Getting Started with Guildrun

Guildrun is a single-player PvE roguelike autobattler from Leyline. You draft a small guild, place heroes on a hex board, then watch combat resolve automatically. The real game happens between fights: shops, events, ranks, items, and relics. This guide is written for the current Demo 0.5.7 mindset — not as a rigid tier list, but as a checklist of decisions that keep early runs alive long enough for you to learn.

Use this page together with Guildrun Hub’s databases: Heroes, Relics, Items, Classes, Keywords, and Builds. For deeper pillars, open the Player Handbook (formation / first contact) and the Growth Route (Rank C→S and expansion). Look up a name when a shop offer confuses you; come back here when you need the big picture.

What you are actually optimizing

New players often treat Guildrun like a collection game: grab every shiny relic, fill every board slot with copies, and hope auto-combat works out. Strong runs do the opposite. They answer four questions after every fight:

1. Did someone have a clear job? Frontline absorb, backline damage, or support uptime.
2. Did the first contact go wrong? Range, splash, and who got hit first usually explain the wipe better than “my DPS was low.”
3. Did I spend Shards on the hero who wins the next fight? Rank spikes beat spreading upgrades evenly.
4. Did my relics ask for tags my team can actually trigger? Rush, Crit, Shield, Stall, Burn, and class tags only pay off if someone fires them every combat.

If a loss feels random, write one sentence about which of those four failed. That habit will teach you faster than any fixed “best team.”

The run loop in plain language

A typical loop looks like this:

Choose → Place → Fight → Diagnose → Spend → Repeat.

• Choose heroes and offers on the route (shops, events, campfires, challenges).
• Place the formation so the right unit eats the first hit and casters get a cast window.
• Fight resolves without micromanagement.
• Diagnose the replay in your head: who died first, who never cast, who wasted a Rush window.
• Spend Shards on ranks, items, rerolls, or the next recruit — not on vibes.

The Demo supports mid-run saves, so you can pause after a hard fight, look up a relic or hero on this site, then continue with a clearer plan.

Difficulty: start low on purpose

The Demo exposes an escalating difficulty ladder (Base through higher rungs, including harsh modifiers such as reduced Shard income or item caps on higher settings). Higher rungs often stack lower ones, so jumping straight to a punishing setting multiplies every mistake.

For your first clears, stay on the easiest setting until you can reliably:

• keep a frontline alive through Act pressure,
• rank one carry to B before a major challenge fight,
• and explain why you took each relic.

Climbing early is optional ego. Learning the shop and formation language is not.

Drafting: jobs first, synergy second

You usually open with a small roster and grow toward a larger field plus a reserve. Treat the board as job coverage, not a museum of favorites.

A stable early formation almost always wants:

• Someone who can take the first hit (Tank / Vanguard–style frontliner).
• Someone who deals clean damage (melee bruiser or ranged carry).
• Optional third job: sustain, shields, control, or economy support.

Perfect tag synergy is a mid-run luxury. If your board cannot survive the next pack, a “dream” Assassin triple is just three corpses in a trench coat.

When you add a fourth or fifth hero, ask: what new job did I unlock? Adding a second copy of the same fantasy before the first one works is a common soft-lock.

Browse the hero roster and filter by class when you are unsure what a name actually does. Dual-class heroes (for example paths that add a second class at Rank B) are powerful, but only after you understand what the current class pool will offer later.

Positioning: free wins you are ignoring

Auto-battlers punish lazy placement. Before a fight, spend ten seconds on:

• Front row: tanks and high Max HP bruisers.
• Back row: fragile Mystics, Mages, and long-range Assassins.
• Adjacency: some auras, shields, and splash patterns care about neighbors — leave gaps if enemies punish clumps.
• Opposite lanes: leap / dive threats may land on the far side; do not park your only healer on the dive tile by default.

After a loss, re-open the formation before you blame the relic pool. Many “bad RNG” runs were actually a backliner eating the opening volley.

Ranks C → B → A → S (growth without panic)

Heroes climb ranks through duplicates or eligible campfire-style events. Think of ranks as a growth route:

| Rank | What it usually means for you | | --- | --- | | C | Base kit. Confirm the hero already does a useful job. | | B | Choose one of three specializations. Some paths add a class and change later offers. | | A | Stat package + a class modifier from the active class pool. | | S | Demo ceiling: another package + final modifier (and often more item flexibility). |

Practical order for beginners:

1. Stabilize the board at C with clear roles.
2. Specialize the hero who is winning fights now (first B spike).
3. Expand jobs instead of stacking identical copies.
4. Draft a modifier that fires every combat, not a cute one-off.
5. Finish the carry; use the last slot for matchup answers.

