WoW Subscriptions Around The World
In January 2007, when Blizzard announced 8 million subscriptions worldwide (Source):
China: 3.5 million
North America: 2.0 million
Europe: 1.5 million
Remaining Territories: 1.0 million
In January 2008, when Blizzard announced 10 million subscriptions worldwide (Source):
China: 4.5 million
North America: 2.5 million
Europe: 2.0 million
Remaining Territories: 1.0 million
Blizzard hasn't made any other official announcements detailing breakdown of subscribers per geography since then, but based on growth trends I think it's pretty safe to assume that their current (12 million) subscribers are approximately:
China: 5.5 million
North America: 3.0 million
Europe: 2.5 million
Remaining Territories: 1.0 million
These figures actually confirm a lot of other estimations out there that put Chinese subscriptions at between 5 and 6 million total. Or basically, "about half" of total worldwide subscriptions.
How many real players are there, though?
In North America and Europe, this information is really difficult to confirm. Blizzard does not release this information to the public so it's up the users to figure out approximate counts.
The Warcraft CensusPlus UI Mod is an addon for WoW where players can poll players online, and all of the information is collected to generate an approximate census of WoW in NA and EU. However, this information is not very accurate, there's a lot of duplication of numbers, and data can be skewed or altered by users.
Right now, for example, total NA and EU players are calculated to be 6,417,374 total users on the census page. However, according to Blizzard, their subscription count in NA and EU is about 5.5 million maximum. =]
In China, however, it's a different story.
NetEase operates WoW in China, not Blizzard, and they're not as restrictive with their policies or information. (Gold/Account trading is perfectly legitimate there by the way.) NetEase publishes raw data across all of their realms, and is available in census format (showing peak logins for both factions and on each realm).
According to Blizzard, total subscribers in China are about 5.5 million. According to their servers, however, total players peaked at about 2.4 million as of this month. That's the most I've ever seen it.
Estimated players in NA/EU
For this, I'll start out by using Blizzard's official announcement from January 2008 when they announced their 10 million subscribers.
China: 4.5 million
North America: 2.5 million
Europe: 2.0 million
Remaining Territories: 1.0 million
During this time, their Net Revenue for World of Warcraft was $275 million. Keep in mind, that this figure also includes WoW sales, Paid Services, Pets, etc.
Since we know how much they pay for WoW subscriptions in each of these geographies, we can get a much more accurate estimation of how many players there are at the most.
China pays ~$7.27 per month
North America pays ~$15 US per month.
Europe pays ~$17.6 (12.99€) per month.
Remaining Territories pay about ~$15 per month.
NetEase, however, has a special contract with Blizzard. NetEase collects subscription payments, and they pay Blizzard a 55% Royalty Fee. So, Blizzard is collecting ~$3.99 per month from each Chinese subscriber.
Using these figures and their subscriber counts, here's what their revenue should have been:
China: 4.5 million x $3.99 x 3 months = $53,865,000
North America: 2.5 million x $15 x 3 months = $112,500,000
Europe: 2.0 million x $17.6 x 3 months = $105,600,000
Remaining Territories: 1.0 million x $15 x 3 months = $45,000,000
That's a total of ~$317 million when it _should_ have been $275 million.
That $275 million figure also includes a ton of other revenue too (game sales, paid services, etc). For example, if subscriptions were to totally match revenue, then there would be 8.6 million players out of 10 million subscribers. However, if there was $50 million worth of game sales and services during that time, then the real players would have been 7 million out of 10 million subscribers.
So, let's try something more recent (e.g. 12 million subscribers) and where we know approximately how many sales there were out of the total revenue to get a better picture.
During the third quarter ending September 30, 2010, total WoW revenue was $289 million. During this period, Blizzard launched WOTLK in China (Aug 31) and they peaked at 12 million subscribers. Since Chinese players did not need to purchase WOTLK, this provides a better revenue figure to work with since it won't include those larger amounts of Game Sales. (Paid Services, Pets, and minor sales will still be included in this total revenue figure however.)
China: 5.5 million x $3.99 x 3 months = ~$65.8 million
North America: 3.0 million x $15 x 3 months = $135 million
Europe: 2.5 million x $17.6 x 3 months = $132 million
Remaining Territories: 1.0 million x $15 x 3 months = $45 million
So, total revenue _should_ have been about $378 million, but it was $289 million instead.
