

• Gear — four slots per operator (1 armor, 1 pair of gloves, 2 kits). Grants DEF plus Attribute/Secondary stats. Crafted from Components + Stock Bills in the Gear Assembly menu, then rerolled via Gear Artificing.
• Weapon — one per operator, fixed to that operator's weapon type. Carries the bulk of base ATK and unlocks three weapon skills as it levels/tunes/potentials up.
• Essence — a consumable material slotted onto a weapon that raises the rank (numeric level) of that weapon's existing skills. It does not add new effects, it amplifies effects the weapon already has.

Because all three feed into the same "is my operator strong enough" question, stopping at just one of them (most players stop at Gear) leaves you unable to actually close your damage gap — Essence and Artificing are where the real ceiling is.

Every operator has a Main Attribute (large ATK weight) and a Secondary Attribute (small ATK weight) among four stats. This is the backbone that every Gear/Essence stat priority decision below is built on.

All four Attributes also feed a shared ATK multiplier: AttributeBonus = 1 + 0.005 × Main + 0.002 × Secondary. This sits inside the standard damage chain (Damage = ATK × Skill% × (1+Bonus) × Crit × DefMult × ResMult, DefMult = 100/(DEF+100), base Crit Rate 5% / base Crit DMG 50%). Practical takeaway: raw Main-Attribute stacking is not optional flavor — it is a direct, uncapped ATK multiplier, which is why Gear and Essence rolls that hit your operator's Main Attribute are worth more than flat ATK% substats in most cases.

This resolves the "where do I forge?" pain point directly: there is no physical forge/anvil to travel to. Gear crafting is a UI system called Gear Assembly, reached by opening the AIC/Explore panel — it is not tied to any map location. It unlocks after completing the Rally and Unite quest in Chapter 1 (Valley IV) and repairing the Hub Base Comms Node; before that quest, the menu simply does not exist, which is the actual reason players report being "unable to find where to forge."

Crafting a piece consumes Components + Stock Bills, with cost scaling by gear level — range: 10 Wood at Level 10 up to 50 Xiranite Components at Level 70.

Gear sets grant a bonus at 3+ equipped pieces from the same set (examples: AIC Heavy, Bonekrusha, Swordmancer, and others) — verify exact set-bonus percentages in-game, as they can shift between updates.

Blueprints for higher-tier gear (Gear Template Crates) are found in explorable zones (e.g. Xiranflow Channel, Tianshi Bureau Academy in Wuling) rather than purchased — worth prioritizing exploration over grinding currency if you're missing a specific set.

Artificing is the mechanism that lets you reroll/raise specific Gear stats after crafting — it only applies to gold-quality gear. Mechanically: you consume same-slot gear as fodder plus an Artificing Catalyst, pick a target stat, and attempt an upgrade.

Priority order: (1) Physical DMG Bonus / Battle Skill DMG Bonus first — these have very few other sources in the whole kit, so Gear is your only lever for them; (2) your operator's Main Attribute (e.g. Agility for an Endministrator-type Vanguard) next, since it feeds the uncapped AttributeBonus multiplier above; (3) spread remaining rolls evenly across multiple pieces rather than dumping everything into one slot — this reduces total material cost to reach a usable floor across your whole set, versus min-maxing one piece to a hard cap.

Each operator is locked to one of five weapon types, and weapon choice, not just rarity, defines which of the three weapon skills unlocks.

Each weapon carries three skills: a generic Attribute buff, a Secondary Stat buff (HP/Crit Rate/damage-type bonus), and a unique conditional effect. 3★ weapons lack the third skill entirely, and lower rarities have weaker skill rank ceilings than 6★.

Leveling caps at 90, with roughly +5 ATK per level and a cumulative cost of 2,524,080 Experience + 341,390 T-Creds to max — level gates sit at 20/40/60/80 and require Tuning to break through. Tuning (5 stages, 0-4) raises the level cap and adds +1 to the rank ceiling of Skills 1 and 2 per stage. Potential (6 stages, 0-5) raises Skill 3's (or Skill 2 for 3★) base and max rank by a combined +1 per stage, up to +5 total. This is where Essence comes in — Potential/Tuning raise the ceiling, Essence raises the actual rank toward that ceiling.

This is the community's most-asked and most-confused topic. Rarity determines how many stat lines an Essence carries and how large a rank bonus it grants your weapon's skills when equipped.

Practical rule: never disassemble a Flawless Essence, even a bad-roll one — it has no Stock Bill value and is only useful as Etching fuel (see below). Everything below Flawless is disposable currency once you have a comparable or better roll.

