Translate this page

Some beautiful music to read the blog with

Sunday, 19 April 2015

Eve Online Subscriber numbers (ex China) - i estimate 405k

[a non-ISK making post but one on CCP / Eve Online]

For what it is worth, i believe the CCP subscriber numbers are about 405k (give or take 20k).  Of which 339k pay in cash, and 66k pay using Plex.

Where is this post born from?

The recent podcast episode from Broadcast from the Ninveah pointed me in a post from The Nosy Gamer blog (which i had missed) that attempts to take two snippets of information from CCP to determine the number of subscribers to the game.

Firstly, CCP stated that "Eligible voters cast 36,984 votes, meaning that we have 15% increase since the last year’s election" [actually, that is not new news].

Secondly, CCP stated (in the same place) that "We have also noticed a 3% turnout increase since last year . . . " [that is new news].

From this information several bloggers and commentators have made their estimates as to the current number of Eve Online subscribers.

When was the last time we had decent Subscriber number data?

Two places.

1) the Prospectus issued during 2013 (for the Convertible bond issue) talked of subscriber numbers for the end of 2012.

2) the dev blog post on the CSM8 election results which gave the votes cast by country + Voters as a % of subscribers + % of subscribers back in April 2013.

The Dev Blog is simpler.  For the UK it states there were 5,787 voters representing 11.64% of the vote, giving 13.74% of voters as a % of subscribers and therefore 10.28% of subscribers (not easy to take in!).

The way i am reading that is in the UK there were 42,118 subscribers (=5787/13.74%) and so 409,707 total subscribers (=42,118 / 10.28%).  It seems to work with all the other countries to give me 409,707 players at that time.

Given the stated 49,702 votes, that indicates that 12.1% of the voters voted.

I am pretty sure that includes Plexed accounts - it does agree with work i have done before and reconciles roughly with information in the Prospectus.

Can we get from there to players at the end of 2013?

Work i have done before suggests there were an average of 476.8k players (subs + plex) during 2013.  See at the bottom section of this post for the workings.

Essentially, I have worked from the Eve Online (ex China) revenues, made assumptions for average time of a subscriber / average monthly payment / Plex vs Subs mix to get an answer for the average number of players in the year.

Given that, i assume that the year end number is about 7% higher than the average at 510k because of the upward trajectory in player numbers during the year.

And now to the current Population

We were told last year that the number of voters in the CSM9 election was 31,294.

We were told this year that the number of voters in the CSM10 election was 36,984.

We were told that the increase in turnout was 3%.  Now, the simple maths from all this is that:

[CSM10 voters/Population now] - [CSM9 voters/Population last year] = 3%

or, inputing what we know:

[36,984/Population now] - [31,294/510,000] = 3%

That suggests that the Population now is 404.8k.

My work on getting the 2013 population numbers

Below is the relevant part from my spreadsheet that tries to look at revenues generated from subscribers.

Not easy.  Also, it looks at average number of subs in the year, not year-end.

Almost all of the workings come from the Prospectus in 2013 for the Convertible Bond issuance.

Starting from the top i have:

  • Average number of subs players in the year (excluding Plex players)
  • Average lend of sub (i assume 10 months given the steep learning curve)
  • How the sub players payment periods are split 1 mth / 3 mths / 6 mths
  • Cost per month sub for each period
  • and therefore the effective monthly payment from each sub
  • Gives total Sub revenues for the year
  • Then i make an assumption of how much comes from Plex in the revenues
  • That gives me the total revenues
  • Below that i try and split those Plex sales into players vs investors

Gives me an average number of players in 2013 as 476,750

For the above, i assume the 2013 year end players are 7% higher than the average given an upward trajectory during 2013 = 1.07 x 476,750 = 510k.

As a check, if i assumed 2014 was going to end 10% below my estimate of the average for 2014 then i would get 0.90 x 449,928 = 405k (which agrees with my estimate above).

Eve Online - ex China
2011 2012 2013 2014F
Number of subscribers 355,000 355,000 404,838 383,734
change 10.9% 0.0% 14.0% -5.2%
average length per subscriber (mths) 10.1 10.2 10.1 10.1
10.1 10.2 10.1 9.9
Price per subscription for 1 mth 60% 60% 60% 60%
Price per subscription for 3 mths 30% 30% 30% 30%
Price per subscription for 6 mths 10% 10% 10% 10%
100% 100% 100% 100%
Price per month
Price per subscription for 1 mth 14.95 14.95 14.95 14.95
Price per subscription for 3 mths 11.36 11.36 11.36 11.36
Price per subscription for 6 mths 9.55 9.55 9.55 9.55
Average subscription 13.33 13.33 13.33 13.33
Subscription revenues 47,851,724 48,325,295 54,516,821 51,674,859
change 8.2% 1.0% 12.8% -5.2%
Sales mix
Subscriptions 77% 77% 76% 75%
Plex sales (to play, not hoard) 23% 23% 24% 23%
100% 100% 100% 100%
Plex sales (to play, not hoard) 14,293,372 14,434,828 17,215,838 15,846,957
change 14.6% 1.0% 19.3% -8.0%
Price per Plex 19.95 19.95 19.95 19.95
Number of Plex sold (to play, not hoard) 716,460 723,550 862,949 794,334
Number per month 59,705 60,296 71,912 66,194
Additional "accounts" as a % of Paid accounts 17% 17% 18% 17%
Total Revenues 62,145,096 62,760,123 71,732,659 67,521,816
change 9.6% 1.0% 14.3% -5.9%
Total Accounts + Paid & Plex (average) 414,705 415,296 476,750 449,928

If required, i can go deeper into my workings and thinkings.

