Fantasy Grounds Toolset Usage Report
  • Fantasy Grounds Toolset Usage Report


    Fantasy Grounds reports the latest RPG system usage! Most games are up a few percent, and it looks like D&D 5E is remaining steady, and Pathfinder 2 has started to make an appearance in the charts. The following charts show usage in the last 12 months (PF2, of course, has only been available for a few weeks).





    Ruleset Total Games
    5E 503,696
    PFRPG 83,268
    SavageWorlds 36,040
    3.5E 18,967
    SFRPG 15,395
    CoreRPG 13,279
    MoreCore 11,782
    Call Of Cthulhu 11,704
    4E 9,274
    Star Wars EotE 8,504
    GURPS 5,981
    AD&D 5,032
    Castles and Crusades 5,030
    DSA 3,676
    PFRPG2 3,398
    RolemasterClassic 2,939
    WFRP 2,858
    Fate Core 2,441
    WH40K 2,332
    VtM 1,987
    Numenera 1,792
    Shadowrun 1,048
    Basic Roleplaying 843
    Mongoose Traveller 1E 771
    WHFRP 2ed 759
    Grand Total 752,796


    Comments 26 Comments
    1. Warren LaFrance's Avatar
      Warren LaFrance -
      With the current spat roll20 is having with a few youtube and twitch broadcasters I wonder if there has been an increase in FG user base...?
    1. Wrathamon -
      surprised by PF2 having a decrease but og pf having an increase
    1. Kite474's Avatar
      Kite474 -
      Quote Originally Posted by Wrathamon View Post
      surprised by PF2 having a decrease but og pf having an increase
      Makes sense. The honeymoon period of the shiny new playtest is done so all that's left (or will be left) is folks who are in earnest trying to test everything out
    1. Celebrim's Avatar
      Celebrim -
      Is SFRPG really that popular?
    1. Wrathamon -
      I know more people who play Starfinder than pathfinder ... like a lot more. Those who dont play 5e seemed to have gone to play SF, and not go back to PF.

      but, that is just my pool
    1. Jay Verkuilen's Avatar
      Jay Verkuilen -
      Once again here's a Zipf distribution plot showing log(Count) plotted on log(Rank). Natural logs, naturally---the only civilized logarithm.

      Name: zipfplot.jpg ► Views: 1395 ► Size: 305.9 KB

      While the Zipf distribution doesn't fit when tested against the more general Mandelbrodt distribution, it's a pretty good first order approximation except in the tail. I would not expect it to fit perfectly due to the fact that the conditions of either distribution are violated due to lack of independence.
    1. Morrus's Avatar
      Morrus -
      Quote Originally Posted by Jay Verkuilen View Post
      Once again here's a Zipf distribution plot showing log(Count) plotted on log(Rank). Natural logs, naturally---the only civilized logarithm.
      For those of us who don’t have a mathematical background (ie me)could you explain what that is?
    1. Mercule's Avatar
      Mercule -
      Quote Originally Posted by Jay Verkuilen View Post
      Natural logs, naturally---the only civilized logarithm.
      I don't know. Wouldn't a domesticated log be more civilized? I mean, natural logs are just kind of feral.
    1. Jay Verkuilen's Avatar
      Jay Verkuilen -
      Quote Originally Posted by Morrus View Post
      For those of us who don’t have a mathematical background (ie me)could you explain what that is?
      The data have counts of games running with a particular ruleset. So 5E is the most common ruleset, followed by Pathfinder, followed by Savage Worlds, etc. The kind of mechanism that generates this pattern of counts is often thought of as representing "preferential attachment,"

      A preferential attachment process is any of a class of processes in which some quantity, typically some form of wealth or credit, is distributed among a number of individuals or objects according to how much they already have, so that those who are already wealthy receive more than those who are not.
      This means a more popular ruleset has a higher chance of drawing a new entrant, with the probability being related to 1/rank. The Zipf distribution, and its relatives like the Mandelbrodt and Yule-Simon, are ways of quantifying counts over the rank order. If the Zipf distribution fit the plot of log(Count) on log(Rank) would be linear. It's not perfect, but it's pretty good.

