World Class Villains

The core metric for assessing the size of a MMO is the number of active players. It’s a simple metric that is easy to understand and compare, even if it isn’t perfect (after all, revenue and  profitability are probably more important regardless of player base size). It’s power is such that good numbers are trumpeted from the rooftops while poor numbers are hidden or even unannounced. “How many players?” is a question synonymous with determining success in MMO terms.

Going back, NCsoft used to do a fantastic thing and release their quarterly player figures along side their quarterly financials. This practice waned – Tabula Rasa might have had something to do with it – so that now all that is presented are quarterly financials for each game. Fortunately, that’s enough to take a stab at how many players City of Heroes / Villains currently has.

UPDATE (15 Feb 10): I’ve been told by Johnny Velocity that another possible reason for the omission of active account numbers from NCsoft’s financials was the release of Aion in Korea. Stockholders were keen on finding out if Aion was cannibalising players from Lineage I and II or if it was attracting new players to NCsoft titles. This was probably something that NCsoft would likely see as sensitive information. Also, only showing revenue makes it hard to work out how much of a contribution microtrans is having to a title’s revenue, which could be extremely useful competitive information about what works and what doesn’t with a sub / microtrans model. Personally I believe that TR did have an impact – it was the first western title that NCsoft was very hesitant in releasing player numbers for – but the above also makes sense. Ultimately it appears that NCsoft found releasing these kinds of numbers more of a burden than a boon and discontinued the practice.

I originally started writing this up in January, had a version crash and lost lots of work. Now that the Q4 2009 figures are out, it paints a really interesting picture of CoH/V.

Who Cares?

CoH/V players do. It was recently discussed on the forums.

I do because I find it an interesting question. CoH/V is nearly 6 years old with an expansion on the way in a market full of increasing competition (and not just Champions Online – any new MMO potentially pulls players away from CoH/V). The analysis was also interesting because it showed me some things about MMO subs I’d never thought of.

Players != Accounts

When people talk about the number of active players in a game, they are really talking about the number of active accounts. Players may have multiple accounts, after all. I’ve tried to avoid this error, but it’s a hard trap to break out of because the terms are often used so interchangeably.

The Stuff I Can’t Control

It’s hard to do a strict “game revenue = number of active accounts” because a number of things go into gaming revenue that can’t be separated out. These items include:

• Box costs or paid content releases – new players have to spend more to buy the box. This throws any kind of player number modelling to a degree, especially during the release of expansions or paid content. In CoH/V’s case, this includes key points in time like the release of CoH and then CoV, sales of additional boxes like the Mission Architect box and the Good vs Evil Edition and also releases like the $10 Booster Packs or Valentine’s Day Pack. Booster packs are coming out about once a quarter and no doubt help boost CoH/V’s revenue.
• In-game advertising revenue – CoH/V was a bit of a pioneer in announcing real in-game advertising that someone was no doubt paying for. The reality is that there might not be any real in-game ads left in CoH/V, but at some point money was coming through this channel.
• Charges for extra character slots, rename tokens, etc – can’t be tracked, but they no doubt bring in some extra dollars.
• Exchange rate – probably the biggest variable, since the figures NCsoft reports are in million Won and CoH/V players pay in $US (or European money, but I’m assuming those currencies don’t dominate). Exchange rates are tricky things – depending on when the money is converted, wild variations can occur (and the exchange rate between Won and US$ did fluctuate over the period tracked). For simplicity’s sake I picked the exchange rate between Won and $US on the final 28th day of the quarter. Why? Every month has a 28th day, it’s close to the end of the month and it seems as good a point as any (and where an exchange rate wasn’t available for the 28th, I’d take the closest date to that). I also looked at a comparison of this method versus using a 3 month average (as provided by TonyV’s data) and found that at most the figures differed by 5% and on average showed no difference at all. Historical exchange rates were taken from here.
• The ‘bulk’ purchase of game time. If you buy 3-month, 6-month or 12-month subscriptions then you aren’t paying $15 a month and are instead paying a lot more in one period, meaning that an account overpays for one period and underpays for others. This adds a level of variation that is impossible to determine on a quarter by quarter basis, but for the purposes of the analysis it will be assumed that the distribution of these account payments are spread equally over the course of the year.

It’s Chartastic!

Here’s the figures and analysis for those who wish to look at the raw data. NCsoft stopped reporting CoH/V’s player numbers after Q3 2008 (oh, and that’s full year quarters Jan – Dec, not financial year quarters Jul – Jun). Concurrent player numbers got one more release before they were removed as well.

First chart up is CoH/V’s revenue in million Won, taken straight from the financials (click to enlarge).

CoH/V's Revenue from Q2 2004 to Q4 2009 in million Won

Key revenue spikes are Q2 2004 (when CoH launched), Q4 2004 (an Xmas sales bounce?), Q4 2005 (when CoV launched) and Q3 2006 (might be a combination of the release of Issue 7 update and the Good and Evil box). However, since then revenue has been much more flat, with CoH/V showing slight revenue growth from Q4 2007 to about early 2009. That’s a good sign, right?

