Change in preference for D&D edition

This was done as a check on survey responses for D&D Research Wizards to see which of two survey blocks was more representative. Gathered in September and November 2020, it seems odd that the editions preference should be so different.

Short answer - more recent block B is more representative. First block A had a strangely high group of 3.5e respondents within it compared to the understanding of the general player population. Using Roll20 player data as reported by Orr Reports we can see the shift in preference for different editions over time. This also happened during a large increase in players on Roll20 so we need to verify this.

It appears that over time 5e has absorbed players from other editions. Initially this seemed to come from 4e particularly where player numbers on Roll20 shrank despite players on the platform growing.

To reality check this we can compare the Roll20 numbers against contemporary surveys where available (2014, 2019, 2020) to confirm this shift in player preference looks coherent. Pathfinder became less played, 3.5e and 4e shrank considerably. This does not support that a representative group would show as many 3.5e players as a fraction of the group today as back in 2014.

The second block of respondents appears to be broadly in line with expectations. It would be interesting to see a breakdown of sources for the first survey - by site, in person interview, etc. to see if there was a particular bastion of 3.5e players that was tapped. As a 3.5e player myself, I would love to know where this hidden realm of like minded souls is.

As a side note, the comparison to Roll20 and twitter surveys suggests there is a block of Pathfinder players out there who have not been contacted by this effort to date.

Sources:

Name	Year	Size	Type
Orr Reports	Multiple		Site data
D&D Research Wizards Update #1	2020	1006	Survey
Twitter D&D Demographic Survey	2020	4021	Survey
Trygstad Thesis	2019	1130	Thesis
Reddit D&D	2014	1734	Survey