Etching is how you push a Flawless Essence's individual stat lines higher after acquiring it, by consuming a second Flawless Essence per attempt. The only tier still being finalized is the final Attribute/Secondary jump, marked below.

Stat category	Level transition	Success chance	Coolant Gel to guarantee
Attribute / Secondary Stats	1 → 2	60%	30
	2 → 3	24%	60
	3 → 4	10.9%	120
	4 → 5	5%	250
	5 → 6 (unconfirmed)	2.7%	450
Skill Stats	1 → 2	10.9%	120
	2 → 3	4.2%	300

How the pity works: every failed attempt awards 10 Coolant Gel (does not consume/damage the target Essence, only the fodder Essence is spent). Once your accumulated Coolant Gel meets the threshold for the level you're attempting, you can spend it to force a guaranteed success instead of rolling — effectively a hard-pity currency, not a soft-pity rate increase.

Efficiency rule: level every stat evenly to +3 first — success rates from 1→3 are cheap (60%/24%, avg ~2.7 Flawless Essences per stat) — before pushing any single stat past +3. Skill Stats should stop being rolled naturally past +1 (10.9% is already expensive) and instead be pushed to guarantee with saved Coolant Gel, since 2→3 at 4.2% burns Essences far faster than it burns Coolant.

Essences drop from Energy Alluvium encounters on the overworld map, with rarity gated by enemy level — low-level enemies drop Stable/Clean, higher-level enemies drop Pure. Flawless Essences do not drop reliably from normal overworld enemies.

• Severe Energy Alluvium — a repeatable, Sanity-gated dungeon variant that all normal Energy Alluvium encounters convert into once you reach Exploration Level 4 (unlocked after the "Beyond the World's Expanse I" main-story checkpoint). This is the actual reliable source of Flawless Essences, and it is repeatable as long as you have Sanity to spend.
• Each Severe Energy Alluvium site has its own fixed stat pool, so which specific stats you can roll depends on which site you run — check a site's pool before farming it for a specific weapon skill.
• Pre-Engraving — before starting a Severe Energy Alluvium run, interact with the rift and select "Pre-Engrave" to lock in 3 candidate Attribute stats + 1 Secondary/Skill stat; the resulting Essence guarantees one of your chosen Attribute stats plus your chosen Secondary/Skill stat, with only the third (final) line left random. This costs Engraving Permits, obtained through Regional Stock Redistribution (region-specific, e.g. Wuling Engraving Permit in Wuling).

Practical takeaway: don't farm blind. Always Pre-Engrave for your target weapon's skill stat before running — it converts a 3-random-line drop into effectively a 1-random-line drop, which is the single biggest efficiency lever in the whole Essence loop.

This is a directional framework, not exact math:

Given the AttributeBonus formula (1 + 0.005×Main + 0.002×Secondary), a point of Main Attribute is worth 2.5x a point of Secondary Attribute for ATK purposes — so when a Gear or Essence roll offers a choice between Main and Secondary Attribute lines, take Main every time unless you specifically need the Secondary's other effect (e.g. Agility's damage reduction on a squishy Caster).

Synthesizing the above into an actionable order:

• 1. Base Gear first. Craft full 4-slot gold gear via Gear Assembly before touching Artificing or Essence — cheap, and establishes your DEF/Attribute floor.
• 2. Level operator + weapon. ATK scales close to linearly with weapon level; this is guaranteed value with no RNG, do it before any gacha-like system.
• 3. Push Skill Ranks via Tuning/Potential. Raises the ceiling that Essence will later fill — pointless to farm Essence for a skill rank your weapon can't reach yet.
• 4. Chase Flawless Essences via Severe Energy Alluvium with Pre-Engraving on. Never run blind once Pre-Engraving is unlocked (Exploration Level 4) — it is the highest-leverage efficiency gain in the whole loop.
• 5. Etch evenly to +3, then guarantee past that with Coolant Gel rather than gambling — per the success-rate table, natural rolls past +3/+1(skill) are materially less Essence-efficient than banking Coolant.
• 6. Run Gear Artificing in parallel, spreading rolls across pieces rather than min-maxing one slot, prioritizing Physical/Skill DMG Bonus then Main Attribute.

A few numbers above are still being finalized (notably: exact Component/Stock Bill costs per gear level, the Artificing pity-progress thresholds, and the 5→6 Etching tier) — treat those as directionally correct but verify in-client before making irreversible material spends.

