Wednesday, January 8, 2014

MMOData Closing Shop

I'm a couple of weeks late with the news, but MMOData published its last chart on 28 December.  The site with the best collection of subscription numbers for MMORPGs will shut down in June.  Ibe Van Geel posted the following reason:
"The biggest reason is that it is getting increasingly difficult to get any useful numbers. Also many of the subscription based MMORPG's went free to play, and their companies tend to not give out useful active accounts numbers."
Van Geel had a pretty good relationship through the years with CCP, even receiving data about EVE Online throughout the Summer of Rage in 2011, showing the ultimate 7.2% drop in subscriptions before the Crucible expansion drove numbers higher again.  But that relationship apparently ended in the summer of 2012, as the last data point for the Tranquility server was for May 2012.  Van Geel had this to say about the Icelandic game company:
"As for the only growing subscription based MMORPG left, EVE Online, I no longer receive subscription numbers. CCP no longer responds to my requests.

"It is a shame really, I doubt that their numbers are declining, but I lost a bit of confidence in CCP, I hope they will give out useful numbers to the public on a consistent basis."
Since then, the only announcement of EVE Online subscriptions was made on 28 February 2013 when CCP put out a press release announcing that the game had surpassed 500,000 accounts.  However, that number included China and players wound up estimating that Tranquility was home to approximately 400,000 accounts.

I'd also like to hope that CCP will once again regularly release subscription statistics.  I realize that the industry practice is to not release numbers unless absolutely necessary, like Activision does for World of Warcraft during its quarterly investor calls.  But one of the things I like about CCP is its openness to the players about things other companies try to hide.  At this point the only indicator the players have of the health of EVE is the average concurrency chart Chribba provides at

So if anyone is interested, Van Geer has made the charts and data he has collected over the years available for downloading.  The links are:
- The archive :
- The blogspot :
And to Ibe Van Geer, thanks for all the work you put into making MMOData such a valuable resource. 


  1. That's too bad. It was a near-impossible task and never appreciated by very many. As with Sir Bruce's numbers, too many people were eager to shout "bullshit" when a given number did not align with their personal wish of how things should be.

    And things have gotten much more opaque since F2P came along. Who do you count? I have a lifetime subscription to LOTRO (best MMO deal ever for me at this point), do I count as "active" always or only on months where I log in or only if I log in for some minimum duration of time? What makes me count as an active player?

  2. The CMS election statistics for participation per country given in are sufficient to calculate subscribers numbers. Since the column labels are not very helpful, here a breakdown: The columns are
    1: Name of Country: c
    2: Number of Votes by that country: $2
    3: Percent of Total Votes: i.e. $2/SUM($2)
    4: Percent of Possible Voters of that country (v(c)), i.e. the participation: $4 = $2/v(c) => v(c) = $2/$4
    5: Percent Subscribers of Total Subscribers (t): i.e. $5 = v(c)/t

    You can substitute 4. in 5.: $5 = v(c)/t = $2/$4/t, therefore t = $2/$4/$5. If you do this, you get values from 408.5k to 410.4k subscribers. I am not sure which date is relevant for the data snapshot, but election was April 2013.

    1. Your method comes out pretty close to what the consensus was. The thing is that CCP would give MMOData the numbers 6 times a year (every other month starting with January).

    2. Yeah, I didn't mean to say, we have a good alternative (we are not even guaranteed to get those statistics for the next election). I like playing with numbers and reading your post I remembered we actually had a +/- 1k data point, which nicely confirms that the consensus you cited was spot on.