Analysis: FFXIV's Consolidation Plan

While completing my Master of Information degree at the University of Toronto iSchool, I have spent a good deal of time analyzing decision-making processes in organizations, from understanding why AOL and Time Warner merged (losing over $200 billion in the process) to understanding why NASA launched Challenger despite warnings from Morton-Thiokol (Kaboom!). It’s easy to look back in hindsight and see that these decisions were bad, but it’s far more important to apply those lessons to decisions that are happening right now and try to unravel them before they unravel the companies making them.

One such bombshell of an apparently bad decision was lobbed out of Square Enix’s office last Thursday (Feb 9, 2012). As I described in my recent feature article, Square Enix announced a round of World Mergers that would reduce the number of active servers for its MMORPG Final Fantasy XIV from 18 to 10. Their plan involved eliminating all the previous servers and letting active subscribers pick one of ten new servers to move to. Inactive players would be allocated randomly. Square Enix would also delete every Linkshell (in-game community, like a guild) and wipe every friends list and blacklist clean.

My reaction of shock and surprise was concisely summed up by three letters: W, T, and F. Why on earth would they do this? How could they possibly have decided this was the best course of action?

In the normal course of knowledge management studies, organizations would be open to researchers interviewing and analyzing their decision-making process, helping both the organization and other organizations to avoid such catastrophes. But the games industry is notoriously closed to academia. It seems to me that most game companies loathe the idea of smart, educated people coming in and helping them find better ways to do things. So normally the story would end here, with academics lobbing conjecture over walled gardens and filling in the blanks with other case studies and their own imaginations.

But thankfully, this story doesn’t end here.

In the days that followed this announcement, Naoki Yoshida, FFXIV’s replacement producer, made three posts on the official FFXIV forums, explaining the decision-making process behind the announcement. These posts, along with relevant details about FFXIV’s current state, might help to explain why such a seemingly ludicrous idea made it to official announcement status to begin with.

Final Fantasy XIV, Square Enix’s MMORPG sequel to the successful Final Fantasy XI, has struggled since launch with poor design choices and lack of content that upset many players. Nevertheless, the game kept running on indefinite free trial, hired a new producer, and swore it would turn things around. Then it started charging a monthly fee, despite the fact that the announced re-launch was more than a year away.

At the time, the decision to start charging a fee was controversial, but it made sense. By this time, most of the people playing FFXIV casually had drifted away (about 12,000 active players down from 600,000 copies sold), and Square Enix figured most active players would be willing to pay a reduced subscription fee to keep playing the game until the 2.0 overhaul. They were right (orange is the number of FFXIV logins during Japanese prime time pre-subs, red is post-subs).

(Caveat: player population numbers are player-generated estimates. However, unless Square Enix wants to provide data showing the contrary, I assume these numbers are pretty accurate.)

This is a crucial piece of information that framed Square Enix's plans throughout 2012. The fact that most players playing prior to the implementation of the subscription fees continued to play after subscription fees were required was interpreted by Square Enix as: "The only players that matter are the ones that are currently active subscribers." All players that had not subscribed were considered lost and no consideration was made for future players. Square Enix basically thought it was making a game for just those 12,000 people. As we will see, this act of sense-making directed producers of FFXIV to believe that their original plan, involving the destruction of all existing in-game communities, was the best course of action.

On February 10^th, the day after the major announcement, Yoshida explains his decision-making process. The decision to consolidate server populations in FFXIV were motivated by the goal of ensuring sufficient players would be available in each time zone to participate in group activities. Of particular interest was the desire to consolidate players in European time zones: the initial plan was designed to encourage all European players to select the New Server #10, in the hopes that this “unofficial” EU server would provide European players with the greatest opportunity to find others sharing their play schedule. The other 9 servers were to see a fairly even distribution of North American and Japanese players.

This desire to consolidate European players can be partially reflected by anecdotal evidence on the official forums. French users taking an unofficial poll during peak Euro hours of users flagged “FR” saw about ~60 players online, well below the ~650 per server seen by the Japanese polls graphed above. I can’t speak German, but I’ll assume their results were similar. More importantly, Square Enix would have actual login statistics to back up their decision; they could see exactly how many people were logging on at each time and from which region, probably prompting the EU population consolidation.

So they wanted a way to “extract” EU players from the other servers and place them in their own world, a perfectly sensible thing to do when population numbers are below the critical mass to ensure a fun gameplay experience. But why shake things up for all NA and JP players too? Yoshida provides his reasoning:

He also recognizes three drawbacks with his solution:

And here lies the key and cornerstone of this entire decision process: “Factor C also may not be as bad because we have no idea when these players will return to the game and if they do, it would be most likely with 2.0. With 2.0, we had plans to implement a World Transfer Service so we believed that should be able to address this problem.” says Yoshida, in his post discussing the cons with the original plan.

Yoshida clearly believed the currently unsubscribed players were a lost cause: they weren’t going to renew their subscription and play until 2.0. Instead, the game’s population would remain largely stable for the remaining year until the re-launch. This conclusion was probably reached by looking at the population numbers before and after implementing the subscription fee: because there was not a significant drop in players, they concluded that players who were not paying now were not interested in playing at all.

