Most gaming communities that ask about Discourse are not asking a feature question. They are asking whether they can keep the room they already run — an 80-node tree, a whitelist application, a live server list, a credits shop, a last-post column veterans scan at 23:00, and a Discord that already owns voice — after a CSS upgrade. The honest answer is usually no, and that is not an insult to Discourse. It is a product-shape mismatch.
This is a fit test, not a scoreboard. The 2026 matrix already compared nodes to categories and tags. The ops close already said the gaming add-on ecosystem is a XenForo pro. The XenForo RP/gaming architecture piece already hung whitelist, listings, fields, Discord, and ranks on a node tree. This article will not rebuild that cupboard. It asks: given a game or RP server community, when does Discourse actually fit, and what should you expect to abandon?
A research note, because it changes how hard you should take any war story: the Discourse notebooks we work from did not include gaming case studies. What follows is reasoned from product shape — official install and trust-level behaviour, chat as a plugin, categories-and-tags IA, the XenForo add-on culture this site already documented — not from a Meta thread titled “we moved our FiveM board.” If you have a public post-mortem, prefer it to this. If you do not, do not invent one.
The fit test (write the answers down)
Five filters. If four say XenForo, stay. If four say Discourse, continue. If you are 3–2, run a pilot category of work (a public dev-blog, a rules FAQ) on Discourse and keep the game board where it is.
- Is the product a tree of inhabited rooms, or a topic stream with tags?
- Which three extras are load-bearing: whitelist, server list, credits/shop, character sheets, faction ACL, resource library?
- Where does voice and LFG actually happen — forum, Discord, both?
- Will veterans accept Latest / infinite scroll, or is the last-post column the culture?
- Who can staff Docker or a hosted invoice, and who can only staff a PHP box?
Filter 2 is the veto more often than people admit. Discourse can host a discussion about a game. It is a poor lateral move for a discussion that is an application form plus a directory plus an economy.
What “a gaming community” usually is
Say “gaming forum” and you may mean any of these. They need different software.
| Shape | What members do all week | Software they are actually asking for |
|---|---|---|
| Patch-notes + social | React to patches, share clips, argue balance | A modern discussion site. Discourse is in the running. |
| Support + bug tracker for a game/mod | Searchable answers, solved tags, version tags | Discourse is a good default. Solved + tags are the product. |
| Server / RP hub | Apply, get whitelisted, pick a character, find a server, read lore, appeal a ban | A forum plus extras. XenForo is the default for a reason. |
| Clan / guild board | Private officer rooms, public recruitment, event planning | Either product, if the ACL is simple. XenForo if the ACL is a faction tree. |
| Marketplace / credits economy | Buy tags, donate for a shop item, spend points | XenForo user upgrades and credits culture. Discourse is a rebuild. |
| Discord-first club | Voice all evening, forum as archive | You may not need a new forum at all. If you do, Discourse Chat will not replace Discord. |
Most “should we move to Discourse” threads from game admins are row three wearing row one’s screenshots. Latest looks clean. The whitelist does not live in Latest.
Product shape: Discourse will not become a board index
Discourse’s default home is Latest. Categories are few. Tags carry the cross-cutting work (version, region, spoiler, platform). There is no last-post column as a first-class index. Infinite scroll is the reading model. Trust levels sandbox new accounts. Official production is Docker or official hosted, not a PHP shared host.
A classic XenForo game board’s default home is a node tree. Officers scan which room had the last reply. Depth is how factions, regions, and “this week’s event” get a door. New accounts can usually post once they confirm, unless you built a validating group. The extras — whitelist applications, server listings, Discord login and role sync, custom character fields — assume that tree and those groups.
You can theme Discourse. You cannot honestly recreate “eighty inhabited rooms plus a last-post scan” without fighting the product every week. People who say “we will just make categories for every old forum” are describing the failure mode the ops close already named: you picked Discourse’s CSS, not its IA.
SF-2 community practice on XenForo is useful here even if you leave: broad nodes plus prefixes beat deep trees. Empty rooms are SEO waste and officer waste. If your 80-node tree is already a ghost town, Discourse’s flatness is not the enemy. The ghost town is. Flattening on XenForo (fewer forums, prefixes for region/status) is cheaper than a platform change. Flattening as part of a Discourse move is honest only if you were going to flatten anyway.