The Automated Industry Complex (AIC) is Endfield's base-building layer: a modular factory system that lets a small team "quickly deploy a fully functional automated production line" instead of manually crafting everything. Every region runs on one Protocol Automation-Core (PAC), the central hub facility, while every Outpost gets its own Sub-PAC sharing the same regional features.

Two hard caps govern how big your factory can grow in a region:

• Protocol Capacity — a slot budget that every placed facility consumes a chunk of. Once maxed, you physically cannot place another facility in that region until you raise it via Regional Development.
• Power capacity — the electrical budget covered in detail below.

Facilities are organized into roughly nine to ten categories depending on which in-game menu you're reading (the AIC overview groups them as Resourcing, Logistics, Production I/II, Power, and Planting; the in-game database further splits out Depot Access, Miscellaneous, and Combat & Support). Most facilities must be individually unlocked through the AIC Factory Plan tech tree before you can build them — you don't get the whole catalog at once.

Items move between facilities and the Depot (your unlimited-slot-count, per-item-capped storage) via Transport Belts plugged into each facility's input/output ports. A belt that looks connected but stops just short of the port is one of the most common reasons a line silently stalls.

Everything starts at a Mineral Bed — fixed map locations that yield Originium, Amethyst, Ferrium, or Cuprium Ore. You extract from them with mining facilities that themselves draw power and must sit inside an Electric Pylon's supply radius before they'll start working:

Yield per node is governed by Mineral Purity, which is a property of the vein's location, not the rig. Purity only ever goes up, and the only way to raise it is leveling Regional Development in that area — higher purity means more ore per mining cycle from the same vein, not a faster rig.

From there, raw ore feeds into Production I/II facilities (crushers, gearing units, reactor crucibles, packaging units, etc.) that convert it into intermediate materials and finally into gear components. The practical implication: a materials shortage is almost never "I need to mine more" — it's usually a bottleneck somewhere in the belt chain or a missing intermediate processor. Trace the chain backward from the stalled output port rather than throwing more miners at the problem.

Every region starts with a base power capacity — 200 power at the outset in Valley IV, confirmed consistently across the AIC overview and independent guides. This 200 is a hard ceiling: total power draw from every active facility (even idle ones, see below) must stay under whatever your current capacity is, and capacity is a separate number from draw that you raise by building generation.

Power physically travels through Electric Pylons (local supply radius around mining/production clusters) and Relay Towers (long-distance backbone). Each cable segment has a fixed maximum reach — commonly cited at 80 meters per hop — so extending power to a distant mining site means chaining relay towers roughly every 80m rather than trying to run one long connection.

Practical design rule: treat your power backbone as its own infrastructure project, laid out and stabilized before you commit to a distant expansion, not improvised after you've already built miners out there and wonder why they won't turn on.

The Thermal Bank is the sole facility that raises your regional power capacity. It converts a fuel item into power over a fixed cycle; connect it to the grid and it contributes its output to the shared pool.

Build cost is 10 Origocrust + 10 Amethyst Part per additional bank, and guides diverge slightly on exactly which tech tier unlocks the facility — the wiki lists Basic AIC I - Power I, while several player guides describe it as gated behind Basic AIC II; treat the precise tier as not fully confirmed but the facility itself, its cost, and its fuel table as solid across sources.

Reading the table correctly matters: raw ore is roughly 6.25 power/second of throughput per bank; an LC battery is 5.5 power/second; an HC battery is 27.5 power/second. Batteries are the only route to real scale — you cannot brute-force late-game power with ore alone — but every battery tier above raw ore requires its own production chain (Packaging Units combining Originium Powder with Amethyst or Ferrium parts), which itself draws power. That dependency is the seed of the entire cascading-failure problem covered next.

This is the single most misunderstood decision in AIC design, and the reason so many bases spiral into repeated blackouts. There is no in-game toggle labeled "reaction priority" vs "output priority" — but the underlying strategic choice is real and it's the crux of stable power design: how do you fuel your Thermal Banks?

• Direct-reaction feeding — pipe raw Originium Ore straight into a Thermal Bank. Low power-per-bank (50 per 8s cycle), but zero dependency: it only needs a miner and a belt. Nothing upstream of it can fail.
• Battery-output scaling — route ore through Packaging Units into LC/SC/HC batteries first, then feed those into Thermal Banks. Massively higher power-per-bank (up to 22x an ore-fed bank at HC tier), but it adds an entire production chain — that chain itself consumes power to run.