This conclusion, however, was wrong. Both active and inactive players expressed their anger at Square Enix’s decision to limit the choice of destination server to only active subscribers. This is because both groups of players had friends in the other group and neither group wanted to be separated. Players who were friends before the subscriptions were imposed wanted to know that they could still play with each other if both resubscribed at a later date.

Square Enix fell afoul of a decision-making fallacy that I see all too often in the games industry: they completely excluded the importance of non-paying players. A non-paying player might not be paying now, but might be paying later; however, if said player is randomly reassigned to a foreign server, they are far less likely to become paying later. Likewise, current players might be playing thinking their friends will come back and join later, but if their friends are taken away to another server, they might give up and stop playing too.

Ironically, this is the exact same problem that Square Enix formerly faced in FFXIV’s predecessor, Final Fantasy XI, where they had implemented a policy to DELETE characters that were left inactive for a year. Quite a few characters got the axe before they realized how bad it was for business (you try launching a “return to the game” campaign with a “oops, sorry we deleted your character” clause). They scrambled to build a character restoration service, but the stigma of character deletion still haunts the Final Fantasy Online brand. So how could they have made nearly the same mistake again?

On February 11^th, Yoshida posts again, this time acknowledging the community’s backlash against the original plan. He recognizes that most users are in favour of a plan that merges two servers into one in a more traditional manner, while also allowing players to optionally transfer off to another world. The key consideration is that, without actively choosing to move, players who once played together will stay together.

Unfortunately, this still does not address the issue of inactive subscribers being denied the option of choosing a server. For users who wanted to stay on the same server, this new formula allows them to stick together, but it conflicts with the original goal of consolidating EU players together on a new server. If there are two EU friends, one of which is active and the other inactive, the inactive player will not be able to move to the new EU-designated server, leaving the friend with a difficult choice.

At this point, there is no final decision to be made, because technical considerations might make this sort of partial merge impossible. There is also no word on the preservation of Linkshells or friends lists. Yoshida says that, for now, the official post will remain on the official FFXIV website.

On February 12^th, Square Enix takes down the official notice, preventing users (such as yourselves) from reading it. Thankfully, copies remain on fansites.

On February 13^th, Yoshida makes a third post, proposing an alternate server merger system in response to feedback from players. The post largely reiterates what was said before about the goals and methods for server transfers, but presents radically different methodology and technical constraints than those which were presented in earlier posts.

With the Feb 13 plan, the existing 18 worlds will be directly merged into 9 new worlds and a new 10^th world will be created for European players. This is basically exactly what players were asking for and what most of us thought the obvious decision was in the first place.

Yoshida also changes his stance on the Linkshell and friends list deletion issue. He states that new testing shows that it is possible to preserve these things, and that this preservation also carries to inactive (non-subscribing) players, ensuring that communities will not be deleted following the merger.

How can such a radical change come about over a weekend? If it was at all possible to preserve these things, wouldn’t there have been a more thorough look into a server merger that would not arbitrarily wipe these key cornerstones of community from the entire game? Or was it tactically important to suggest that it was not possible to preserve them at all in order to bolster the original case for the scatter-merger plan?

Yoshida does not that it is only possible for users staying on the “destination” server for two merged servers to keep their Linkshells and friends lists. Users who opt to transfer to another server, such as the new EU-timezone server, will still face a total Linkshell and friends list wipe, just like the general population did in the original plan.

With that in mind, will people transfer over to World 10? Or will this new world simply be even more underpopulated? Yoshida does recognize this possibility, and states that further transfers and mergers are possible, especially if World 10 fails to meet expectations. Which prompts me to ask: why make a world 10 in the first place? Why not pick the two servers with the most active EU players, merge them, and designate them as the unofficial EU server? Or would that result in an overpopulated server?

Sadly, without proper data, external analysts who could otherwise propose viable solutions for these server consolidation woes can do little but propose vague suggestions. Perhaps the way forward for Square Enix would be to reveal some concrete population data and allow users to make informed decisions about the future of FFXIV, rather than forcing both producers and players to continue fumbling about in the dark.

Of course, there had to be one last slam against players who are currently unsubscribed from the game: originally, in the case of character name conflicts, the oldest (first created) name would prevail. Now, active subscribers prevail over inactive ones, so users who had played the game since launch day could come back to a forced re-name, their character’s name stolen by an opportunistic subscriber who watched the announcement of the two worlds being merged and created a character with the same name on the other world. That’s another one of those important life lessons Square Enix should remember: players will exploit and abuse any opportunity to screw over other users.

It’s amazing that such a simple thing such as merging servers could go so wrong for an operator of an MMORPG. Contrast this with, for example, Rift, which recently underwent a series of “soft mergers” by turning underpopulated servers into “trial” servers and encouraging players there to use the game’s free character transfer service to jump to a more populated world.

Would things have gone differently for Square Enix with a Knowledge Manager on their team? Most assuredly. Will they, and other game companies, avoid these sorts of mistakes again in the future? I suppose that depends on whether or not they learn the value a good KM can bring to a team.

Further Reading: The Knowing Organization by Chun Wei Choo