Do Tabletop RPG Sites Block AI Crawlers? 2 of 6 Do

Tabletop RPG is the rare hobby vertical that lands slightly above the corpus average for AI blocking — but the more striking detail is how few of its sites publish a robots.txt at all. Of the 10 sites we track, only 6 returned a parseable policy, and 2 of those gate at least one AI crawler.

2 of 6 Tabletop RPG sites block at least one AI crawler.

That works out to a 33.3% block rate, a touch above the corpus line. Across the 934 sites in this edition with a readable robots.txt, 277 block at least one AI crawler — a 29.7% rate. Tabletop RPG edges past it. What makes this slice distinctive is the missing-policy share: four of the biggest marketplace and community sites publish no readable robots.txt at all. This report names the blockers, the allowers, and the no-policy sites, with every figure a verbatim count from public robots.txt files sealed on June 14, 2026.

Who Gets Disallowed Here

Of the 10 tabletop RPG sites we checked, 6 returned a parseable robots.txt. Two of those — enworld.org and koboldpress.com — disallow at least one AI crawler token. A community forum and a publisher: the kind of properties most invested in protecting original written content from training sets.

The allowers are dndbeyond.com, roll20.net, tribality.com, and thealexandrian.net — a mix of the dominant digital toolset, a virtual tabletop, and two long-running content blogs that leave access open. Then there is the distinctive group: dmsguild.com, drivethrurpg.com, gnomestew.com, and rpg.net returned no parseable robots.txt at all. That is not a block; it simply means there is no published rule to read for some of the hobby's largest PDF marketplaces and forums.

Two of the 6 readable tabletop RPG policies — enworld.org and koboldpress.com — disallow an AI crawler.

The low policy-publishing share is the story under the story. For a hobby this text-heavy, with so much original rules and fiction content, you might expect more sites to weigh in on AI access. Instead, four of the biggest just have no published stance. For a hobby a step lower on the ranking with a fuller set of readable policies, the disc golf AI-access report makes a useful contrast.

The identity of the no-policy sites sharpens the point. dmsguild.com and drivethrurpg.com are the dominant PDF marketplaces for the hobby, carrying enormous catalogs of original rulebooks and adventures — precisely the content a publisher might be expected to fence off. That they publish no robots.txt at all is striking: it means the most license-sensitive content in the vertical currently sits behind no stated AI-access rule whatsoever. rpg.net and gnomestew.com, a major forum and a long-running advice blog, round out the silent group with deep archives of community discussion.

Whether that silence reflects a deliberate openness or simply an unmade decision is something the snapshot cannot tell us, and we do not guess. What it does establish is that the readable slice — the 6 sites with a policy — is the honest basis for the block rate, and that the four absent policies are a live question rather than a settled allow.

Why Tabletop RPG Lands Where It Does

A 33.3% block rate puts tabletop RPG just above the corpus average — unusual for an enthusiast vertical, most of which sit well below it. The reason is the nature of the content. Game rules, adventure modules, and homebrew fiction are exactly the kind of original written work whose creators have the strongest reason to keep it out of model training. So the two sites that do publish a policy and care about gating, gate.

Tabletop RPG sites post a 33.3% AI-crawler block rate.

At the same time, the platform layer — the toolsets and virtual tabletops where play actually happens — leans open, because their value is utility and reach, not the licensing of static text. That split between content publishers gating and platforms staying open is what pushes the readable slice just past the corpus line.

dndbeyond.com and roll20.net illustrate the open platform logic well. Both are interactive tools — a character-and-rules manager and a virtual tabletop — whose product is the experience of play, not a corpus of static prose, so there is little for an AI crawler to extract that competes with their business. tribality.com and thealexandrian.net are independent content blogs that have chosen reach over restriction, the opposite call from enworld.org and koboldpress.com despite working in the same medium.

That two text publishers gate while two others stay open is the clearest sign that this is an unsettled question inside the hobby rather than a consensus.

Where This Sits in the Corpus

The window below centers on tabletop RPG and its nearest neighbors in the block-rate ranking — the categories directly around it — not the full 112-category list. It shows the band this hobby occupies.

The company is telling. Tabletop RPG shares its 33.3% reading with ham radio, wine, and motorcycles, and sits just above yoga, scuba, and beekeeping. These are mostly content-rich enthusiast verticals where a meaningful minority of sites gate.

Against the extremes, tabletop RPG sits in the busy middle of the ranking, above the corpus average but nowhere near the news-and-gaming cluster. A fully-open sibling sport with no blockers is covered in the pickleball AI-access report.

Which Bots Are Blocked Most Across All 934 Sites

When a tabletop RPG site does gate, the tokens it reaches for are the leaders of the corpus-wide bot leaderboard. Across all 934 sites with a readable policy, blocking concentrates on a familiar top tier.

