On the Popularity of Games

Which Role Playing Game is the best role playing Game?

Wait, no.

Which is the most popular?

The truth is, we live in relative darkness about what games are being sold and what games are being played. To date, our best resource has been ICv2 which just tells us the top few best selling games through the retail channels.

This has allowed an awful lot of us to live in ignorance.  Didn't release a supplement this month? Nobody is playing your game!

What's more, is that the people in charge of the companies have no motivation to dispel any illusions the players have.

And it's not really a mud-slinging type industry—After all, products that focus on "fixing" things and are built around not having "problems" other gaming systems have don't do well. Or so I believe, and who's to tell me otherwise?

The Orr Group, that's who.

Who are the Orr Group? They are the Roll20 programmers who consumed Tabletop Forge when several people quit the project. (A slight exaggeration, Joshua Owen approached Roll20.) They provide a Virtual Table Top interface for people playing on-line games.

Now people don't like to have their illusions shattered. Anyone who doesn't like these numbers or facts will just come up with excuses about how it doesn't count home games, or people playing games on G+ online might not be using these pieces of software.

Well that's true. it's also true that those people don't understand statistics and sampling data either. After all, I play B/X so I don't need Roll20, so I'm not counted.

Let's look at the numbers we get for the third quarter of 2014.

THE ORR GROUP INDUSTRY REPORT is an overview of tabletop gaming ruleset popularity based on usage data from Roll20 Virtual Tabletop. Third Quarter results are as follows: Based on a sample size of approximately 25,000 games and 15,000 players.
Notes on how information is gathered:
  • “Games” = % of games where the Game Master said “this is what we’re playing.”
  • “Players” = % of players who were active in the quarter and filled in the “this is what I enjoy playing” field on their profile.
  • Percentages will total more than 100% because each player/game can have more than one game type designated. This is extremely common for players and rare for the games.
  • This only takes into account games/players which were active (e.g. game was played, player played at least one game) during the 3rd Quarter of 2014.
  • This is meant to be a representative sample, as it only takes into account games and players who filled out relevant fields on their game page/profile. There was significantly more than 25k games and 15k players who played in this quarter, but many do not fill out these fields.
  • The list of available games to select were curated by Roll20’s staff based on previous freeform survey results, and will be changing in the future.
Notes on Roll20 / THE ORR GROUP:
Roll20 began as an effort to keep developers and The Orr Group founders Riley Dutton, Nolan T. Jones, and Richard Zayas in touch via long distance gaming. Since launching via Kickstarter in April of 2012, Roll20 has attracted more than 600,000 users as a free service. The program continues to be funded by subscribers who receive features that assist advanced gameplay.

Did it match your expectations? Was the Popularity of the games played related to your internal narrative of which games are popular and which games are not?

Does it matter if more people are playing 1st edition than whatever game you are playing that isn't Dungeons and Dragons?

Hack & Slash 
FollowGoogle +NewsletterSupport


  1. Seems odd to me that RuneQuest and variants (except call of C'thulu) are missing entirely.

  2. I have only used Roll20 for RQ, I know I am an oddball, but I'd have thought there would have been a few more.Mind you this is skewed towards systems that benefit from a grid.

  3. minor nitpick: doesn't Josh Owen spell his first name Joshuha?

  4. I'm honestly surprised at the results. When you look at the games played Pathfinder wins, but the most people are playing D&D 3.5. Then there's the fact that the three latest editions of D&D account for a combined 41.26% of all the games played on Roll20 while 86.49% of all players are participating in sessions involving those games. If you add AD&D into the D&D category then you've got 43.91% of all games!

    I expected the combined D&D total to closer to 30% and that the highest percentage of players would be playing Pathfinder, not D&D 3.5. Do you know if they're going to be doing these again?

  5. " Anyone who doesn't like these numbers or facts will just come up with excuses about how it doesn't count home games, or people playing games on G+ online might not be using these pieces of software.

    Well that's true. it's also true that those people don't understand statistics and sampling data either. "

    This isn't at all true. I can easily imagine a situation where players of certain games are more likely to want to play via computer. I.e., there might be a sampling bias.

  6. Using Roll-20's usage data is actually a limited sampling source. In recent U.S. elections, many polling agencies assessments of eloctorates proved wrong because they were relying on Land-Line telephone polling, while ignoring that many young people didn't possess land-lines, and that many working poor were rarely home to take part in such polling.

    Similarly, it hardly comes as a surprise that when polling users of a single service, the results will skew towards the services that have made efforts (or had efforts made) to make their product compatible with the service. It makes sense that some games are less suitable to online play. Much like all user polls, the results reveal much about the audience of a product, but are often a poor representation of the entire population. The Roll-20 data is useful in revealing how their service is used, but extrapolating that data across _all_ gaming is not statistically sound.

  7. Do you have a link to the original report?

    1. It was sent by e-mail. The quoted text is the unedited text of the press release, and the graphic is the one that was sent with it.

  8. Wow, the sample size is 25,000?! How many games are running on roll20 I wonder?

    If these numbers are from everyone who has filled out a profile that means that the percentages emphasize people who use roll20 to find games/players. I'm one of the 0.75% who run an OSR game and I have a profile that says as much, but my players don't have profiles.

    The numbers don't surprise me, Pathfinder is the big fish and roll20 is tailor made for Pathfinder/3.5. This info doesn't bode well for 5e because not that many players are using roll20 to find games...

    The high percentage of 4e games & players is interesting. I think the VTT environment is ideal for miniatures focused rules, because you can play the same game without buying all the miniatures. I want to get some Mordheim going on roll20!

    Also, you do didn't have to out yourself by including BESM in the picture like that, no one had to know :)

  9. I'm confused by some of the "games" listed. It makes sense to break out different versions of a game (like D&D), but what about the White Wolf stuff? They list World of Darkness, Vampire: The Masquerade, New World of Darkness, and White Wolf (Any Game). That would be like listing D&D 3.5, D&D 4, Forgotten Realms, and D&D (Any Game).

    I'm sure that these things are artifacts from the profiles that didn't translate well into the expected data, but I think they have a pretty big impact on conclusions drawn. "Dirty data" is always problematic and can drastically skew results. And since they're using non-standardized responses, there's no telling what you're going to get.

    Aside from D&D/Pathfinder (who are the clear leaders), any small change in definition or grouping is going to have a big difference. Grouping all editions of Star Wars, while breaking apart VtM from the rest of WoD is going to lead to very different conclusions than if you included all WoD games together or broke Star Wars into D20, D6, and Saga. Also, I can't be sure if D&D 3.5 includes other D20 games, like Modern.

  10. Nice info. It is a pity that they only took it from survey data though. Surely there is a way for them to see what campaigns have been active, how many have been active in them and what the campaign lists as the gaming system.

    It also ignores (of course and I feel stupid putting this here) games being played at an actual table and also games that are not really suited for Roll20. For example diceless games and I can almost guarantee there are a lot more FATE games going on too but Roll20 is not really a FATE thing. I know it is represented and can be done but it would not be my first choice to play the system online.

  11. Sampling bias. Roll20 is most attractive to people who play games on a grid (PF, 4e), so we can expect those games to be over-represented here. Also, I expect a lot of grognards would shy away from Roll20, regardless of system, and would be under-represented.

    Interesting data, though. I'm surprised 5e doesn't have higher numbers.


Related Posts Plugin for WordPress, Blogger...