The trap: once your grid is battery-dependent, a battery production line is both a power consumer and your power supply's fuel source. If anything upstream (a jammed belt, a stalled Packaging Unit, a moment of over-capacity) interrupts battery output, your Thermal Banks run dry, capacity collapses, and everything downstream — including the very facilities that make batteries — loses power simultaneously. That's the cascading failure: the fuel supply chain and the thing it powers are the same system, so a failure anywhere propagates everywhere.

Working decision framework:

• Never fully retire your ore-direct Thermal Bank once you've moved to batteries. Keep at least one bank on a dedicated, independent Originium miner as a permanent floor — this is your manual "kickstart" line that keeps producing a trickle of power even if the entire battery chain goes dark, giving you enough to manually restart the rest.
• Scale with batteries for bulk capacity, but size the battery-production power draw into your budget before switching it on — don't add battery lines and assume the power they'll eventually produce already covers their own operating cost.
• Physically separate power-generation facilities into their own Sub-PAC cluster distinct from manufacturing, so a manufacturing-side jam doesn't directly compound a power-side one.
• Bring capacity up in increments — add one battery tier, stabilize, confirm surplus, then expand — rather than switching on a large new production block all at once.

Total power loss happens when demand exceeds supply and the base is forced into an emergency/restricted state — reported consistently as dropping you to a bare 200 power reserved solely for dismantling or reorganizing, not for running production. Two under-appreciated causes drive most of these events:

• Overexpansion without matching generation — the single most common trigger. Adding facilities faster than you add Thermal Bank capacity.
• Protocol Stashes — a passive drain that guides flag as the most common reason players get stuck in a blackout loop. They consume power constantly and, unlike a normal facility, cannot simply be switched off — switching a facility off stops its production but it still counts toward power draw, whereas a Protocol Stash must be fully removed ("stashed away") to actually refund its power cost. (Reported by a single guide source; treat the general mechanic — that idle-but-placed facilities still draw power — as the safer takeaway even if the exact terminology varies.)

Recovery sequence, in order:

• Remove or disable the highest passive-drain structures first (Protocol Stashes and any facility you don't currently need) to pull total draw below the emergency threshold.
• Rebuild baseline generation with ore-fed Thermal Banks — they need nothing but a miner and are the fastest facility to bring back online.
• Restart battery production once ore-fed generation is stable, to climb back toward higher capacity.
• Re-enable mining and manufacturing lines last, and incrementally rather than all at once.

Prevention checklist:

• Keep a visible surplus margin (don't build right up to your capacity ceiling).
• Treat every placed-but-idle facility as an active power consumer, not free storage.
• If you must cut load in an emergency, reduce Combat Facilities first — reported as the recommended first cut since they're rarely mission-critical to keep running continuously.
• Monitor total draw vs. capacity in the Automation-Core detail screen regularly rather than only after something has already gone dark — restoration after a full blackout takes considerably longer than prevention.

Protocol Capacity (the facility-slot budget) and power capacity both scale primarily through Regional Development Level, which also raises zipline limits, unlocks Combat Facility transfer limits at higher tiers, and increases Mineral Purity in that region's veins. Approximate progression (values as reported by community trackers; treat exact numbers as subject to patch changes):

Decision implication: Protocol Capacity, not raw materials, is usually what actually limits how ambitious your factory blueprint can be in the early-to-mid game. Before copying a large community blueprint, check its facility count against your current regional cap — a beautifully optimized layout that needs more slots than you've unlocked simply won't fit.

Blueprints let you save a section of your AIC layout and reuse or share it, which is the fastest way to skip the trial-and-error of designing your own production lines from scratch.

• Saving: enter Top View inside your AIC, drag-select the section to capture, save it (adding a name, description, and tags), and submit it for review — status flips to "Share" once approved.
• Sharing: use "Share with Others" to generate an alphanumeric sharing code, postable to Reddit, Discord, or any community hub. Pair it with a screenshot — codes without a visual preview are far less useful to browsers deciding whether a layout fits their needs.
• Importing: open Shared Blueprints in your own AIC menu, select Import, and paste the code.

Where it breaks — the traps new players hit:

• Server lock: codes are server-specific. A blueprint code generated on one regional server (e.g. a North American server) will not import on another (e.g. an Asian server). Always confirm the poster's server before importing.
• Unlock prerequisites: a blueprint will only actually function if you've already unlocked every facility type it uses and completed each facility's introductory tutorial — importing a layout with facilities you haven't researched yet leaves you with placeholders you can't build.
• Not everything is saveable: some structures and building types are excluded from the blueprint system entirely, so even an approved blueprint may not capture 100% of what you see in a screenshot.
• Mastery gap: a blueprint solves layout, not understanding — it won't teach you why a line is arranged that way, which matters the moment you need to modify it for your own resource availability or power budget.