Problem is that this growth in revenue in M Won disappears when you convert the revenue to $US. It appears that the $US weakened against the Won over this time period, meaning that every dollar CoH/V earned was worth more Won. Here’s another chart of the CoH/V revenue in M Won and $US (no current amounts shown to keep it cleaner – click to enlarge):

CoH/V's revenue in both million Won and estimated million $US from Q2 2004 to Q4 2009.

What this shows is that although CoH/V is bringing in the Won, it’s doing so off a shrinking $US revenue (my assumptions on currency conversion being open to criticism here). The exchange rate between the US and South Korea has meant that decline in US revenue has been hidden to casual observers. Personally I’d not thought of MMO revenue as being at risk to currency fluctuation, but there you go.

When you match up the $US figures and the active subscription numbers, this declining revenue makes sense (click to enlarge):

CoH/V's active subscription numbers and estimated million $US revenue Q2 2004 to Q4 2009.

The last reported official active subscribed accounts has been showing a downward trend since the release of City of Villains. It would appear that recent initiatives haven’t been able to reverse that trend.

Other Stuff to Bore People With At Dinner Parties

Some other interesting things that came out of the analysis was:

• how much the earnings per subscribed account can vary. Players typically talk about paying $15 a month ($45 a quarter) but that’s not exactly true. There is an earnings lag that occurs after a box launch as players cash in their free month – revenue can be collected in one quarter but the free month is used in another quarter. The other interesting thing is that players aren’t necessarily paying $15 a month. Revenue figures include more than just subscription revenue, so you would expect to see earnings per subscribed account (EPSA) to be higher than $45 a quarter. It isn’t. All I can assume is that there are a significant proportion of CoH/V’s player base who are bulk purchasing play time – CoH/V players might be paying as low as $11.95 a month for an account (or about $36 a quarter), which suggests that although having a loyal, long-term player base is great for a MMO, in revenue terms you still need the players who are just paying month-to-month for maximum returns.
• but even with that in mind, dividing quarterly revenue by $45 still gives an acceptable estimate of the number of active accounts. Using this method will tend to underestimate player numbers (barring box releases) by about 10%, at least in CoH/V’s case. That’s not bad for estimation purposes.

CoH/V's earnings per subscribed account, Q2 2004 to Q4 2008.

• on average, the highest concurrent account figures from a quarter represented 13.4% of CoH/V’s overall active accounts. I’ve seen the ‘1 in 10’ rule used before to try to estimate such things and in CoH/V’s case it appears to be within that range.

Accuracy, Or I Meant To Do That

So, how accurate are the numbers? The ones taken off the official financials should be golden. Ones that involve some kind of calculation are open to being questioned.

One thing that gives me some indication that the revenue conversions are about right is that Cryptic indicated they’d earned over $100m from CoH/V for NCsoft. Adding my estimated $US million revenue figures from CoH’s launch to Q3 2007 (the actual sale to NCsoft went through in Q4 2007) it comes to a shade under $108 million – at the very least, it’s consistent with an external data point.

All Of Which Brings Us To…

Q4 2009. In which several things of note are evident:

• CoH/V sees a significant drop in revenue earned. Revenue was 3927 million Won – approximately US$3.36 m – which is the lowest ever recorded quarterly revenue figure for CoH/V. There have been concerns that September 2009 could have been a damaging period for CoH/V, with the release of Champions Online, Fallen Earth and Aion and these Q4 2009 figures seem to reflect that yes, damage occurred.
• This result sees CoH/V earn 22 909 million Won for 2009, which falls short of the 25 000 million Won that was set as CoH/V’s revenue target in the Q4 2008 Earnings Release.
• CoH/V’s revenue target for 2010 is 19 000 million Won.

For Paragon Studios, none of these things are good. If we use the (flaws noted) method of dividing CoH/V’s quarterly revenue by $45, this would indicate a total of 74 708 active subscribed accounts for this title. This figure is likely an underestimate, but it would point to CoH/V having fallen to its lowest active subscriber point in the game’s history.

How many active subscriptions does CoH/V have anyway? Less than 90 000 in my estimation. The number of active players (players, not accounts) would be less than that, especially if multiple accounts and zombie accounts (accounts that are all paid up but haven’t been used for an extended period) are considered. Zombie accounts are still revenue sources, but they come from players who have possibly forgotten they are subscribed to CoH/V or are only paying to keep the account active for veteran rewards. According to the very experienced Victor Wachter, ‘passive accounts’ can generate significant revenue for a MMO. Multiple accounts were more common to CoH/V in the days when players couldn’t buy additional character slots, but they would still be part of CoH/V’s account base.

What Does This Mean For the Future of CoH/V?

Really, I think this entry is long enough. Speculation will occur next time.

DOOOOOOOOM?

Not necessarily. I’m sure that Paragon Studios have other initiatives in the fire.