Let's say however, that there are a total of 3 million real players out of 5.5 million Chinese subscribers, and re-adjust those numbers. That's still $342 million though.
Hmm.. okay, let's say China player count is 50% of official Subscriber counts (like what it should be). And, all of the other geographies are 3/4 of their official counts.
That's a total of $267 million, which is much closer to the $289 million total. That also leaves $22 million in potential game sales, pre-paid services, pet sales, etc. If that's the correct number, that would mean that there are 7.6 million real players out of 12 million "Subscribers" worldwide. Does $22 million sound too low or too high for all of those purchases during a 3 month period though?
Let's get an absolute maximum figure: Total real revenue was 76% of what it should have been, so if you apply that number to total subscribers that gives you approx. 9.17 million players. That's the absolute most it could ever be, assuming that there were no game sales, no paid services, no pet sales, absolutely nothing but subscriptions during that 3 month period.
Of course, we know that's not true.. if we knew how much Blizzard sales (outside of subs) are during a 3 month period we could get a much more accurate estimate. For now though, real players are probably about 7.5 to 8 million out of 12 million total.
Previously, it was estimated (taking a different approach) that the maximum possible number of players were about 7.2 million out of 11.5 million total. (That's equivalent to 7.5 million out of 12 million.)
However, there's one important item I've been leaving out. In April 2009, NetEase paid for a 3-year license agreement in the amount of $30 million for the right to operate World of Warcraft. In addition to that flat rate, they also pay recurring license fees, royalties, and consultancy fees to Blizzard. (Several sources estimate "minimum annual revenue shares of $180 million" for example, meaning that they must pay Blizzard at least $15 million per month. Is that minimum subscription, or is that on top of royalty fees?) There could very well be another $30-45 million, for example, included in their revenue figure that has nothing to do with subscriptions. (That would reduce 7.5 to 6.8 million.)
However, for arguments sake I'm using the more inflated figures in Blizzard's favor.
The Magic Number
One thing to remember too is that subscription numbers are always fluctuating from month to month, and there are always different factors at play (seasons, server issues, ban waves, new expansion packs, boring content, etc.) There are always players coming and going from the game (e.g. new players or attrition.) Over one quarter, 100,000 players might leave because of boring content.. but 60,000 new players might join.
Blizzard only announces when they peak at certain numbers (maximum number of simultaneous "Subscriptions"), but if you look at all of the possible real players who have joined, played, and stayed or left the game, it's possible that World of Warcraft has cycled through 20-40 million unique individuals around the world over the years. That's actually really cool. =]
(But it's impossible to know the number for sure unless Blizzard releases all total WoW Classic digital/retail/everything sales since day 1.)
Over the past 3 years, though, the subscription numbers have plateaued (no longer growing at explosive rate seen before) and it seems like the "magic number" they're hovering around for real players is probably between 6.5 and 8.5 million (at any time playing World of Warcraft). That's still a freakin' huge number of players for a 6 year old game.
This gives you a really good idea of Fantasy MMORPG POTENTIAL too. Given that you have an awesome game, great gameplay, an addiction factor, and your game is as popular as it could ever be - your maximum potential number of players will be about 6.5 to 8.5 million players. As a publisher, if you can attain that amount, you've reached the maximum potential possible for that game. I don't think SCI-FI MMORPG can reach the same heights as Fantasy.. but Real Life Simulations (Call of Duty / GTA) could probably exceed it as that genre has a larger market demographic.
I already have an old entry that explains Active Subscriptions, and multiple blog posts covering many facets of this topic. However, I decided to make a new entry to help further explain that there is a difference between real players versus Active Subscribers™. But, I'll be taking a different approach.
Blizzard Has 12 Million Subscribers
In the older post (August 2010), I examined real player numbers during a time when World of Warcraft's Subscriber count was officially 11.5 million.
However, this became a major problem for many readers (here and on other forums). The biggest excuse used was, "Well.. these numbers must be totally wrong because Blizzard's last announcement stated 12 million players for a fact! This is really old now too, so none of these numbers apply to today."
