Tuesday, October 25, 2011

WoW Player Activity

There is this chart making it's rounds on the internet for quite some time now. It shows player activity between 2005 and 2009 as measured by Warcraftrealms.com and its addon CensusPlus.

There's also this chart of the last year of player activity at Warcraftrealms.com.

Now, the big question, of course, is how this data correlates with subscribers. After all, everybody would like to know how deep the slump between patches actually is for Blizzard. But they insist on only telling us the subscriber peaks.

Since I love riddles and playing detective, I compared this data with the official player numbers by Blizzard between 2005 and 2009. We have very good data from this time intervall, because Blizzard was so surprised and proud about the success of WoW that they constantly told us the newest peak; often every month. The data comes from mmodata.net

Little reminder: TBC was released January 2007. WotLK was released November 2008.

The red line is the official subscriber number. The blue line is the player activity scaled with the factor 11. As you can see, during classic WoW the player activity was very good at predicting subscriber numbers. Later it tended to underestimate the number of subscribers. One reason is probably that WOWCensus only works for NA/European realms and WoW started to grow rapidly in China and other places.

If anybody has some WoWCensus data for a larger timeframe, or knows why January seems to be counted twice in that chart, please leave a comment.


  1. I wonder why we get a rogue legendary and not a warlock legendary.

    Or, surprise, surprise, the caster legendary didn't really help to make warlock more played. Clearly it will lead to more rogues.

    And what are you trying to say with the last graph? That since WotLK player spend less time in game then before?

  2. I am just saying what I am saying, Kring :).

    I would be interested in finding out more about this Warcraftrealms census data. And I did this little analysis to find out how credible it actually is. The result is: Well, it's at least not completely useless.

    I also used this opportunity to link some WoW activity data. I know that some readers will react hostile towards it and are right now pondering about how to elegantly comment that WoW is not declining at all. Hi guys! :)

    But, anyway. I think this post has some valuable information and maybe someone leaves an insightful comment regarding this Warcraftrealms.com page. I have a long tradition on this blog figuring out how well WoW does. This post is just one in a row.

  3. I like data and charts. And just because it. like almost all business data, is not 100% accurate or "provable" does not prevent it from providing insights and being interesting.

    However, I think there is a real problem with the WoW numbers: China. Customers dropped during LK as a couple of million Chinese "subscribers" literally overnight due to the dispute with the company/government. AB quit releasing the figures presumably when they would have shown a huge and first ever subscriber drop. (Probably rightly so; a contract dispute in China says nothing about WoW's prospects in the rest of the world and little about WoW's long-term prospects in China. ) Nor have I seen anything that says where the latest .5 million drop was evenly distributed. I.e. did it decline 5% everywhere or decline 1.5m in US/Europe and gain 1m in Asia for a net -0.5m. I.e. we do not know if US/E subscribers grew slightly or dropped 25% in the last year.

    Regardless, declining activity levels are not a good thing for the publisher.

    So with little insight into what the US/E subscriber count is, the warcraftrealms data is not nearly as useful as it could be.

    It seems to me that the AB corporate philosophy has changed from where peaks were a PR # released periodically to where we get a number every fixed interval, i.e. in the quarterly 10-K filing. We should get a new data point next month.

    Currently a trivial number, but going forward as the API and websites get more robust, we may some day need to be careful what "playing the game" means. CCP said they eventually wanted you to be able to do any non-combat thing in the game from the web site. Certainly I think all the gold bloggers who currently spend hours browsing/buying/selling the WoW AH (e.g. iPhone app or when they should be working/in class) are playing the game even though they are not logged onto a legacy client.

  4. I know Daeity's blog - Digital Castration - covered WoW subscription numbers in a lot of detail some time ago. Very interesting reading if you havent already checked it out.

    And as you mentioned in your post Nil's, the WoW fanboys didnt take too kindly to most of his conclusions unfortunately. Stuff like "Who cares how many subscribers they have, its irrelevant" or "You don't have all the information so STFU you're wrong".

    Of course these same people will whip out the old "Well WoW has 12 million subscribers so your criticism on any aspect of the game is wrong" argument in their very next post. Typical hypocrites.

  5. Eyedol, you always have those on both sides. It's not really fertive to try to discuss with preachers (of either side).

    However, you have to precisely differentiate between numbers and conclusions. Just because x goes down, doesn't mean it goes down because you say y caused it. x goes down = fact, y made x go down = your conclusion.

    I like this seemingly new approach, Nils. For me, the numbers always (should) come first and then you draw the conclusions. Instead of setting up a theory and trying to prove it by numbers.

  6. Here is another way to resolve these issues. On the financial statements (detailed ones) is a line item that tracks how much revenue Activision Blizzard makes in WOW related activited each quarter.

    This activity isn't fully explained and I'm sure includes things like transfers and pet sales. But it is an indicator. For the second quarter they made 389 million which was down from over 420 million in the 1st quarter.

    Maybe this was due to sales of the game as it does not state if game sales are included or excluded. But if we can get more info here we can easily tell if they are taking in more or less money.

    Speaking of that the company set a new 52 week high in their stock price and in fact hit a new multi-year high (I think going back to 2009). But this might also be due to all the other games that Activision has and not directly related to WOW even if it is their cash cow.

    Finally, cash on hand for the company dropped by 300 million for the 2nd quarter. They will release 3rd quarter number on November 9th I think.

  7. @Goodmongo: Thanks a lot for that info!

    @Srosh: I don't know what conclusion you are talking about. The post deliberately doesn't include any conclusions :)

    @Eyedol: Didn't know that blog. Thanks!