On Guildrun Hub, open any hero page (for example Fiona or Irini) to see Rank C–S art, abilities, and specialization names. That visual ladder helps you remember what you are buying when the campfire appears.

Items vs relics: different jobs

Items usually sit on a hero — local power, triggers, and personal survival.

Relics rewrite the run: economy rules, team-wide keywords, quests, and engines that fire every fight.

Beginner rules of thumb:

• Do not spray rare items evenly across the roster. Feed the unit that already has a job.
• Early relics that improve Shard income, reroll cost, or consistent combat payouts often beat a flashy combat relic you cannot trigger yet.
• An off-theme Epic that needs Rush while nobody on your board Rush-triggers is worse than a Common that completes a tag you already own.
• If a run-defining relic appears but your specializations point the wrong way, consider whether a pivot is cheaper than ignoring the relic. Multiclassing and Rank B paths exist so starting class is a suggestion, not a prison.

Use Relics and Keywords when a description mentions Rush, Crit, Shield, Stall, Burn, Frost, or a class name. Clicking the keyword shows related pieces so you can judge whether the rest of your shop will support it.

Economy: Shards are tempo

Shards buy recruits, ranks, items, and rerolls. Greedy players either:

• spend everything instantly and cannot afford the Rank B that would have won the challenge, or
• hoard forever and die with a full wallet while the board stays Rank C.

A healthier pattern:

• Keep enough liquidity for the next important rank or a key item.
• Buy economy help early if the shop offers it — interest and per-win Shards compound.
• Reroll with a purpose (“I need a frontliner” / “I need a Shield item”), not out of boredom.
• Before a boss or challenge node, empty the wallet on power you can use this fight.

If higher difficulties cut Shard income, every wasted reroll hurts more. Practice clean spending on Base first.

Keywords worth learning early

You do not need the whole encyclopedia on day one. Learn the language that shows up in shops:

• Rush — short aggression windows; great with attack-speed and tempo comps.
• Crit — burst and execute-style payoffs.
• Shield — absorb before HP; pairs with support Mystics and durable Warriors.
• Stall — delay / control clocks that enable later payoffs.
• Burn / Frost / Poison — status engines that want dedicated applicators.
• Class tags (Warrior, Tank, Vanguard, Assassin, Duelist, Mystic, Mage) — many relics care who you actually field.

When an ability or relic text highlights one of these words, treat it as a contract: build the team that can sign it every combat.

A simple first-run plan

You can ignore meta names and still clear with discipline:

1. Draft one reliable frontline and one clear damage dealer.
2. Keep the formation simple: tank front, carry back.
3. Prefer combat and shop routes that teach you enemies and income — not only the scariest looking path.
4. Get one hero to Rank B before the first major gate fight.
5. Mid-run, name your engine out loud (“Shield Mystic + economy relics” or “Rush Duelists”) and stop drafting against it.
6. Final shop: spend remaining Shards, fix placement for the boss pattern, then commit.

If you want curated starting comps while you learn, skim Builds. Treat them as examples of job coverage, not scripts you must force every seed.

Common beginner traps

• Four copies of the same fantasy before anyone has a second job.
• Parking the healer on the dive hex.
• Grabbing every relic before the team can absorb variance.
• Ignoring mana / cast cadence on Mystics and Mages while stacking raw Attack on them.
• Forcing yesterday’s winning board into a seed that offered a completely different relic engine.
• Blaming RNG without checking who died first and who never cast.

After each failure, change one variable: placement, who got the next rank, or which relic you accept. Changing everything at once teaches nothing.

How to use Guildrun Hub while you play

• Lookup a hero’s active, passives, and Rank B names before you lock a specialization.
• Search a relic name, then follow its keywords to related options.
• Open a class page when you are deciding whether a dual-class path is worth the future modifier pool.
• Return to this guide when you feel lost in the shop — the answer is usually jobs, positioning, or Shard timing, not a missing legendary.

Closing mindset

Guildrun rewards players who treat every fight as feedback. Stabilize a boring, correct board. Specialize the unit that is already working. Let relics choose the late identity only after the early board can survive. Climb difficulty when your losses become explainable, not when you simply want a harder badge.

Clear one calm run on Demo 0.5.7 with that loop, and the rest of the systems — Endless, harsher rifts, deeper multiclass pivots — will suddenly read as options instead of noise.