Just to clear things up, the same concept of real players vs Subscriber counts apply no matter what Blizzard's Subscribers presently are.
For example,
Blizzard announces 9 million Subscribers, but there are 5 million real players.
Blizzard announces 10 million Subscribers, but there are 6 million real players.
Blizzard announces 11 million Subscribers, but there are 7 million real players.
Blizzard announces 12 million Subscribers, but there are 8 million real players.
Based on those trends, I think it's safe to assume that if Blizzard announces 13 million Subscribers (and absolutely no variables change), there are probably about 9 million real players.
So, if this post is read towards of the end of this year, when Blizzard reaches 13 million players for example, the same concept applies. This should be a no-brainer.
Blizzard Only Makes Announcements When They Peak
When Blizzard reaches a certain Subscriber milestone, they make an announcement. There might be an announcement if they reach 12.5 or 13 million Subscribers for example.
On Oct 28, 2008 Blizzard "surpassed 11 million subscribers". On Nov 21, 2008 Blizzard "reached 11.5 million subscribers."
In February 2010, Mike Morhaime stated that "there are currently 11.5 million World of Warcraft subscribers, the same number of users the game had in December 2008." That's from Dec 2008 to Feb 2010! There was no growth during that time, but they didn't make any announcements of Subscriber decay.
In fact, there was quite a substantial amount of Subscriber decay during this time.
From June 2009 to October 2009, The9 was transitioning ownership of WoW to Blizzard's new partner NetEase. During this time, all servers were offline and there was no WoW in China.
When servers were brought back online in October 2009, NetEase was ordered during the following month (Nov) to stop charging current players and stop new registrations until the Chinese government could approve the game. From November 2009 to March 2010, NetEase was not allowed to collect subscription fees or register new players.
Ignoring this fact, let's just examine the 4 months when servers were offline and there were no subscriptions or players.
That was a loss of approximately 5-6 million players world wide. There were no announcements made, e.g. "Blizzard announces World of Warcraft is down to 5 million worldwide!" After all, why would they?
However, during this entire time news websites around the world continued to promote 11 or 11.5 million Subscribers, even though the number was closer to half.
On August 5, 2009, Activision Blizzard held their quarterly earnings call for the quarter ending June 30, 2009. Now, keep in mind that there was no WoW in China during June of 2009.
Activision Blizzard stated in their SEC filing that World of Warcraft "is played by more than 11 million subscribers worldwide."
In that same Form 10-Q, they also stated: "Our new licensee for World of Warcraft in China will not be able to begin offering the game in China on a paid basis until we obtain all required regulatory approvals, and we do not know when that will happen. World of Warcraft was unavailable to players in mainland China from June 7, 2009 (local), the date on which our prior license arrangements expired, until July 30, 2009 (local), when our new licensee began making the game available to our previous players in China in a test format without charge."
(Note: They still didn't have all approvals, so the date was extended beyond July 30, 2009 - but that was the limit of the information that they knew at the time.)
So, even though they were around 5-6 million official Subscribers during June - Blizzard still stated in an SEC Filing that WoW is played by more than 11 million subscribers worldwide.
This is completely honest and true, no matter what you might believe. It's all about wording, and that's one of the points I'm trying to make.
1. They only make announcements when they peak Subscribers.
2. When they say "20 million Subscribers", that doesn't necessarily mean that they actually have "20 million subscribers" (or anywhere close to that number) at that time.
3. When they say "20 million Subscribers", that doesn't necessarily mean real players.
It's all based on wording and their last official census announcement.
Right now, for example, they have "12 million Subscribers". Even if something bad were to happen, Blizzard can still tell news outlets "At last count, we are at 12 million Subscribers" or they can say:
We're currently at 12 million Subscribers worldwide!Even if Subscribers (or real players) are actually closer to 4 million. It's still a completely true statement.. based on their own findings or methods.
*based on internal figures and calculations.
Why Do You Keep Capitalizing "Subscribers"??
I want to ensure that there's a distinction between Blizzard's definition of "Subscribers" and what you might consider a subscriber.
Yes, Blizzard has a special internal definition of what a "Subscriber" actually means. If that's not proof alone that there are big differences between real players and Blizzard Subscribers - I don't know what is.
