Why is D&D still Medieval?

[Willie the Duck]

"How to Lie with Statistics", a $5 to $10 book that's possibly one of the best non-math explanations of all the ways stats, studies, and survey data can fail, be manipulated, or get misrepresented.
https://www.google.com/search?gs_ssp...PEPvPG8qA821:0

Ah, you've made an assumption based on the title of the book.

I'm not sure what you expected them to do with your posting. Just dropping the link in here without also including an explanation as to what parts of the survey in question you think are non-representative and why is vaguely like wandering into the middle of a discussion and shouting 'Fallacy!' as loud as you can instead of formulating a thought and explaining it to others. Sure, those of us with a data science background can impute* your point from missing data, but for anyone else, you've just dropped a link to a book slightly too expensive to buy just to figure out another poster's point for them (mind you, they could have asked 'and what conclusions did you draw from this book, that you consider pertinent to this discussion?')

Basically all surveys of people will have multiple inherent biases and most surveys have areas of ambiguity that require interpretation. Professionals and experts involved with making and running surveys try to mitigate these flaws and still semi-regularly come to incorrect conclusions. Based on past experience with WotC, it's surveys and results I simply have no faith that they would pay for professional survey and analysis services. So, not saying that anything iffy is going on, just that I think they have shoddy surveys that they analyze based on their own unchallenged assumptions.

The most common source for the statistic in 5e is D&D Beyond data. However, I have yet to see evidence that the D&D Beyond userbase is statistically similar to the general D&D playerbase, so it could potentially be a non-indicative sample. Furthermore, not every PC built in DDB is played, so we have to take into account which characters actually receive playtime;I have no idea how one would accomplish that.

There we go, actual meat to the matter. I tend to agree. WotC's surveys definitely seem to be skewed. Usually with regards to what questions they bother asking. Regarding P. G. Macer's point about D&D Beyond's character's made being non-representative of total gamer populations actual character's played (and at what proportion), I'd also agree. In particular, I suspect that D&D Beyond includes a whole lot of first level characters that are rolled up, but not played (or perhaps printed off and played in person, without updates back to D&DB as to how they progress or how long they are played).