      In this case it makes a lot of sense: Gaming is a social activity and a more popular ruleset will, on average, draw more players. The same pattern appears in many other contexts: Words or phrases in languages, populations of cities, distribution of species, links to web pages, citations to journal articles, and fans of teams being examples.
    1. Nylanfs's Avatar
      Nylanfs -
      Yea, graphporn!
    1. LordEntrails's Avatar
      LordEntrails -
      Quote Originally Posted by Warren LaFrance View Post
      With the current spat roll20 is having with a few youtube and twitch broadcasters I wonder if there has been an increase in FG user base...?
      Yes, I know of a dozen or so. But that's not statistically relevant.

      This data is only through the end of September. So the Roll 20 fallout won't be shown. Next quarter data for FG, and especially Roll 20 (if the Orr group releases it), will be what actually indicates the significance of the spat.
    1. Nylanfs's Avatar
      Nylanfs -
      I wonder if Doug has a stat for Demo sessions, or perhaps demo downloads that could be tracked.
    1. TerraDave's Avatar
      TerraDave -
      Quote Originally Posted by Morrus View Post
      For those of us who don’t have a mathematical background (ie me)could you explain what that is?
      Its means winners really win.

      So, for example number 1 is twice as many as number 2 which is twice as many as number 3 and so on.

      What the plot shows, where 5E is actually above that orange line, is that its really really winning.

      Which some of us have been saying for a while.
    1. Jay Verkuilen's Avatar
      Jay Verkuilen -
      Quote Originally Posted by TerraDave View Post
      Its means winners really win. What the plot shows, where 5E is actually above that orange line, is that its really really winning.
      Certainly it is on Fantasy Grounds.

      One potential cause from deviations from Zipf's law is violations of independence of the counts, which is pretty likely in this case. For example, there are likely to be DMs who have multiple games and are, all other things being equal, more likely to run the same ruleset. It also doesn't distinguish defunct from running games (at least I don't think it does).
    1. Nylanfs's Avatar
      Nylanfs -
      No, AFAIK all the games that Doug reports on are from the GM client pinging the FG license server, so those are all active games, planning sessions or simply playing around. None of them would be non-active campaign. I remember this being discussed somewhere in one of the earlier posts.
    1. Perry Chalmers's Avatar
      Perry Chalmers -
      Quote Originally Posted by Nylanfs View Post
      No, AFAIK all the games that Doug reports on are from the GM client pinging the FG license server, so those are all active games, planning sessions or simply playing around. None of them would be non-active campaign. I remember this being discussed somewhere in one of the earlier posts.
      IIRC, it actually only counts session where players connect via the server alias, but doesn't when players connect via the GM's direct IP, it also doesn't count GM prep.
    1. Nylanfs's Avatar
      Nylanfs -
      You can also track some info from Steam also.
      https://steamcharts.com/app/252690#All
    1. darjr's Avatar
      darjr -
      So those are actual games? And it may exclude some home games using the platform? Am I reading that right? It’s under reporting actual games?
    1. LordEntrails's Avatar
      LordEntrails -
      Those numbers only include sessions where at least one player connects and the session lasts at least so long long (maybe 1 hour, not sure) and that those sessions connect to the FG license server.

      Connections to the FG license server,
      ARE made when;
      - A free license connects to a host/GM using the Ultimate license
      - A client uses the Alias system to connect to the Host/GM

      Are NOT made when;
      - A standard license player connects to a host/GM using a direct IP address

      The numbers measure something very different than the Roll20 numbers.

      But what these numbers show are trends and relative popularity. It measures a consistent set of numbers which should be very representative. So you can see how 5E compares to StarFinder and you can see how fast any given system is growing (or not).
    1. Nylanfs's Avatar
      Nylanfs -
      It would also not include people that have to use a VPN service because of IPv6 address issue.
    Comments Leave Comment