Subject Lines and Header Titles might say one thing, but please read the fine print. All MMO companies do this with their numbers, it's not just limited to Blizzard and this shouldn't be a big secret (apparently, though, most gamers are just in complete denial about this distinction).
Here is Blizzard's official definition of a Subscriber™:
World of Warcraft’s Subscriber DefinitionSo basically,
World of Warcraft® subscribers include individuals who have paid a subscription fee or have an active prepaid card to play World of Warcraft, as well as those who have purchased the game and are within their free month of access. Internet Game Room players who have accessed the game over the last thirty days are also counted as subscribers. The above definition excludes all players under free promotional subscriptions, expired or cancelled subscriptions, and expired prepaid cards. Subscribers in licensees’ territories are defined along the same rules.
- a Subscriber is not one single player.
- a Subscriber is any account that has a paid subscription fee (if you haven't logged in, but still have an active subcription - then yes it counts).
- a Subscriber is ALSO any account with a first month's subscription.
- a Subscriber is ALSO any person who has an active prepaid card.
- a Subscriber is ALSO any "Internet Game Room" players who have accessed the game over the past 30 days.
What this means:
- If an account gets banned and a new account is purchased during the same month, that counts as two active subscriptions for one user.
- You pay for the month at the start of the month. When you get banned, you still paid for that entire month and Blizzard won't be refunding your money. The banned/cancelled account will no longer apply once you reach the end of the month.
- A "first month's subscription" is very common especially for gold farmers and gold sellers/traders who need to create new accounts frequently.
- Even though "expired prepaid cards" are mentioned, the cards in fact never expire. You can purchase one in NA/EU, but wait several months/years to activate it for example. When you purchase a prepaid card in China, you can keep adding time to it for as long as you want.
- Blizzard does not provide a definition of an "active prepaid card", so it's possible but highly unlikely, that prepaid cards mean any card that have time on them (whether that time has been used or not.) Since Blizzard does not have immediate access to Point of Sale data, they would base "active prepaid cards" on restocking quantities. If a retailer requires 50 new cards, for example, it's assumed that there are now 50 "active prepaid cards" in the hands of Subscribers, and those would therefore count as a Subscriber.
- Since the game is free in China and it's inexpensive to play, many users have multiple accounts and multiple pre-paid cards. 1 player might represent 2-4 Subscribers for example. These are all considered active pre-paid cards as long as there are still minutes on it.
- Pre-paid cards in China hold minutes (2000 minutes or 4000 minutes for example). If the cards still have 5 minutes left on them, and they are thrown out, that still counts as an Active Subscription. Until the card is completely expired (used up), then it counts as an active prepaid card as per Blizzard's definition.
- There are players worldwide who have multiple accounts (multi-boxing), not a large amount but it exists. One player might represent 8 Subscribers for example.
- "Internet Game Room" players are an interesting scenario. In China for example, players can goto "Game Rooms" (e.g. gaming cafés instead of playing from home) where they can pay for 1-2 hours of gameplay for example. Any user who plays within the 30 day period is counted as an official subscriber. So, you could login to WoW from a Game Room, play for 1 hour to check mail or guild activities and then not sign-in again for an entire month - that still counts as an Active Subscription. Imagine the sheer number of unique logins from Game Rooms? "Game Rooms" as Blizzard calls them are extremely popular in China.
Not Done Yet
Since this entire thing is going to get pretty big, I'm going to divide it up into parts.
From what you've seen so far though, you can probably guess that there is indeed a big difference between what you view as a subscriber (real player? one account?) and what Blizzard defines as a Subscriber.
Next up, I'll show you what Blizzard has said about Subscribers around the world, how many real players are in China, and how many are estimated to be in North America and Europe.
In NA/EU, Blizzard does not publish actual player counts (versus Subscribers). However, in China the WoW Servers are run by NetEase and they do publish actual player counts. I think you'll be surprised by the results.
I wonder if there are any websites out there that keep a running total of leaked games? (Well, major game releases anyways.)