Gear (Endfield's equippable artifice system) is not purchased or simply drop-farmed — its core materials are manufactured through the AIC, gated behind the Gearing Tech node in the AIC Factory Plan. That node is what unlocks Gearing Units and the ability to convert raw ore into gear-grade components.

• Intermediate components come from dedicated processing facilities — for example, a Gearing Unit processes Origocrust and Amethyst Fiber into Amethyst Components, one of the base building blocks for lower-rarity gear.
• Liquid-state reactions (used for higher-tier component chains) run through the Reactor Crucible, a Production II facility (reported: 50 power draw, 20 Ferrium Parts to build, unlocked via Wuling AIC Plan - Wuling AIC I - Liquid Reaction) that combines powders with Clean Water, Sewage, or Precipitation Acid into solutions and effluents at roughly 30 items/minute per input line.
• Higher-rarity gear requires clearing specific story missions before its blueprint becomes available at all — no amount of factory optimization substitutes for mission progress on the highest tiers.
• Catalysts, used alongside manufactured components during the actual artifice/upgrade step, are purchased separately from each region's Stock Redistribution System rather than produced in the AIC.

Practical takeaway for planning a gear-farming factory: map your target gear piece's material list first, identify which materials are AIC-manufactured vs. mission-gated vs. Stock Redistribution purchases, and only then design the production line — building a full manufacturing chain for a component you could have bought outright, or that's still mission-locked, is wasted Protocol Capacity.

The single most common source of confusion for new players is assuming Endfield has one unified pity counter like most other gacha games. It does not. Headhunting (Operators) and Arsenal Exchange (Weapons) are structurally different systems — different currencies, different pity lengths, different guarantee logic — and progress in one never affects the other.

Note the Arsenal Exchange quirk: one "pull" (called an issue) is not a single item — it hands you 10 weapons in one transaction. This changes how its pity counters should be read; see the Arsenal Exchange section below.

Base drop rates:

Headhunting layers two separate safety nets on top of these base rates, which is more generous than most competitors and worth understanding precisely:

• Mini-pity (5★ floor): at least one 5★-or-better Operator is guaranteed every 10 pulls — if pulls 1-9 in a block whiff on 5★+, pull 10 is forced to be 5★ or higher.
• Soft pity (6★ ramp): if no 6★ has appeared by pull 65, every pull from pull 66 onward gains a flat +5.0% additive rate, so pull 66 sits at 5.8%, pull 67 at 10.8%, and so on.
• Hard pity (6★ guarantee): pull 80 is a forced 6★ if none has appeared yet (still subject to the 50/50 below).

Based on modeling the exact rate curve above, the expected number of pulls to your first 6★ of any kind is approximately 54 pulls, with a median of 67 pulls. Cumulative odds of hitting a 6★ by a given pull:

The practical takeaway: the soft-pity ramp is steep and short (only 14 pulls wide, 66→80), so in practice most 6★s land in the 66-75 window rather than being smoothly distributed — budget for "around 70," not "around 40."

This is the mechanic that trips up players coming from other gachas, so it deserves to be stated plainly: every 6★ you pull (soft-pity, hard-pity, or lucky early hit) has an independent 50% chance to be the banner's featured Operator and 50% chance to be a random off-rate-up 6★. Unlike most other gacha games, losing the 50/50 does NOT guarantee the next 6★ will be featured. You can theoretically lose the 50/50 multiple times in a row within one banner.

What actually caps the pain is a separate, banner-specific counter:

• 120-pull guarantee: if you haven't obtained the featured 6★ within the first 119 pulls on that Chartered banner, pull 120 is forced to be the featured Operator. This is a hard ceiling, not an average.
• 240-pull dupe layer: a second, layered guarantee additionally hands you a potential token (i.e. a duplicate/copy) of the featured 6★ by pull 240 if you keep pulling on the same banner — useful context if you're chasing Potential ranks rather than just the first copy.

Based on a Monte Carlo simulation over 400,000 trials using the rate table and 50/50+120 rules:

In plain terms: roughly one in three players will lose the 50/50 badly enough to ride the safety net all the way to 120 pulls (≈60,000 Oroberyl). Budget for 120, not for the 50/50's theoretical "half the time it's 55 pulls" — the tail risk is common enough that it should drive your saving target, not the average case.