Chat plugin versus Discord
Gaming communities already have Discord. The question is never “should we discover chat.” It is “what is the forum still for.”
Discourse Chat is an in-app plugin. It is useful for people who live in Latest and want a side channel without opening another client. Trust levels throttle it (TL0 defaults include a 20 messages / 30 seconds cap; TL1 loosens). It is not a voice mesh, not a stage channel, not a fifteen-server permission tree, and not where your raid group already idles.
Discord already owns voice, ephemeral LFG, and the “we are online right now” social graph. XenForo’s job next to Discord is usually login, role sync, and announcements — the XF Discord article is explicit: pick one job. Discourse’s comparable official-adjacent story is outbound chat-integration (notifications into Discord, Slack, Telegram, and friends) plus Discourse Chat in-app plus Discord as a login on hosted. That is not two-way guild sync. Do not plan a move on the hope that Discourse will absorb the Discord.
A healthy split that works on either forum product:
| Happens on the forum | Happens on Discord |
|---|---|
| Rules, lore, patch essays, appeals that must be searchable | Voice, LFG pings, “who is on tonight” |
| Whitelist / application if you still have a form product | “You passed, here is the role” ping |
| Ban archive, patch notes, guides | Officer chatter that would junk a category |
| Long arguments you want to cite next season | Arguments you want to die at 2 a.m. |
If 90% of weeknight life is the right-hand column, a forum redesign will not save activity. If 90% is the left-hand column and Discord is a noisy side channel, Discourse can be a better left-hand column than a neglected 80-node XF — provided the extras in the next section are not load-bearing.
What fails (or becomes a rebuild)
Be clinical. These are the usual load-bearing pieces of a game/RP hub, and Discourse’s stock/plugin catalogue is not the XenForo third-party catalogue.
Infinite scroll and the missing last-post column
Veterans who “check the board” are often checking which room moved. Latest tells them which topic moved. Those are different mental models. Officers who moderate by scanning a tree will feel blind. You can teach them filters (watching a category, using the New / Unread lists). You cannot give them the old muscle memory.
If your staff’s only moderation UI is the last-post column, budget retraining or stay. This is cultural, not a theme-component away.
The 80-node tree
Discourse will flatten you. The XenForo → Discourse importer literally flattens node depth greater than 2 into tags. Even on a greenfield Discourse you should have a handful of categories (General, LFG, Support, Lore, Staff) and tags for region, faction, spoiler, patch.
That can be better than eighty empty rooms. It is still a retraining project. Faction officers who “owned a forum” now own a tag and a notification level. Some will treat that as a demotion. They are not entirely wrong: a category has permission walls; a tag does not, unless you work at it.
A worked IA for a discussion-first game community on Discourse (not a hub with applications):
| Category | Tags (required where noted) | Who posts |
|---|---|---|
| Announcements | patch, maintenance |
Staff |
| General | meme optional, spoiler tag if you are that kind of game |
Everyone past TL0 limits |
| Help / bugs | Required version or platform |
Everyone |
| LFG | Required region or mode | Everyone |
| Lore / stories | spoiler | Everyone |
| Staff | — | Staff group |
Five public rooms. The old “EU / NA / OC / PvE / PvP / Weekend / Serious” grid is tags. If that table feels like it is missing your product (applications, directory, shop), that is the signal.
Whitelist / application boards
XenForo can run a stock-ish application workflow: a forum, prefixes or fields, groups that see pending versus accepted, a promotion when staff click. That pattern is a reason game boards stay.
Discourse can store an application as a topic in a restricted category. Staff can use official Assign (when you have it) and a tag (pending / accepted / denied). That is a process, not a product. You will write the process. You will not install a XenForo-shaped application add-on and go home.
If whitelist is how people enter the Minecraft server, not how they enter a chat room, a topic-plus-tag workflow can work. If whitelist is a 12-field character sheet with staff-only columns and a queue UI, you are looking at a custom plugin or you are staying on XenForo.
Server listings
A FiveM / Minecraft / Discord-room directory is a first-class XenForo pattern (often an add-on plus custom fields plus a widget). Discourse does not have an official “live server ping grid.” You can:
- Keep the directory on a small XenForo or a dedicated site, and use Discourse for discussion.
