Monday, April 20, 2015

Another Look At EVE Online Subscription Numbers In 2015

As we approach EVE Online's 12th anniversary, the speculation continues about the game's subscription numbers. Admittedly, I contributed to the guessing game with a post three weeks ago speculating on what the subscription numbers1 were as of Fanfest based on the CSM election results dev blog. My post was a response to others who were touting divergent subscription numbers. While CSM member Mike Azariah was touting growth in the subscription numbers, ShadowandLight from the Legacy of a Capsuleer podcast was participating in (if not trying to start) a nasty whisper campaign, stating on EVE Radio on 26 March that an "inside source" told him that EVE was down to "146,000 subs."2 When I took all of the best information available, I came up with a subscription range of 319,000 to 342,000 on Tranquility.

The topic has also fascinated the writers at Massively Overpowered, the reincarnation of the site that closed in February. Brendan Drain wrote the second article on the site looking into the numbers provided by my post. While his numbers fall within the upper range of my estimation, we both come to the conclusion that the number of subscriptions has fallen 18% over the past two years.

While I only attempted to ascertain the basic facts on the subscription numbers, Drain put the numbers into perspective by attempting to compare what I had calculated and comparing the result to the industry as a whole. He pointed to a study done by SuperData, a company that provides "market intelligence covering the market for free-to-play gaming, digital console, mobile, PC downloadable, streaming media and eSports." The study, titled "MMO Market Report 2015", predicts that revenue for pay-to-play (subscription) games will fall to $2.3 billion this year, a 17.9% decline from the $2.8 billion earned by P2P games in 2013.

Comparing revenue and subscriptions is a bit like comparing apples and oranges. For example, dual and multiple character training, both introduced in 2013, had the potential of both reducing the number of subscriptions and increasing revenue. Other changes, like the institution of jump fatigue in the Phoebe release, possibly resulted in subscription losses not related in any way to a player's preference for free-to-play games over P2P games. But Drain's main point is still valid. In a shrinking market for P2P games, seeing a 12-year-old game lose subscriptions due to potential new players preferring F2P games over P2P games is expected. Seeing growth in such a market environment is remarkable.

In my opinion, CCP ended what I call the cult of the growing subscription in June 2013 when dual character training was introduced in Odyssey. In the era in which EVE launched in 2003, most games relied strictly on box sales and subscriptions for revenue, so the subscription number was a useful number for players to see if their game was thriving. After all, why invest in a virtual world that is in danger of closing? But in the modern age, with even subscription games having in-game cash shops and offering other services as micro-transactions, we can no longer judge a game studio's, or even a game's, health based strictly on the number of subscribers. For that, we know need to become accountants and peruse the quarterly, semi-annual, and annual financial statements of game companies, providing we have such access. CCP's financial statement for 2014 is due out soon and I'm anxious to dig into those numbers.


1. Some will say that accounts paid for using PLEX are not really subscriptions. For those people, replace the word subscriptions with the term "paid accounts", as paid accounts over 30 days old make up the voter pool in CSM elections.

2. ShadowandLight has since admitted he didn't know what the number he was spreading around actually meant, which leads me to believe he was just throwing mud around and hoping something would stick.


  1. Your point about multiple character training is a really good one, I'm sure some people consolidated their alts. But since you can only be logged into one at a time, I'm not convinced it will be enough people to make a big difference.
    I don't have the full superdataresearch report (because it hilariously costs $2,500), so I don't know how the pay-to-play numbers were composited, but I would guess that it's either revenue from only pay-to-play revenue streams or total revenue from only pay-to-play games. In either case, most of that revenue will be from subs or box sales, so it's not so much apples and oranges as it is comparing green apples and red apples. This is especially the case with EVE, as EVE subs and PLEX sales accounted for over 90% of CCP's revenue up until at least CCP's 2013 financial statement.
    We don't know how much revenue alternate streams like the NEX store, collector's edition, books and other EVE merchandising have brought in, but each of these has been a response to the same downward market trend and recognising a need to diversify EVE's business model. I think it's pretty clear that the subscription-only model is practically dead and even with supplementary revenue streams, the very idea of a mandatory subscription may be next in line.

  2. I noticed in your article that you mention the EVE monument numbers. That's all active characters, not accounts. So divide the number by 3.

  3. It's not, the monument contains only one character per account. CCP picked the character on each active account with the most skillpoints.

  4. The CSM election numbers don't give a real basis from which to estimate actual sub numbers. They mention percentages only, relative to past elections, and everyone who based their analysis on those percentages made unsupported and/or speculative assumptions as to the base sub numbers to which the percentages were being applied.

    And, yes, the CSM numbers were published in this fashion as a deliberate move on CCP's part - they have no wish for actual sub numbers to be released or calculated, since the numbers have been trending steadily downward, since their last public annoucement of sub numbers several years ago. It would simply be bad PR to acknowledge a significant drop in sub numbers, for whatever reason.

    As for DCT and MCT, those account consolidation effects on sub numbers leveled off more than a year ago and certainly no longer contribute to the ongoing drop in player accounts.

    The 2014 financials should give a better picture, depending on how much creative accounting CCP can still apply to their balance sheets. They have already written off a lot of DUST and WoD losses, so they don't have much more room to wriggle.

