Methodology

Core principles

1. Each patch, we update the site. We verify as much data as we can, add new pages for what the patch introduces, and upgrade existing guides where they need it.
2. Pages are signed.Each entity page surfaces a visible “Last reviewed by Trakan/Irissia” byline that links back to Trakan / Irissia. No anonymous edits.
3. Stale flags surface.Pages last reviewed more than 60 days ago render a “Pending re-test” pill until one of us has verified them against the current patch.
4. Pages are disposable.Anything we publish can be re-tested and re-ranked. We don't protect bad takes for consistency.
5. One game, two writers.We don't scale by adding authors. We scale by limiting scope to Mongil and going deeper.

How tier ranks are produced

Each character is ranked by how they actually perform on current end-game content for their role, not by theory-crafting in a spreadsheet. The rank reflects what we've verified in-game; if a unit was untested at the rank's deadline, the page says so.

The tier scale

• SS — top of their role on current content, regardless of team. Replacing them with an S-tier alternative costs a clear.
• S — top performer with a meaningful condition (specific team, specific encounter). Slot in if you have the condition; otherwise an A-tier in that role.
• A — competitive in their role; rank dependent on roster gaps.
• B — playable but outclassed for most current content. Niche use cases noted on the page.
• C / below— niche-only or outright outdated. We don't pretend these are choices for new accounts.

What the rank is NOT based on

• Theory-crafting calculations alone (without in-game checks).
• Community consensus on Reddit / Discord (useful signal, not the call).
• Pre-release tier predictions (we re-test in launch builds).
• Developer marketing copy.

How build pages are produced

• We pick one canonical build per character role. Alternative builds (e.g. supportive vs. selfish DPS) get separate sections on the same page, not separate pages.
• Substat priorities are written, not just bullet-ranked. If “Crit Damage > Crit Rate” only applies above 65% crit, the page states that breakpoint, not just the priority.
• Outdated builds get the same 60-day stale flag as the rest of the site.

Patch cadence

Each patch we run the same loop: read the official notes, verify the data we can verify, add pages for what's new, and revise existing guides where the patch invalidates them. Re-test priority goes to characters explicitly buffed or nerfed, then to anyone touched indirectly (counter element shifted, supporting unit adjusted, etc.). Anything not touched within 60 days surfaces a stale pill on the page until we verify it.

Community contributions

Join our Discord — the best place to help us improve Mongil Codex. Spot a wrong number, want a page that doesn't exist yet, or just want to talk Mongil with other players? That's where it happens.

• Build submissions go through /builds. We don't auto-publish — every community build is reviewed against the current patch before going to the public feed.
• Map routes, character teams, and speedruns: same review gate.
• Submitting requires a Discord-linked account. Speedruns require verifiable video.

Corrections & disputes

The fastest path: tell us in the Holst Network Discord. When a number on a page is wrong we acknowledge the report in the relevant channel, fix the page, and update its “Last reviewed by” date. If the original claim materially misled readers (not a typo — an actually-wrong recommendation), we add a one-line correction note at the bottom of the page so the change stays visible.

For tier-rank disputes specifically: bring receipts. Screenshots, clear times, build context. We'll re-test in-game and either change the rank or write a paragraph explaining why we disagree.