This is the second most-asked question and it's easy to misread how it works. Here is the precise breakdown:

Practical implication: it is safe to pull on a Chartered banner right up to a soft-pity-adjacent number (e.g. 55-60) even if you don't plan to finish this banner, because that progress is not wasted — it rolls into the next limited banner's hard-pity count. It is not safe to assume the same about the 120-pull featured guarantee or about weapon pity.

Basic Headhunting is the permanent, non-time-limited banner (live since Jan 22, 2026). It draws from a fixed standard 6★ roster — as of writing, Ardelia, Pogranichnik, Last Rite, Ember, and Lifeng — and, critically, has no rate-up and no 50/50: any 6★ you pull here has an equal chance of being any operator in that five-person pool.

• Same 91.2%/8%/0.8% base rates and the same 65-pull soft pity / 80-pull hard pity structure as Chartered banners.
• 300-pull one-time selector: the first time your cumulative Basic Headhunting pulls reach 300, you receive a one-time selector letting you pick any one of the five standard 6★ Operators outright. It does not expire and does not refresh after a second 300 pulls.

Recommendation: because Basic Headhunting pity and Chartered pity are fully separate pools, spending Oroberyl here is a genuinely different investment from limited-banner pulls, not a "waste" — but given the 300-pull threshold for the guaranteed pick, it only pays off as a deliberate long-term project, not spare-change spending.

The weapon gacha is easy to misread because its numbers are quoted per issue (one 1,980-Arsenal-Ticket transaction that returns 10 weapons at once), not per individual weapon. You may also see the same milestones expressed in "individual weapon pull" terms (×10), which is why you'll see both "10th issue" and "100th weapon" describing the same checkpoint — they are not contradictory, just different units.

• Mini-pity: at least one 5★-or-better weapon guaranteed within every 10 weapons (i.e. within a single issue).
• 6★ pity: 3 consecutive issues without any 6★ forces a 6★ on the 4th issue.
• Rate-up mechanic: a 6★ weapon has a 25% chance (not 50/50 — noticeably worse odds than characters) to be the featured rate-up weapon.
• Rate-up guarantee: if 7 consecutive issues pass without the featured weapon, the 8th issue guarantees it.
• Arms Offering milestone system: the 10th issue grants a selector box (pick any non-featured 6★ weapon); the 18th issue grants the featured rate-up weapon outright; this then alternates every 8 issues thereafter (selector, guaranteed rate-up, selector, guaranteed rate-up…).
• No carry-over: unlike character Chartered-to-Chartered pity, weapon issue-progress resets completely when a weapon banner rotates.

Because rate-up odds are only 25% per 6★ (vs. 50% for characters) and there is zero cross-banner carry-over, Arsenal Exchange is structurally the more expensive, higher-variance system. Treat weapon pulls as optional min-maxing, not a budget priority equal to character Headhunting, unless the featured weapon materially defines a specific character's kit.

Currency conversions:

Free Oroberyl comes from daily tasks, weekly Operational Manual objectives, the Protocol Pass, story missions, exploration/gathering chests, and the Credit Store; permits (Basic/Chartered HH Permit) come from similar sources plus the Acquisition Center and limited-time events.

Monthly free-income estimates vary: roughly 3,000 Oroberyl/month from dailies plus ~1,800-2,400/week from weeklies (≈2 ten-pulls/month from routine tasks alone) is one benchmark, while a broader estimate of ~100 free pulls across a full ~49-day version cycle applies once one-time story/exploration/event rewards are included. Treat both as directional ballparks — one-time content-drop bonuses (new area unlocks, Protocol Pass tiers, story chapters) can swing the real number substantially between patches.