Here are the big ones I can remember, including these 2 most recent ones:
Killzone 3 (PS3) - Feb 12, 2011
Crysis 2 (PC) - Feb 11, 2011
COD Black Ops (X360) - Oct 18, 2010
Halo: Reach (X360) - Aug 20, 2010
Halo 3 ODST (X360) - Aug 28, 2009
The Sims 3 (PC) - May 18, 2009
Spore (PC) - Aug 31, 2008
GTA IV (X360) - Apr 23, 2008
Halo 3 (X360) - Sep 21, 2007
GTA San Andreas (PS2) - Oct 20, 2004
Halo 2 (XBOX) - Oct 13, 2004
World of Warcraft (PC) - Jan 4, 2004
Gran Turismo 4 (PS2) - Nov 2, 2003
DOOM 3 (PC) - Nov 3, 2002
Half-Life 2 (PC) - Oct 2, 2003
Crytek is blaming pirates for the leak. Which is weird.. wasn't it leaked by an internal (possibly disgruntled) employee? Pirates can be blamed for spreading the game, but not for leaking it in the first place.
Did Blizzard blame piracy when a member of their senior management team leaked confidential subscriber data, financials, and product release slates? Even Carmack had it right when the DOOM 3 leak happened, he blamed ATI for the leak - not pirates.
Did you noticed something else about these leaked games by the way?
They went on to experience HUGE commercial success and awards. These were top selling games worldwide! Even after the GTA IV console version was leaked, it still went on to sell what.. over 19 million copies by now? Doom 3, GT4, The Sims, Half-life 2, Halo.. these are the best selling games in the world.
The only trend I see here is that leaked games go on to become hugely popular and financially successful games.
If the game is bad, though, you can just blame piracy for the bad sales. Right, Crytek?
Game leaks, and other forms of media, have been pretty common for decades actually. I'm sure some news sites may try to link the current popularity of Wikileaks to leaks like this.. "the reason for these leaks is because we live in a Wikileaks generation." Even though that has nothing to do with it.
Now that I'm thinking back, what other kind of stuff has been leaked? I remember the X-Men/Hulk/Harry Potter films being leaked.. DVD screeners, tons of TV series leaked months in advance. Haha.. ah yes, I remember when the pilot episode of Doctor Who was leaked back in 2005 and BBC spokeswoman Annie Frederick claimed that it wasn't a publicity stunt and that the leaked version was completely different from their final version - apparently, the leaked version was unfinished and an incomplete rough cut. After the "real" version came out, it was discovered that the only real difference between the two was the intro music. Good one Annie. Since viral marketing was all the rage back in '05, her claims led me to believe that it probably WAS a publicity stunt after all.
Anyways.. for other (real) internal leaks, it just goes to show that employers need to improve their hiring standards and screening processes, and once they get good employees, they need to treat them very well. There's nothing worse than a disgruntled employee, they'll end up costing you more than any possible revenue they could have made for you. (Things like weak performance, theft, sabotage, influencing other employees, complaining about your company and changing perceptions, creating hostile work environments, security issues, being a bottleneck, and vandalism tend to cost companies a lot of money.)
* UPDATE:
Here were some other games that leaked. If I find any others, I'll keep updating. (If you know of any, just leave a comment and I'll update/change dates too if I got one wrong.)
Assassin's Creed (PC) - Feb 24, 2008 (Ubisoft blamed the disc-replication firm Optical Experts Manufacturing for the leak.)
Fable 2 (X360) - Oct 16, 2008
Fallout 3 (X360) - Oct 8, 2008 (Bethesda blamed the press for leaking a review copy.)
Far Cry 2 (X360) - Oct 16, 2008
Manhunt 2 - Sep 7, 2007 (Sony Europe employee was blamed and later fired.)
Painkiller (PC) - Dec 31, 2003
S.T.A.L.K.E.R. (PC) - Dec 30, 2003
S.T.A.L.K.E.R. Call of Pripyat (PC) - Sep 28, 2009
World of Warcraft Expansion Packs (Alpha TBC, WOTLK, CATA months before release.)
I think Manhunt 2 is proof alone that controversy and game-leaks can sell even bad games well. Manhunt 2 sold 500,000 more copies than they should have. =]
* UPDATE (02/18/2011):
Bulletstorm (X360) has apparently been "leaked", but in fact the street date was just broken by a retailer. So it's not really considered a leak, but it has been out for 4 days before retail.
