Korea's star board: why life skills are load-bearing
✍️ Alice・愛麗沙
The Korean 별자리 (constellation board — no official English or Chinese name; Taiwan has not shipped it) is not a separate system. It is the star board inside Soulstream: you connect stars outward from the centre, and the connected stars form constellations that grant permanent bonuses.
One rule on that board answers the whole “is training production worth it” argument: connecting a 항성 (major star) costs red, blue AND yellow Soul — all three. Only a 왜성 (minor star) takes a single designated colour. So an account that only stacks combat power cannot buy its way up the board with red Soul; blue Soul comes solely from life power. Korea denominated the top of its own growth board in a currency only production generates.
What to do today: keep training life skills, and stockpile nothing. Push logging, mining and shearing first — they feed processing-facility upgrades, the single biggest bucket of life power. Safe to ignore: per-star costs (no public table exists anywhere), the gold price ladder, and the 3-for-2 exchange rate. The real bottleneck is not Soul at all — it is connection wait time (1 minute per minor star, 5 per major).
First, what it is and is not
There is no standalone “constellation system” in Korea. Every Korean source describes a star board inside Soulstream whose connected stars form 별자리; the community uses that word for the clickable nodes themselves. It is also completely unrelated to 별의 인장, the Star Seal system that attaches effects to equipment slots — the two share nothing but the word “star”.
Our existing Soulstream coverage handles the income side: every character on a server has its combat power, life power and charm summed, and those totals set how fast three Soul currencies generate. This article covers the spending side — and that is where the production question is actually settled.
One rule explains the whole design
Stars come in two kinds. A 항성 (major star) gives a powerful or unique effect and can be levelled individually. A 왜성 (minor star) gives flat stats — attack, defence, and so on.
The cost split is the interesting part. In kakao.gg’s own words: 「항성은 붉은·노란·푸른 소울 3종류를 모두 필요로 하며, 왜성은 지정된 한 종류만 사용합니다」 — major stars require all three Soul colours, minor stars only one designated colour.
Set that against where each colour comes from:
| Soul colour | Stat it draws from | The five server-wide brackets | Auto-production per hour |
|---|---|---|---|
| Red Soul (붉은 소울) | Combat power (전투력) | 1–50,000 ／ 50,000–100,000 ／ 100,000–150,000 ／ 150,000–180,000 ／ 180,000–200,000 | 0.018 → 0.108 Soul/h |
| Blue Soul (푸른 소울) | Life power (생활력 — the same stat life-skill training raises) | 1–30,000 ／ 30,000–60,000 ／ 60,000–90,000 ／ 90,000–120,000 ／ 120,000–140,000 | 0.018 → 0.108 Soul/h |
| Yellow Soul (노란 소울) | Charm (매력) | 1–20,000 ／ 20,000–50,000 ／ 50,000–80,000 ／ 80,000–110,000 ／ 110,000–130,000 | 0.018 → 0.108 Soul/h |
Source: namu.wiki's Soulstream page (the five-bracket table) + mabilife.com's guide (2026-07-29) + this site's own KR community scan (2026-08, which independently recorded the same 0.018–0.108/h shape). The sum is **server-wide**, not your own characters — mabilife says so in as many words: 「특정 캐릭터 6명만 집계되는 방식이 아니라 서버 내 캐릭터들의 능력치 합이 적용됩니다」 ("it is not that only 6 specific characters are counted; the sum of the characters within the server applies"). What is still open: no source publishes a live server's actual total, so nobody can say which rung a real server sits on — do not use these figures to work out "how much do I need to contribute". Membership adds 0.036/h and a connected star adds 0.054/h.
Every row in that table is required to connect a single major star. An account that maxed combat power and never touched production stops dead at the first major star, no matter how much red Soul it holds — because blue Soul comes only from the server’s life-power total.
That is the answer to “why would I train production”, and it is a structural one rather than a sentimental one: Korea priced the strongest nodes on its end-game board in a currency that only life power creates.
The line we will not cross
Push that one step further and it becomes false, so here is the limit: your own life power does not set your own blue Soul income.
Auto-production runs on the server-wide total, and on a populated server that total is almost certainly pinned at the top bracket already — a single Korean player at the Season 1 processing-facility gate carries 13,000 life power alone, against a top-bracket ceiling of 140,000. So “if I train more life skills the server generates faster” does not hold up on a living server.
What does hold up is the design statement — and one lever an individual really can pull: at 던바튼 고로 you can trade 3 Soul for 2 of another colour. Covering a blue shortfall with red burns a third of it as tax. That is the actual price tag on skipping production: not zero, but a 33% conversion penalty.
Time is the real constraint
Soul is not what runs out first. The launch-day guide that mapped the board’s positions put it plainly: 1 minute per minor star, 5 minutes per major star, every time you connect.