• Save in units of 120, not 80. Since roughly a third of pulls will need the full featured-character guarantee (see the 50/50 section), budgeting for the 80-pull hard pity alone will leave you short about a third of the time. Aim your savings goal at 120 pulls (60,000 Oroberyl) per character you actually intend to guarantee.
• Don't fear "wasting" pity by stopping early. The 65/80 counter carries into the next Chartered banner, so pulling partway on a banner you're lukewarm about is not lost value — only the 120-pull featured guarantee and any in-progress weapon issue count are lost when a banner ends.
• Treat Basic Headhunting as a separate long-term fund, not overflow spending. Its pity does not interact with Chartered banners at all; only commit here if you're deliberately working toward the 300-pull selector or specifically want one of the five standard 6★s.
• Deprioritize Arsenal Exchange unless the weapon is build-defining. With only a 25% rate-up chance per 6★ (vs. 50% for characters) and zero pity carry-over between weapon banners, it is the least F2P-efficient system in the game — spend Origeometry on Oroberyl first, Arsenal Tickets second.
• Watch for non-gacha character paths. At least one operator has been offered through a separate "Bond Quota Exchange" token system rather than the standard banner pull pool on a recent Chartered banner — check each banner's specific reward track before assuming every featured-adjacent character requires pulling.
• If you're chasing Potential (dupes), plan around 240, not 120. The layered dupe guarantee (see above) means committed spenders aiming for early Potential ranks should treat 240 pulls as the realistic ceiling for banner investment, not 120.

Endfield deliberately splits Operator strength into four systems that cannot substitute for one another. Understanding the dependency chain is the single most useful thing a new player can learn, because it explains why simply dumping EXP into a favorite Operator stalls out fast.

• Level (1-90): raw stat scaling (ATK, HP, and derived attributes). Increased with EXP items and T-Creds.
• Ascension / Promotion (E0-E4): unlocks higher Level caps and unlocks nodes on the Ability Matrix. You cannot Level past a cap without Ascending first.
• Ability Matrix (skill leveling): Basic Attack, Battle Skill, Combo Skill and Ultimate each rank up independently, gated behind Ascension tiers.
• Potential (P0-P5): dupe-based stat and Combo Skill/Ultimate enhancements, gated behind pulling the same Operator again.

Because Ascension unlocks both the next Level cap and Ability Matrix nodes, it functions as the true bottleneck of the chain — you cannot buy your way around it with EXP or Skill materials alone.

Operators level from 1 to 90, but the climb is broken into hard walls at Lv. 20, 40, 60 and 80 — each requires an Ascension before leveling can continue. Operators below Level 20 also gain a trickle of EXP simply from defeating enemies in the open world, so early leveling is nearly free.

Cumulative cost to push a single Operator from Lv. 1 to Lv. 90 is approximately 1,792,290 EXP and 385,420 T-Creds. EXP materials drop from Protocol Space ("Operator EXP" domains), story missions, Stock Redistribution, and the Acquisition Center.

Promotion (Ascension) is the gate between Level brackets. Each tier costs T-Creds plus a rank-specific material, and the jump in cost from E3 to E4 is steep and shared across your whole roster, making it the most common mid-game wall.

Practical read: the E4 rare materials (Triphasic Nanoflake, Quadrant Fitting Fluid, etc.) are the same pool used for Mastery skill upgrades and top-tier weapon Tuning, so E4-ing multiple Operators simultaneously competes directly with maxing any single Operator's Ability Matrix. Don't E4 a bench unit before M3-ing your main DPS's Battle Skill.

Each Operator has four independently-leveled abilities — Basic Attack, Battle Skill, Combo Skill, and Ultimate — each progressing through base ranks up to Mastery 3 (M3). Structurally this mirrors the ability ranks in the original Arknights but applies it to all four ability slots rather than one.

Maxing a single ability line to M3 costs approximately 172,200 T-Creds, roughly 82 Protoprism, and roughly 118-130 Protohedron depending on patch version.

Priority order within the Ability Matrix: Battle Skill first, then Ultimate, then situational Combo Skills, with Basic Attack last. Battle Skill and Ultimate carry the largest share of an Operator's damage or utility per material spent; Basic Attack scaling is comparatively flat and only worth maxing once everything else on that Operator is done.

Potential is Endfield's duplicate system, running P0 (base, no dupes) through P5 (max), requiring six total copies of an Operator. Pulling a copy you already own converts automatically into an Operator Token, which is spent to raise Potential by one stage.

• Effects scale with rarity and Operator design, but generally include base stat increases plus enhancements to Combo Skill and Ultimate behavior (not raw Level or Battle Skill).
• Excess tokens can be exchanged for AIC Quota — a max-Potential 5-star token converts to 20 AIC Quota, a 4-star token to 5 AIC Quota.

Why Potential is correctly prioritized last: every dupe pull competes directly against pulling a new featured Operator on the next banner, and against saving Oroberyl toward a 120-pull featured guarantee (see Gacha Pity below). Since Potential does not touch Level, Ascension, or the Ability Matrix, an under-leveled P5 Operator is still weaker than a properly built P0 one. Spend dupes on Potential only after your core roster is fully leveled and skilled.