* UPDATE (03/06/2011):
Homefront (PC) - Mar 6, 2011 (Kaos Studios indirectly blamed Steam)
* UPDATE (11/07/2011):
Battlefield 3 (PC) - Leaked Oct 16, Release Date: Oct 28
Gears of War 3 (XBOX360) - Leaked Jul 2, Release Date: Sep 20
Batman: Arkham City (XBOX360) - Leaked Oct 13, Release Date: Nov 21
RAGE (XBOX360 / PS3) - Leaked Sep 30, Release Date: Oct 4
Deus Ex: Human Revolution (X360 / PS3) - Leaked Aug 18, Release Date: Aug 23 (Retailer broke street date.)
Resistance 3 (PS3) - Leaked Aug 20, Release Date: Sep 6
The Witcher 2 (PC) - Leaked May 8, Release Date: May 17
Call of Duty: Modern Warfare 3 (X360) - Leaked Oct 25, Release Date: Nov 8
Mass Effect 3 (X360) - Leaked Nov 5 2011, Release Date: Mar 6, 2012 (Bioware blames MS XBOX Live "Human Error")
An interesting read on 1UP following those most recent Take Two announcements.
Oh! A while back, Pachter also predicted GTA 5 would be released in 2010. =]
Here were some of his other past predictions/comments:
1. Mafia II unlikely to be profitable and would not recoup the cost of game development. (Source)
Take Two still has DLC development underway for Mafia II this year. On Steam, it was in the Top 10 of best selling games for 2010 (even though Mafia II came out just 4 months before the end of the year.)
2. Starcraft 2 would sell 6 million units on opening day, however it ended up being 1.5 million instead. (By the end of the first month, they sold 3 million units worldwide and by end of February they sold 4.5 million units.)
After first day sales were released by Blizzard, he adjusted his prediction and made it 5 million total copies within 2-3 months. However, it took 7 months before it reached 4.5 million units. =]
I didn't think it was really fair though, making such a big change after seeing all of the sales details. Not that it helped either though..
3. He was highly skeptical of Red Dead Redemption, and stated that it was not blockbuster material.
4. On a related note, he's predicting that LA Noire will be comparable in sales and just as successful as Red Dead Redemption (after he apologized for underestimating RDR in the first place.)
Personally, I have my doubts about LA Noire though.. it looks like a great game, but it's like releasing a natural disaster movie in the middle of vampire film/TV season. I've been watching Google/Twitter trends and monitoring other various patterns, but there's very little interest in LA Noire so far (especially in comparison to other big name releases coming out during the same period).
I'm worried that the game might end up being too linear or storyline driven, with most of the game taking place with cut scenes and NPC interaction (after all, they spent a lot of time and money on voice acting and facial imaging). If there are too many NPC conversations and detective mini-games that are impossible to lose, it takes away from the freedom and excitement of exploring in RDR and GTA. What if there's little in the way of side quests (errr... "investigations") and everything in the game is based on one single storyline or one investigation after the other? Then it just becomes an interactive visual novel.
The other thing that worries me is that people are mostly talking about how "great it looks", but very little about the actual gameplay, story, characters or music. I still think first month's hypetastic sales will be very good mind you (I'm predicting ~2.5 million worldwide for the first month), but I don't think they will be as high as RDR in the long run. I also think Michael Pachter will be left disappointed with his own sales predictions. =] Hopefully, Team Bondi is developing a very stable game too.. this game is going to make or break them.
5. He thought that Ubisoft DRM was a good idea.
6. He said Borderlands would fail badly.
7. He's predicting that the Activision-Bungie game will sell 10-15 million on opening day. I'll have to come back to this one at a much later date.
His predictions, of course, have a lot of IF's and BUT's. IF it's as popular as Halo, IF it's released on multiple platforms, IF they have a good sales day.. blah blah blah. Come on, where's the hard science and figures?
From what I've seen online, he wasn't too bad going back 5-7+ years ago but things have gone very bad for him since late 2008. So, it's probably a good idea for him to leave his predictions more open ended like he did with the Activision/Bungie unannounced game.