- Fake a directory with a category of topics and a tag per region, updated by humans. It will rot.
- Budget a plugin. Name the maintainer.
“We will just make a category called Servers” is how you get a graveyard of last-year’s IP addresses.
Credits, shops, donor perks
XenForo’s points / credits / shop culture is user upgrades plus third-party economies. Discourse hosted lists gamification as a Business-plan extra; that is badges and scores, not a donor shop with SKUs. A Discourse group can be a donor lounge. A Discourse plugin can be anything a Ruby developer ships. The XenForo resource directory already has the shop you were about to buy.
If the shop is load-bearing, Discourse is a product abandonment. Call it that in the staff meeting.
Character sheets, factions, deep ACL
Custom user fields exist on Discourse. They are not the same diagnostic as XenForo’s fields-plus-promotions-plus-Analyze permissions. Faction trees, regional officers, “this wing of lore is visible to accepted characters only” is the class of ACL the ops close already warned you about. Discourse category security via groups is enough for staff / members / public / one donor lounge. It is a grind for twelve factions with overlapping officers.
RP communities that are mostly searchable play-by-post in a few rooms can live on Discourse (tags for canon vs. sidebar, a lore category, TL0 sandbox against drive-by trolls). RP communities that are a character database plus twenty private boards are XenForo-shaped. The architecture article is that cupboard; do not expect Discourse to grow it.
Ranks that look like a military
Discourse trust levels are not ranks. TL3 Regulars demote on a 100-day rolling window. TL4 is staff/manual or SSO. The moderation model owns the details. Game boards that sold “Sergeant / Lieutenant / General” as identity will have a cultural fight. You can add a gamification plugin (hosted Business lists one) or flair via groups. You cannot make TL3 into a trophy without lying.
Media libraries and resource dumps
XenForo Media Gallery and Resource Manager are official extras. The Discourse importer does not bring them across. A game board whose value is 50,000 screenshots and a mod archive is not a Discourse migration; it is a library project plus a discussion site.
When Discourse is suitable
The fit is real for a subset. Say yes when most of this is true:
- The community is discussion-first: patches, guides, bugs, clips, arguments. Not a directory, not a shop, not a 12-stage application.
- You want a short category list and you were already ashamed of the empty tree.
- TL0 sandboxing would have saved you from the last three raid-week spam waves.
- Staff are community people who will learn Latest, not ACP stylists who live in widgets.
- Discord stays Discord. The forum becomes the searchable half.
- You will pay official hosted (public 2026 grid: Free / Pro around $100/mo / Business around $500/mo — re-check the live page) or you already like Docker on 2 GB+ RAM.
- You can name the three extras you are abandoning, in writing, to the members who paid for them.
- Greenfield — a new game, a new server, no 2014 bookmarks — is the easy mode.
Typical good fits: a studio’s official discussion site; a popular mod’s support room; a game with a patch cadence and a wiki-adjacent culture; a Discord-first club that wants a public, indexed FAQ and patch archive.
Typical bad fits: FiveM / Rust / Minecraft networks whose forum is the whitelist and the server list; large RP sandboxes with character sheets and faction boards; any board whose homepage is a live directory.
Chatty versus archival: a week in each chair
Monday on a XenForo game hub. You glance at the node tree. The whitelist forum has twelve new prefixes. The EU listing widget is stale; you ping the server owner. A donor opened a ticket because a promotion did not add them to the VIP node. Discord announced the patch; the news forum still needs the thread. Officers ask why the last-post column on “Lore / Deep / Region 4” is empty. It is empty because nobody has inhabited that room in a year. You do not delete it because a lore writer might come back.
Monday on a Discourse game discussion site. You open Latest. A TL0 user cannot embed four screenshots of a bug; you point at the welcome topic. A Regular flagged a phishing “free nitro” post and the sandbox plus watched words already hid the last one. LFG is a category with a required region tag. There is no listing widget. There is no VIP node. Discord still owns voice. Someone asks where the last-post column went. You do not have one. The lore writer’s old deep-tree room is a tag on Lore, or it is a topic.
Neither Monday is morally better. One is a hub operator’s Monday. The other is a discussion operator’s Monday. If your members bought the hub, do not surprise them with the discussion Monday.
Hybrid patterns that are not cowardice
You do not have to move the whole city.
