Every gacha and pity system, explained from the official rate tables
✍️ Alice・愛麗沙
⚠️ Rates below are Taiwan’s official published probabilities (NEXON NOW disclosure, fetched 2026-08-17) unless labelled otherwise; KR-only mechanics are tagged KR. A future global service’s rates are unknown — regional disclosures can differ, so treat this as the TW/KR record, not a promise.
The headline rates
| System | Rate | Hard floor |
|---|---|---|
| Premium fashion pull (epic) | 2% per draw | 100-pull box pity — and it re-arms: after claiming, further pulls count toward a second claim in the same rotation |
| Premium pet pull (epic) | 1% per draw | Hard pity at 200 pulls |
| Fashion legendary synthesis (epic→legendary) | 20% | Shard system (below) |
| Fashion/pet combine (3 same-grade → 1) | Published per-grade rows | 10% “Great Success” doubles output on success |
Two structural notes. Legendary pets come only from combining, never from pulls. And fashion grades run Normal → Advanced → Rare → Elite → Epic → Legendary; promotion below Rare ignores slot, from Rare up it demands three same-slot, same-grade pieces (board-verified rules, official ladder).
The four pity mechanics are different machines
The game says “pity” about four mechanics that behave nothing alike. Confusing them is how people misbudget:
1. Point-bar pity (enhancement). Each failed enhancement attempt at +11 to +20 banks points; at 100% the next attempt succeeds outright, consuming no materials, then the bar resets. Failure never destroys or downgrades the gear. From +15, a catalyst adds +1% success per attempt. This one is a cost ceiling: the worst case is computable in advance, and our odds page computes it.
2. Counter pity (pet box). A dumb counter: 200 pulls, guaranteed. Nothing carries between banners.
3. Shard-conversion pity (fashion endgame). Failures pay out shards that auto-convert at 100:
- Failed synthesis (Epic→Legendary): 5 shards per failure, per slot → at most 20 failures before a guaranteed legendary fashion box for that slot.
- Failed promotion (Epic→Legendary): 7 promotion-ticket shards per failure → at most 15 failures to the guaranteed ticket.
- Failed pet ascension: 7 shards per failure, same 100-shard conversion → 15 straight fails is the hard pity.
The design consequence people miss: at these steps, trying now costs nothing extra — every failure is a step toward the floor. The steps without shards (mid-ladder fashion promotion) are the opposite: a failure just burns material, and waiting only buys confidence, not better odds.
4. Fragment pity (runes). Failed ★8 rune upgrades and dismantles grant Unknown Rune Fragments; 100 of them guarantee a random unowned ★8 legendary. (KR’s Season 2 later added a broader legendary pity, retroactively — labelled KR; TW hasn’t received it.)
The mean is a lie — read the 90% column
A 2% item averages 50 pulls; a 0.5% item averages 200. But averages describe nobody: at 0.5%, one player in ten is still empty at 460 pulls. Before starting any chase, the number to read is not the mean but the pull count at which 90% (or 99%) of players have succeeded — that’s the budget that protects you from being the unlucky tenth. Our probability database computes the 50/90/99 columns for every row the operator publishes, which is exactly the arithmetic the official page doesn’t do for you.
Worked honesty from the same tables: the “price per pull” isn’t even a list price on TW — fashion tickets sell only inside packs (the monthly premium pack works out to ~122.5 M Coins per ticket), so every cost estimate is a package average.
What the disclosures don’t cover
Being precise about the boundary matters more than the rates themselves:
- Enhancement +1 to +10 exact rates are shown only in-game (community: near-100%).
- Mystic Dice have no published table at all — every rule known about them is TW player testing.
- Curse trigger odds (from failed high-stage enhancement) are undisclosed; even the stage where curses start is community-disputed (+4 vs +5).
- Dungeon drop rates: never published anywhere.
Where an official number exists we use it; where it doesn’t, this site says so instead of inventing one. That single habit is most of what separates a usable gacha guide from a confident-sounding wrong one.
The one-paragraph spending rule
Never start a chase without its floor in view: enhancement’s worst case is computable, the pet box caps at 200, fashion’s endgame failures bank shards — but mid-ladder promotions and anything without a published rate are open-ended. Budget to the 90% column, not the mean; count the pity you’d forfeit by stopping early; and remember the fashion box pity re-arms, which quietly makes the second hundred pulls better value than the first. Everything deeper — what’s worth buying at all — lives in the spending guide.