Bob World Builder's 2026 D&D survey: what 5,000 players said, and what it does not show
On 15 September 2026 the YouTube channel Bob World Builder published Player Data Doesn't Lie: D&D is Changing, an 18-minute reading of a fan-run copy of Wizards of the Coast's own player survey, answered by roughly 5,000 people, against a similar copy from 2024 answered by about 9,000. The headline the video draws is a contradiction: respondents like Dungeons and Dragons less than they did two years ago, like its direction more, distrust its publisher by a wide margin, and keep playing it anyway. The numbers are worth reading, and so are the caveats the video itself lays out.
· 5 min read · TerrainTale
What the survey is and who answered it
Wizards of the Coast runs periodic player surveys whose results it does not publish. Bob, the channel's host, copies the questions into an anonymous form each time, shares it, and publishes the answers with a spreadsheet linked from the video. The 2026 form asked the same first question as 2024, how do you feel about Dungeons and Dragons, with answers from I hate this game to I love this game, plus new questions on trust, direction and actual play shows that the 2024 copy did not have.
The sample is the channel's audience, mostly. The video says so more than once, and it says the audience of Sly Flourish, a channel that shared the survey too, moved the same way. The average respondent age was 39. Nothing here is a census of D&D players; it is a large, self-selected group of people who follow D&D on YouTube, measured twice with the same questions.
The feeling about the game fell
The video's first finding is the drop in positive feeling. In 2024 about 75 percent of respondents said they liked or loved the game; in 2026 about 60 percent did. I love this game fell from about 35 to 24 percent, I like this game from about 40 to 34, and I don't like this game rose from 8.9 to 19.1 percent. Outright hate went from 1.6 to 3.9 percent, which the video takes some care to describe as technically doubled while remaining small, a phrase aimed at anyone planning a headline.
Bob's own explanation is the halving of the sample. About 9,000 people took the 2024 copy and about 5,000 the 2026 one, and he reads that as fewer people caring strongly enough to fill in a survey now that the run of big news from 2020 to 2025, the licence dispute of 2023 among it, has passed. That is speculation and he labels it as such.
- Like or love the game: about 75 percent in 2024, about 60 percent in 2026.
- Dislike: 8.9 percent to 19.1 percent.
- Sample: about 9,000 responses in 2024, about 5,000 in 2026.
The feeling about its direction rose
The second finding runs the other way. Asked in 2024 how they felt about the direction of the game, respondents were mostly negative. Asked in 2026 whether they agreed that D&D is headed in the right direction, strong negative feeling fell by about 5 points and plain negative by about 15, with neutral up 5 and positive up 10. On the publisher, strongly negative feeling about the direction of Wizards of the Coast fell from 54 percent to about 37, and neutral tripled from 9.3 to 27.9 percent.
The video connects the improvement to the announcements at Gen Con on 30 and 31 July 2026: the Universes Beyond line of licensed crossovers, starting with D&D: World of Warcraft on 17 November 2026, a Dark Sun setting on 30 January 2027 and a Star Wars book on 15 June 2027, as EN World reported at the time. It also notes that Universes Beyond is the one part of that news some respondents dislike. The two questions were not worded identically in the two years, and the video says the comparison treats them as if they were.
Trust is the number that did not move
The 2026 copy added two agree-or-disagree statements. About 70 percent disagreed that they trust Wizards of the Coast with D&D, and about 80 percent disagreed that Wizards cares about them as a player. Set against 60 percent who like or love the game and, by another question in the same survey, played it in the last month, the video's conclusion is that a bad opinion of the company is sustainable as long as the opinion of the product holds. It puts it as D&D is safe.
That reading lines up with what Wizards is saying itself. On 14 September 2026 Wargamer published an interview in which Dan Ayoub, head of D&D, called the 2023 licence episode a rug pull and said, in his words, trust takes a long time to build and a second to break. The survey suggests the second half of that sentence is still where the audience is. Bob has about 4,000 free-text answers to the question of what Wizards could do to improve, unread as of the video, and promises a separate video on them.
Dimension 20 is close behind Critical Role
The last section is the one the actual play audience will argue about. Asked which D&D shows they watch, respondents chose I don't watch any about as often as Critical Role, which Bob had expected to lead by a wide margin. The larger surprise was Dimension 20 sitting close behind Critical Role, despite most of its seasons being behind the Dropout subscription while Critical Role's roughly 400 episodes are free on YouTube. Bob's guess is that a little more free material would put Dimension 20 first, and he frames it as popularity shifting after about ten years of Critical Role on top.
That is a guess from a survey of a YouTube audience, not a ratings figure. But it matches the week's schedule: Dimension 20's Toylight premiered on Dropout on 16 September 2026, and Critical Role's Campaign 4 continues on Thursdays, so anyone who wants to test the claim at their own table can ask which show the players are quoting.
Does it hold up?
As a measurement of the whole hobby, no, and the video does not claim it is one. As a repeated measurement of one large audience, yes, and the direction of each change is more trustworthy than any single percentage. The drop in affection and the rise in optimism are both large enough to survive a self-selected sample, and the trust numbers are so lopsided that no plausible sampling error rescues them.
The most useful caution in the video is its own: the same true fact, hate doubled, can be an honest sentence and a misleading one. Anyone quoting these figures should carry the sample sizes with them. The spreadsheet is public, linked under the video, and the questions are worth putting to your own group before the next session.