Runes: the system that actually decides your gear's power, not your gear

✍️ Alice・愛麗沙

⚠️ The imprint success-rate table below is Taiwan’s official NEXON NOW disclosure (fetched 2026-08-17) — the one hard number in this article that isn’t a community estimate. The flat combat-power-per-rune figures are a Korean community measurement (single source, confidence marked MED), and the rune catalogue itself (~530 entries) is compiled from Korean wikis across three KR content seasons, most of which Taiwan has not reached yet — this page states which parts apply to TW today.

The stat on your gear is a lie about where it comes from

The single most common new-player misreading on this game: “I found a purple item.” What actually happened is you found an item with a rune imprinted on it, and the colour shown in that gear slot is the rune’s rarity — Advanced → Rare → Elite (purple) → Epic (pink) → Legendary (orange), with a sixth tier, Mythic, added in Korea’s later seasons. The item’s own base attack/defence comes from its item level, and has nothing to do with the colour. This is why “a plain white base can outperform a purple one” isn’t a contradiction — the base stats are identical at the same item level; every point of difference you’re seeing is the rune, not the item.

The catalogue is large, and growing as more Korean content gets compiled in — the count below is a snapshot as of this page’s last edit (2026-09-01), not a fixed number. Across every Korean season compiled into this site’s database as of that date: 530 runes total — 255 legendary, 110 elite, 90 epic, 31 unconfirmed, 26 rare, 18 mythic (mythic is Korea Season 2+ content, not live on TW). By slot: 230 armour, 163 accessory, 96 weapon, 41 emblem. Not all of this exists on Taiwan yet, and the exact catalogue size will have moved since this snapshot — search by slot or class at /en/runes/ for the live, current count rather than the number above.

Imprinting can fail, and failure curses the gear

Imprinting isn’t a sticker — it’s a roll, and Taiwan’s operator publishes the odds directly:

The asymmetry in that right-hand column is the whole system in one fact: emblem-slot imprinting never fails, at any star grade. Every other slot fails a real fraction of the time from ★2 up, and failure isn’t a wasted attempt — it’s an active penalty. The operator’s own text states it plainly: a failed imprint curses the gear (stats drop, disassembly and repair are blocked) until you use a curse-removal potion. One practical note the official text doesn’t emphasise: being cursed does not stop further enhancement attempts on that piece, so there’s no need to panic-clear a curse the moment it happens.

This table also answers “should I stockpile high-star runes before imprinting them?” — at ★5–6 you’re failing three attempts out of four, and each failure costs you a curse-removal potion on top of the failed roll. Veteran habit is to batch high-star imprints into one sitting with removal potions stocked ahead of time, rather than imprinting each one the moment it drops. Low-star (★0–1) is unconditional 100%, which is why “fill every slot first, upgrade later” works as a beginner strategy without any probability risk.

One fact buried in the official text that almost nobody cites: there are two versions of the same-named ★6 rune. A “Legendary Prototype (★6)” and an ordinary “Legendary (★6)” of the same name carry identical special effects, but the Prototype’s base stats are higher — Prototypes drop only from real combat content (dungeons, abyss, team raids), never from boxes. If your rune matches someone else’s name and star count but their panel reads higher, check this before assuming a display bug.

What a rune is actually worth, just for existing

New players tend to think a rune’s value is only its effect text. In fact every rune adds a fixed amount of combat power just by being equipped, scaled by grade — before its actual effect does anything:

With ten imprint slots per class (confirmed identical across all 18 TW-live classes), a full legendary-★6 loadout caps this specific channel at 2,500 combat power — a real number, but worth sizing against the game’s own benchmarks: Taiwan’s own 2026-08-19 event asked for 19,000 combat power as its qualifying threshold, so maxing every rune’s grade alone covers about a tenth of that bar. The other ~90% comes from the rune’s actual effect and from enhancing what you’ve already got, which is the second growth line: imprint enhancement (+1 through +20, spending shards and gold), where every level adds another +5% to this same flat bonus. Grade and enhancement are two independent levers on the same number, not one system.

Five supply lines, ranked by how reliable they are

On Taiwan’s server today, in order of dependability: weekday dungeons and the abyss (the main farm — probability-based loot, but available every day); black holes (the red “leads to deeper levels” variant skews toward emblem runes more than the blue variant); field bosses and event rewards (time-limited — miss the window, it doesn’t come back); self-select boxes like Goro’s Box (think about which season you’re in before opening — a box’s contents can predate content you haven’t reached); and fusion — combining three same-rarity runes into one of the next rarity up. Fusion’s odds, expected cost, and whether it’s currently worth doing are a deeper topic on their own (short answer for Taiwan’s current season: once a slot reaches legendary, stop fusing and redirect resources to imprint enhancement instead — the reasoning is season-timing-dependent and not fully translated into English yet).

One entry-level rule that covers most early mistakes: don’t break down your only usable rune in a slot just to chase a fusion upgrade. Six filled slots always beats one pretty slot — that priority holds until you’re gearing past the early game.

There’s also an invisible sixth supply line: the daily shard trickle. Disassembling spare runes, plus routine loot from black holes and the abyss, both produce Rune Shards — which exist for exactly one purpose, feeding imprint enhancement (the same +1-to-+20 system mentioned above). It’s easy to undervalue because nothing about it feels like an event, but every hole you clear on a normal day is, mechanically, savings toward your next enhancement stage.

Disassemble, overwrite, discard: three buttons, two of them irreversible

• Disassemble is the correct destination for a spare rune — it produces Rune Shards (plus some Spirit’s Traces), and shards have exactly one use: imprint enhancement at the blacksmith. High-star runes get no confirmation prompt for this action — Korea’s community has documented real cases of players disassembling a mythic-grade rune by reflex.
• Overwrite-imprint puts a new rune on already-imprinted gear — the old rune is disassembled automatically, and the enhancement points already banked on it are lost (with partial material refund). The official text states this outright. Using this to “wear something cheap now, swap to the good one later” is a completely valid strategy — just budget for the fact that the cheap rune is a consumable, not an investment.
• Discard is the only button that makes a rune disappear with nothing recovered. The urban myth runs backwards on emblems specifically: disassembling an emblem fully refunds the rune imprinted on it — discarding is the one to be careful with, not disassembling.

First week, condensed to three steps

Step one: fill every slot on your body with some rune, regardless of colour — an empty slot costs you more than a low-grade one does. Imprint the emblem slot first, since it’s the one guaranteed success in the whole system. Step two: prioritise accessory and weapon runes that carry your class’s actual skill tags over generic ones — which runes your specific class wants is a search-by-class question, answered at /en/runes/ rather than a static list, since the database updates as new content lands. Step three: disassemble your duplicates into shards and start feeding imprint enhancement immediately, rather than warehousing them for “fusion later.” For every official probability table this system publishes — imprinting, fusion, enhancement — the numbers-first version is at /en/oddsdb/; check it before committing resources, not after.

Written in English from Taiwan’s official imprint-success disclosure, the flat combat-power-per-rune measurement, and the full rune catalogue, alongside this site’s Traditional Chinese original. Two deeper topics mentioned above — imprint-enhancement order/stop-loss, and fusion timing before a season changeover — are not yet translated into English; their Traditional Chinese originals (fusion) have the full arguments and build-time cost tables.