That is why every Korean guide recommends the same operational habit — queue several stars at once so the waits overlap, and start long connections before sleeping or leaving for work. It is also why “auto-production stars first” tops every priority list: a narrow income pipe makes everything downstream slow. The production segment sits at 12 o’clock from the starting point.
Cost rises with distance from the centre. And here is the honest part: there is no public per-star cost table, not even in Korea. The community went as far as building a Soulstream simulator, yet its author, asked outright how much Soul and time a complete board takes, said he had not worked it out — and two months later someone was still asking him to add per-target Soul totals. The only concrete figures in circulation come from one player pushing rightward on launch day: roughly 10 single-colour Soul per stage and 20 on the red side, which led him to conclude that direction was not viable yet. One player, one direction — not a cost table.
Where Soul comes from
The launch-day author listed the sources as dungeon rewards, gold, events and membership. Repeat loot, the package-point shop and a direct Soul-package purchase sit alongside those.
The gold route is 30 purchases per colour per week, per player — not 30 shared by the whole server. This pass corroborated that from a new angle: the same post says 「매주 30개씩, 매주 90만골 들여서 구매 가능」 — 30 of each per week for about 900,000 gold. Three colours times 30 purchases is 90 purchases, and at the opening band’s 10,000 gold each that is exactly 900,000. The arithmetic only closes on the per-player reading.
The ladder by cumulative purchase count: 1–600 at 10,000 gold (rising 20 each), 601–1200 at 22,000 (rising 50), 1201–1800 at 52,000 (rising 100), 1801–2000 at 112,000 (rising 200). Hence the standard Korean advice — let auto-production run and buy only the shortfall; forcing all 30 every week is not automatically worth it.
Resetting breaks everything downstream
A reset returns your Soul. The complication is what else it does. kakao.gg lists four things to check first: how much Soul comes back, which major stars are deactivated when a path is released, how long reconnection takes, and whether you have the gold and materials ready for the engraving work afterwards. The most practical note is the last one — time it against your combat schedule. Rebuilding the board during a raid week means eating the gap yourself.
Priority order
Early: auto-production stars → attack-power major stars near the centre → convenience (bag, weight) → engraving-cost reduction → supplementary stats.
Mid-to-late: attack effects for your main content → class- and build-specific combat effects such as crit → continued engraving-cost cuts → charm and life-power production conditions → high-cost outer major stars.
“Near the centre” is the operative phrase: since cost scales with distance, central attack stars buy the most combat value per Soul.
Where sources disagree, and one retracted rumour
The unlock condition is reported two ways. mabilife names main quest 「다시 한 번, 여신에게」 and guide quest 「소울스트림에 대하여」; namu records it as opening after main quest 「[3장: 팔라딘]」. These are more likely compatible than contradictory — the named quest probably sits inside that chapter — but nothing read this pass proves it, so both are stated. Forced to choose, we follow mabilife: it names specific quests rather than a chapter, and it was written for players actually unlocking the feature.
One retracted rumour is worth recording so it stops resurfacing. On launch day a player claimed only max-level characters counted toward the server total. The next morning the same player corrected himself and called it misinformation. It is false, and its own author withdrew it.
What to do, and what would change this advice
Do: keep training life skills. The reason is now structural rather than hopeful — the board’s strongest nodes are priced in blue Soul, and blue Soul comes only from life power. Life power’s four sources — bag pouches (max 3), processing-facility upgrades, life-skill levels, and rare tools (8, worth up to 880 points combined) — are all outside the cash shop. Of the three Soul colours, this is the only one nobody can outspend you on. Order stays what it always was: logging, mining and shearing first, because they feed facility upgrades and facility grade is the largest single bucket of life power. Nothing needs stockpiling — the system reads character stats, not items you can hoard in advance.
Ignore: the per-star cost maths. No public table exists, and even if one did, the board is a personal choice that memorising early will not accelerate. The gold ladder and the 3-for-2 rate can wait too. Carry two things forward instead: major stars cost all three colours, and the true bottleneck is connection wait time.
What would invalidate this. The “all three colours” rule rests on two guide-level sources agreeing, not on Nexon’s own documentation — an official listing or an in-game screenshot showing major stars taking a single colour would collapse this article’s central claim. The 1-and-5-minute connection times rest on one launch-day observation and may have been patched. The cost figures are one player’s single-direction sample and must not be used as a table. And the standing rule applies: Korea shipping something is not a promise Taiwan gets it, or gets it unchanged — the guild ship went live in Korea on 2025-08-21, opened accidentally on Taiwan for nineteen hours on 2026-08-05, was pulled in emergency maintenance, and still has no official reopening. Lag predicts the shape, never the date.
Conveniently, the advice here does not depend on that risk: even if Soulstream never reaches Taiwan, training life skills still pays through potion margins, crafting chains and guild-ship processing demand. Korea just added one more reason, and moved it from “this might matter someday” to “this is already written into the price”.