This is the single most repeated question in the community, and the answer converges on one order:

Roster width matters more than any single system. The most common early-game mistake is spreading materials across four or more Operators instead of committing to two full teams. Two fully-built teams consistently clear more content than four half-built ones.

Level 80 is the practical stopping point for most DPS Operators; pushing 80→90 costs roughly 2,000 Sanity-equivalent investment for only a 14.3% base attribute increase, and is only worth it on your single best-built hypercarry after every other priority is satisfied.

Free-unit note: the Endministrator (player avatar) and Ardelia (the free 6-star healer) are day-one investment priorities since they're guaranteed, always available, and slot into nearly every early team.

Every meaningful farming action — EXP materials, Ascension materials, Skill materials, weapon Tuning materials, Essences — is ultimately paid for in Sanity. Understanding its math is more important than memorizing any single item cost, because Sanity is the true constraint on how much progression you can do per day, not currency stockpiles.

Recovery outside natural regen comes from consumable items (Sanity Syrup / Emergency Sanity Booster-type items, roughly +40 Sanity per use) and from converting the premium currency Origeometry directly into Sanity — the first daily exchange for 40 Sanity costs 1 Origeometry, with the Origeometry cost increasing for each additional exchange the same day (soft rate-limiting to discourage pure pay-to-skip).

Practical implication: at 200 Sanity/day, a 60-Sanity stage only allows about 3.3 clears/day from pure regen — this is why the game funnels players toward Protocol Space (material-efficient farming with scaling rewards by difficulty) rather than repeatedly clearing story content for EXP.

This is the progression pain point almost no beginner guide covers, because it only matters once you reach Valley IV endgame content. Essences are weapon/gear-tier upgrade drops with randomized secondary stats, farmed from Severe Energy Alluvium runs. Left to pure RNG, getting a "clean" roll on the stat you want can take dozens of runs.

The Valley Engraving Permit removes that randomness for one run: use it before an Alluvium clear and you pre-select one guaranteed stat/attribute on whatever Essences drop from that run, at the cost of consuming the permit.

Economy-level takeaway: natural weekly Sanity regen (200/day × 7 ≈ 1,400/week) funds only about 21 Alluvium runs per week (about 24.5 with a monthly Sanity pass). Since that's roughly the same order of magnitude as weekly permit restock, players do not meaningfully out-accumulate permits by farming harder — the recommended strategy across guides is to bank a buffer of roughly 20 permits and spend them only on your actual best-in-slot stat targets, not on every run.

Endfield's pity system is not a simple "pity carries, 50/50 loss guarantees next" model like mobile Arknights — its structure directly changes how you should budget Oroberyl/Origeometry against the leveling costs above.

Base 6-star rate is 0.8% (5-star 8%, 4-star 91.2%). The weapon banner runs different odds (4% / 15% / 81% for 6/5/4-star weapons), with its own 80-pull featured pity and a 100-pull selector for non-featured 6-stars.

Why this matters for economy planning: because the 120-pull featured guarantee does not carry over between banners, currency saved toward "almost 120" on an banner you don't finish is not wasted at 80 (you still walk away with a 6-star) but the featured-unit guarantee resets. Budget Oroberyl in blocks of 120, not 80, if a specific unit is the goal — pulling to exactly hard pity and stopping is the most common way players end up with an off-banner 6-star instead of the unit they wanted.

• Let Levels 1-20 come free from open-world kills; don't spend EXP items until an Operator is committed to your final 2 teams.
• Ascend reactively — only when a Level cap or a needed Ability Matrix node is actually blocking you, not preemptively across your whole roster.
• Within the Ability Matrix, always fund Battle Skill and Ultimate before Combo Skill or Basic Attack.
• Commit to two teams (roughly 6-8 Operators) rather than partially building four or more.
• Prioritize Level and Ability Matrix on the Endministrator and Ardelia (free, guaranteed, broadly useful) before investing in early gacha pulls.
• Spend Sanity on Protocol Space / material-specific domains over repeat story clears once story EXP income drops off.
• Save Origeometry/Oroberyl toward 120-pull blocks per targeted banner rather than stopping at 80 hard pity, unless a random 6-star is an acceptable outcome.
• Don't rush Valley IV Engraving Permit spending — stockpile toward a buffer (~20 as a stable target) and reserve them for stats you've confirmed you actually need, since weekly Sanity income does not comfortably outpace weekly permit restock.
• Treat Potential (dupes) as the last system funded, after Level/Ascension/Ability Matrix are handled on your core roster.