    Based on EVE Offline numbers, however, I'd estimate that CCP conservatively lost at least 10-15% of accounts in 2014, primarily casual industry players who were effectively kicked in the nuts by the ill-conceived Crius changes. As was predicted, these players appear to have quietly let their multiple pre-paid high sec industry accounts run out through the year, perhaps keeping a single account for PVP (or not). The losses due to casual player accounts unsubs appear to have finally leveled off by end of 2014 (again, as predicted), so we hopefully won't see a repeat of something similar in 2015.

    Ofc, this will depend on how Fozzie Sov actually plays out....

  5. Even if EVE is losing players, what would be the point when
    the server gets turned off? If the game
    is still making a profit and the server gets turned off because at some point
    in the future the game is not as profitable as it was some point in the past
    that would be stupid. On the other hand,
    if the server gets turned off because operating expenses < income, that
    would be the point to throw in the towel.

    What are the bare bones requirements to keep the game on? Internet connection for the server, power for
    the server, someone to watch the server and fix anything that breaks, physical space
    for the server including cooling, and maybe a budget to upgrade the hardware
    over time. There is some portion of the player
    base that enjoys the game enough to keep paying even with all the changes. As long as the last update to the game before
    all the dev’s are fired is stable, profit could be squeezed from the people who
    want to keep paying until the subscriber base can’t pay the electric and
    internet bill.

    In the death-knell of eve, imagine blades being pulled as
    the subscribers give up and the power to keep all the blades on is greater than
    the paying subscribers can cover.
    Eventually there is one care-bear alone with 3 accounts who has claimed sovereignty
    in all of space. Eve is running on a VM
    somewhere in a data center but since there is almost no one playing, no one notices. In the lone subscriber’s mind, he won
    EVE. Everyone else won when they left.

    Hell, if the subscriber numbers ever got so bad that the
    game could be hosted on a VM somewhere and there were no development expenses
    the game could limp along indefinitely as long as the hand full of subscribers
    paid enough to cover the hosting cost and likely extra hosting charges for the
    size of the database of all the items belonging to long gone characters.

  6. Yep, the CSM numbers only enable us to calculate a percentage drop. But all I did was apply that percentage drop to known data points and reasonable extrapolations on the number of global subs to reach a conclusion on where subs sit globally today, which is around 410k.

    This figure makes the assumption that the ratio of players between Serenity and TQ has remained relatively stable (a fairly safe assmption). The financials won't give a more accurate picture unless CCP's 2014 report actually includes direct figures on subs.

  7. "... reasonable extrapolations..." and "...fairly safe assumption..." are what I consider to be unsupported and/or speculative. Extrapolations and assumptions are not actual data.

    The financials, however, will paint a clear picture of how CCP is doing as a company - and, correspondingly, how well EVE Online is doing as a product, particularly inasmuch as EVE Online is CCP's only real revenue generator. And, the revenue numbers - as compared to 2013 - will be fairly indicative of whether the number of player accounts has gone up or down.

  8. @noizygamer when you say "I since admitted" regarding my statements being told there was 146k paying subs, I flat out from day one stated the source and info were unverified from the very 1st time I spoke of it on Eve Radio. We still have no hard evidence on subs, the best guesses put active subs at barely above 300k and who knows how many are paying vs plexing. I'm not sure why you continue acting like I wrote the Holy Scripture on Eve subscribers and penned in 146k as gospel but your starting to lose a lot of credibility when you have recorded audio of my exact statements but then try to write it in a different light. Frankly it seems your starting to use various forms of misinformation or bending reality to try and spin to fit your narrative. Listen to my exact words from the various shows about the topic and feel free to *fully* quote me, I appreciate your future journalistic integrity in advance.

  9. I guess you didn't read the notes at the bottom of the post. Your spreading of unverified information in order to make CCP look bad is why I described what you were doing as engaging in a nasty whisper campaign. Thank you for verifying what I wrote in these comments.

  10. Discussing information, verified or unverified, is allowed and encouraged when the organization (CCP) refuses to provide hard details. Even with your "in-depth" analysis (which you do by standing on the shoulders of others who did the work) is at best a shot in the dark hoping that you can correlate voting data with subscriptions. You must admit that at BEST your engaging in extremely shaky statistically fishing, with zero actual hard data to go on for years (not your fault though in that area). How is spreading such information any better or any worse since none of it can be verified until CCP releases their 2014 fiances?

    My real issue though is you have recorded audio to use for your "quoting' yet you refuses to do so, probably because it wouldnt fit your narrative trying to paint me in a bad light. It's quite sad actually, you probably could be a decent reporter for gaming news in general yet let your biases blind you to doing things like factual reporting and using industry standards such as referencing. It's no secret your hatred of ISBoxer and multiboxing in general. You tend to make statements trying to lump "hackers" with people who use 3rd party tools to multibox, but somehow give a pass to all the other 3rd party tools out there which undoubtedly let people have an "unfair advantage".

    You've also continually refused to debate me in a public forum regarding ISBoxer / Multiboxing / 3rd party programs but feel perfectly comfortable lobbing grenades from behind your keyboard without having people debate you. I find that a bit... unsettling.

  11. Okay, that's it. No more feeding the troll.