Discourse for public docs + XF for the hub. Patch notes, bug tracker, and “how to install” on Discourse (or even on a studio Discourse). Whitelist, listings, and donor shop stay on XenForo. Two logins unless you invest in SSO on both sides. Two staff cultures. Worth it when the public SEO surface is the painful part of XF and the hub is the valuable part.
Keep XF, flatten the tree. Broad rooms, prefixes, delete empty nodes. This is the SF-2 rule and it is cheaper than Docker. If the complaint was “the board feels old,” fix the style and the tree before you change the engine.
Keep Discord, do not add a second chat. If you move to Discourse, turn on Chat only if people will leave Discord. Two chat products is how LFG fragments.
Pilot a single game mode. New season, new server, Discourse-only. Old servers stay on XF until they die. You learn whether your officers will flag and tag before you burn the permalinks.
The stay-or-migrate framework still applies: lock-in, staff, URLs, cost, size. Game extras are just lock-in with louder users.
Cost and hosting, gaming-sized
A hobby XenForo game board: license $195 (12 months updates; $60/year to extend), plus a PHP VPS, plus the add-ons you actually use (Resource Manager $70, Media Gallery $70, Enhanced Search $60, branding removal $350 if you care — public 2026 purchase page; VAT extra). Cloud exists (Starter $60/mo listed, up through Business $250/mo) if you want no SSH.
A hobby Discourse: software $0, VPS that is happy at 2 GB+ RAM, or hosted Free / ~$100 / ~$500. Rebuilds take the site down for minutes. Game communities that already run game servers often can run Docker. Game communities that live on a $6 cPanel box often cannot.
Staff time dominates both invoices. Retraining officers who hate infinite scroll is a real line. So is rebuilding a listing plugin.
Decision table
| If this is true | Do this |
|---|---|
| Whitelist + listings + shop are the product | Stay on XenForo. Read the architecture, whitelist, and listings pieces. |
| Officers live in the last-post column | Stay, or budget a quarter of retraining and accept resignations. |
| You need twelve faction ACLs | Stay. Discourse groups will make you sad. |
| You are a studio / mod team and the forum is support + patch notes | Discourse is a fit. Hosted if you can pay. Tags for version. |
| You are Discord-first and want a searchable archive | Discourse or a flat XF. Do not rebuild voice. |
| You hate the 80-node tree but love the extras | Flatten on XenForo. Do not change engines to avoid a cleanup. |
| You already decided to move | Use the import runbook. XFRM/XFMG/page nodes die. Passwords: plan a reset. |
| You want “modern” and have no extras | Discourse or even Flarum if you must stay on PHP. |
What to tell members if you move anyway
Do not lead with “better software.” Lead with the losses.
We are moving discussion to a new platform. Latest replaces the board index. Your old forum rooms become a few categories plus tags. Applications / listings / the shop are not coming across unless we name a replacement in this post. You will set a new password. Discord stays where it is. Staff will run office hours for two weeks.
If you cannot publish that paragraph because a sentence is still “TBD,” you are not ready to move.
Related reading
- XenForo cupboard for RP/game hubs: build a roleplay or gaming forum
- Application workflow: whitelist requests
- Directory: server listings
- Voice and roles: Discord and XenForo
- After you still want Discourse: import runbook, trust levels, first week
Takeaways
- Discourse is suitable for discussion-first game communities (patches, support, guides) and a poor lateral move for hubs whose product is whitelist + listings + shop + faction ACL.
- We had no NotebookLM gaming case studies. Reason from product shape, not from a fake migration diary.
- Discourse Chat is not Discord. Keep voice and ephemeral LFG where they already are.
- Latest and tags replace the 80-node tree and the last-post column. Flattening can be healthy. Pretending you will keep eighty categories is how you hate the software.
- Load-bearing XF extras (whitelist, listings, credits, XFMG/XFRM) die or become rebuilds. Many game boards should stay.
- Hybrid is allowed: Discourse for public docs, XenForo for the hub; or flatten on XenForo and do not migrate.
If you wrote “discussion-first, Discord stays, extras named and killed, staff will learn Latest” and it did not feel like a eulogy, Discourse can host your game community. If it did, you already have your answer. Stay on XenForo and spend the quarter deleting empty nodes.