[Edit: corrected Typo, Ninveah instead of Nirvana!]


  1. I'm sure Kirith will be pleased to learn that he is a demigod....

    Ninveah, not Nirvana :P

    Rob K.

  2. I'm sure Kirith will be pleased to learn that he is a demigod....

    Ninveah, not Nirvana :P

    Rob K.

  3. Your math is wrong, I'm afraid. Even if we treat the number of subs in 2014 as an unknown, X, the fact that Leeloo says turnout was up 3% allows us to conclude that EVE subs were down between 10 and 15% from 2014 to 2015.

    There were 31294 voters in 2014, 36983 voters in 2015. If that represents a 3% turnout increase, simple long division lets us calculate what happened to EVE's subs in that one year. Here's the table:

    On the left, pick how many non-China TQ subs you think EVE had in 2014. At the top, pick what you think Leeloo means by "3%" (whether the actual was 2.55%, 3%, or 3.49%). Where the two cross are EVE's subs now, and on the left is the one year decrease.

    For various reasons which I lay out here:

    I personally believe there were 380k TQ subs during the 2014 election. My count is supported by the number of names on the monument, which was fixed by number of subs right around the same time. So I estimate there are between 330k and 340k TQ subs now.

    1. *on the right is the one year decrease

    2. Does the Monument include the names of players paying by Plex?

    3. I personally can assure you, it does.

      Kara Bol

  4. I'll have to admit I'm a bit confused. In your table, you seem to have the number of subscribers as of Fanfest 2014 listed as 383,734. So if you know that figure, why did you plug 510,000 into your formula for determining the turnout percentage? You want the number as of Fanfest, not an estimate for the end of the year.

    Using the numbers in your post, the turnout percentage for the CSM 9 election was:

    31,294 / 383,734 = 8.1551%

    This equates with Ripard's estimate of an 8.2% turnout for the CSM 9 election. With CCP Leeloo announcing a 3% increase, I would use 11.2% in the formula for determining the CSM 10 turnout:

    36,984 / Number of paid accounts = 11.2%

    With number of paid accounts turning out to be 330,214.

    1. The top part of the table looks at the average number of players paying using a credit card etc during the year - that is the 383,734. The bottom part of the table then looks at the number of players paying with Plex, on average during the year - that is 66,194. To give a total average for 2014 of 449,928.

      And then I need to somehow deduce the year end number.

      For fanfest 2014 I used the prior column which had the 2013 numbers in it

    2. Then it looks like you are double-counting subscriptions paid for using PLEX. Because it doesn't matter whether you pay with RL currency or PLEX, you still count as a subscriber.

    3. Well, that depends. In the 2012 Prospectus CCP differentiated between "subscriptions" and "Plex sales" in their revenue number.

      My table tries to differentiate between people paying using "subscriptions" and people paying with Plex.

      I doubt i am double counting - there is a chart on page 75 of the Prospectus that shows the number of monthly subscribers to Eve. The wording leading up to that indicates that CCP views monthly payers as "subscribers" where as those that use Plex are counted differently.

      Though, when it comes to CSM voting, they are all the same - confusingly.

      Where i am most likely to be wrong is in my assumption that, on average, players pay $13.33 a month subscription to play.

      My main sense check for my 2013 average of 476,750 players is to observe that they created revenues of $71,732,659. If i assume that the average player lasts 10 months then that suggests that the average payment was $15 per month (a mix of subscription payments at $13.33 and Plex payments at $19.95).

      However, it is fair to say that an average year number of players does not nail down what the actual number of players was in any point in time.

    4. Worth bearing in mind that i am say that there are 339k players that pay cash or credit card (i.e. subscriptions). And another 66k who play using Plex.

      It is likely that those 339k pay US$ for the Plex and those 66k pay ISK for the Plex in game to finance their monthly game time.

      It is possible we are talking the same numbers . . . .

    5. I believe you are making this a lot more complex that it really is. We'll just have to agree to disagree.

    6. Fair enough - you have a great blog by the way :)

    7. Thanks. I just thought of something. If you are going to differentiate between subscriptions and accounts paid with PLEX, then you probably need to use the "subscription" number in that chart you used as the grand total, as that number appears to be the number of accounts eligible to vote in the CSM election. Then you can take the size of the pool of voters and divide it up between those paying with RL cash and PLEX.

  5. seems CCP confirmed there are 480k names on the Monument

    If so, that is close to my 510k estimate.

    If i plugged 480k into my maths then i get 388k players today (instead of the 405k).


Note: only a member of this blog may post a comment.