Across all 934 sites, CCBot is the most-disallowed crawler token, named in 204 published policies.

This corpus-wide picture frames the tabletop numbers. enworld.org and koboldpress.com are not inventing a blocking pattern; they are opting into the same dominant bot tokens that the rest of the web's content publishers reach for first. A different enthusiast vertical that gates only a single site appears in the reef keeping AI-access report.

Corpus-wide, 277 of 934 sites block at least one AI crawler.

Reading the Sealed Numbers

Every figure comes from one sealed, point-in-time snapshot. We fetched each tabletop RPG site's public robots.txt, parsed its user-agent and disallow directives, and recorded whether any AI crawler token was disallowed at the root. A site that returns no robots.txt is logged as having no published policy — not a block and not an explicit allow. In this research, nothing is estimated, modeled, or extrapolated; each count is read directly from the files as they stood when the snapshot closed.

The snapshot is content-addressed so the numbers stay reproducible:

1. Collect. Fetch each site's /robots.txt and keep the raw response verbatim.

2. Parse. Extract user-agent groups and disallow rules, matching the AI-crawler token list.

3. Seal. Hash the full snapshot and freeze it under sha 760275d49a628cc3.

4. Aggregate. Compute category and corpus totals only from the sealed records.

Because robots.txt is voluntary, these counts measure stated intent, not enforcement. A disallow line is a request a crawler may or may not honor. For tabletop RPG in particular, the gap between the sites we checked and the sites with a readable policy is unusually wide, so we are careful to report the block rate against the readable slice rather than the full set we attempted — anything else would either bury the no-policy sites or silently recast their silence as a choice they never published.

Frequently Asked Questions

Q: Which tabletop RPG sites block AI crawlers?

A: enworld.org and koboldpress.com are the two of 6 readable tabletop RPG policies that disallow at least one AI crawler token — a community forum and a publisher. The other readable policies, including dndbeyond.com and roll20.net, leave access open.

Q: Why do four of the biggest tabletop sites have no robots.txt?

A: dmsguild.com, drivethrurpg.com, gnomestew.com, and rpg.net returned no parseable robots.txt in this snapshot. That means they publish no rule for a crawler to read — neither a recorded block nor an explicit allow — which is why only 6 of the 10 sites we checked have a readable policy at all.

Q: Why does tabletop RPG block more than most hobbies?

A: Its content is original written work — rules, adventures, and homebrew fiction — which gives publishers the strongest reason to keep it out of training sets. That pushes the readable slice to a 33.3% block rate, just above the 29.7% corpus average, while the platform layer like roll20.net stays open.

Q: Does a disallow rule in robots.txt actually stop a crawler?

A: No. robots.txt is an honor-system standard, so a disallow directive states a preference: a well-behaved crawler honors it and a non-compliant one can ignore it. The figures here capture what enworld.org, koboldpress.com, and their peers publish, not what any crawler is forced to do.

Key Takeaways

Tabletop RPG edges just above the corpus average for AI blocking, driven by content publishers gating original text — but its most distinctive trait is how many of its largest sites publish no policy at all.

• 2 of 6 tabletop RPG sites with a robots.txt block at least one AI crawler — a 33.3% rate.

• enworld.org and koboldpress.com are the named blockers; dndbeyond.com and roll20.net stay open.

• Four major sites, including drivethrurpg.com, returned no parseable robots.txt at all.

Put AI-Access Data to Work

The realistic buyer of this monitoring is horizontal. An AI-search and GEO agency tracking client-eligible corpora can fold tabletop RPG into a weekly sweep: re-crawl the 6 readable policies and watch the 4 no-policy sites, then alert the account team the moment drivethrurpg.com or dmsguild.com publishes a robots.txt that disallows a crawler, or an open site like dndbeyond.com adds an AI-crawler token — both signal a client's content shifting out of answer-engine reach. A competitive-intelligence analyst watching AI-access drift across many content-heavy verticals can run the same recurring job to catch the first publishers in a category to gate.

The category-native second buyer is a TTRPG rulebook and PDF marketplace product manager who wants storefront and SRD pages eligible for AI discovery; they would monitor their own robots.txt against this baseline so a stray rule never silently delists them while koboldpress.com and enworld.org gate by choice. US Tech Automations runs this scheduled robots.txt and llms.txt monitoring with change alerts and an AI-access dashboard. See how the agentic monitoring workflows detect policy drift automatically.

Source: US Tech Automations Research — Closing Web edition; figures are verbatim counts from public robots.txt files sealed June 14, 2026 (snapshot sha 760275d49a628cc3).