Of Game Servers

The Server Dilemma

This week there has been a bit of a war waging in the ArcheAge community between two factions.  The first faction wants them to open new servers so that they can play the game and not have to wait in 10+ hour long queues.  The second faction doesn’t want new servers to be opened at any cost, because it serves to dilute the community, and in a game where land ownership is a crucial part of the experience there is no real way to merge the servers.  I feel like this is one of those arguments when both sides have equal merit.  Generally speaking when a company spins up new servers to help cover the launch weekend users, they also end up having a bunch of completely dead servers months later that have to be merged.

The ultimate problem with new servers, is that we want to play with our friends.  By nature an MMO is a social experience, and at this point most of us bring with us a large group of friends and acquaintances that we want to be in close proximity to.  As such lots of players will roll on another server only to get the feeling of actually playing the game they crave, but ultimately later on re-roll once the opening weekend crush is over to be in the same place as their friend.  So I feel like new servers by nature are a damned solution to the opening weekend problem.  Any servers that spin up are by nature less valuable to established players than the original servers.  They will have less of an economy, and less resources to run any group content.  Having done this multiple times, there is a certain pride in playing on a “day one” server.

Abolishing the Server

For a long time I thought the solution to this problem was to simply do away with the server infrastructure entirely.  Elder Scrolls Online launched with a Mega Server infrastructure, where instead of pool resources into smaller named and classified clusters of servers… they just dumped all of them together in one of two larger data centers US and EU.  This allowed them to simply add more nodes to the cluster to gently smooth out the server load without actually interrupting the users experience.  I have to say after the what has to be hundreds of game launches I have experienced…  this concept made for one of the absolute smoothest I have seen.  We were able to get in and play without needing to fuss about what server our guild was going to be on.  It solved so many issues and let us get in and play with our pool of players.

The problem there however is that as the population dwindled and our regularly nomadic guild did what they were good at… and moved on to the next big thing, we were left without much of a community.  The only intimacy we had in that game was that which we brought with us.  Coming back to Final Fantasy XIV it has refreshed my understanding of just how important that server community is to the overall feel of a game.  The night we bought player housing, we had various folks from different guilds popping by our house and welcoming us to the neighborhood.  Our housing district has not one but two dedicated linkshells for communicating with “neighbors” and after months of doing hunts and fates we’ve joined other really social linkshells that flesh out the rich community feel that I had been craving.  So while the Mega Server answers one question, I feel like it comes up lacking in the social aspect greatly.

A Transitional Population

The Rift launch looked like pretty much every other game launch in memory.  I compared a game launch to a natural disaster the other day, because really it is very similar.  There is this crashing wave of users that hit the servers, and that which they don’t break they flood to the point of bursting.  After the storm clouds cleared they were left with the same problem as always.  Some servers had massive queue times, whereas the others where relatively dead.  World of Warcraft tried to solve this problem by allowing players free character moves between overpopulated servers to very specific under populated ones… and as a result a bunch of thriving communities were born out of the leftovers of higher populated ones.  We would not have the Scryers or Wyrmrest Accord servers were it not for this process, so it feels like it mostly worked well.

Rift however went a step beyond this in allowing any character to transfer to another server for free.  There were some restrictions initially about transferring from PVE to PVP, but over time these went away as they patched in the “Faction as Fiction” concept.  This allowed players to self sort, and overall it seems to have worked.  I’ve moved around quite a bit during the time since this went live in that I started on Shadefallen a server that was whisked away in a great depopulation of servers.  From there I moved to Faeblight, and then to Deepwood to hang out with Liore and the Machiavelli’s Cats…  and then back to Faeblight again because I missed being on a Role-Playing server.

This solved the problem of being fettered to a server and having your friends disappear on you.  You could freely transfer elsewhere to play with a different pool of players.  One of the big problems I have with World of Warcraft right now is that my friends are so splintered across so many different servers.  While you can do cross server grouping to an extent, it just isn’t quite the same as being on the same server and in the same guild.  At $25 a pop it would simply be cost prohibitive to move my army around to another location, even though Argent Dawn where we are currently has seen it’s better days.  With the Rift situation it becomes relatively easy to move your guild to a new location and set up shop again on more favorable shores.  The big thing that Rift is missing however is an equivalent to the BattleTag system that gives you a simply way to connect all of your friends with one single ID.

The Hybrid Approach

I feel like maybe there is a hybrid to these options out there that we just haven’t seen yet.  I really like the concept of the mega server in that it removes the work of even having to discuss what server everyone will be located on.  It also opens things up.. so if you meet someone that plays the same game as you…  there is never that moments later heartbreak when you realize you are on different factions or different servers.  What I think we are missing is the concept that Elder Scrolls Online talked about when they were launching, that never quite panned out.  Originally the idea was that while everyone would be on a mega server, we would be able to self sort into pools of players with like interests.  This would apply an almost dating website series of questions to identify what types of goals you are focused on when you play an MMO.  Then through these questions you would be sorted into the cohesive community that best represents your interests.

So when you are wandering around the Capital city, you are seeing the players that represent your tastes in MMOs.  If you like being around role-players and crafters, you would end up being shuffled into a very collaborative environment.  If you like competing with other players and doing battle to determine who has the most skills… then you would be shuffled into a very competitive focused environment.  I feel like a scenario like this would give me some of the permanence from the community that I found I craved when I came back to Final Fantasy XIV.  I like being involved with the community just outside of my guild.  Guild is a comfortable home for me, but it isn’t the end of my universe, and I crave interaction of new ideas and new players.  I feel like this self sorting mechanic would allow for both seamlessly.    You could have different types of players in the same guild, but self sorting into their own little “perfect community”.

The Solution

Essentially I have seen a lot of solutions for the same problem, and each of them has fallen short.  That doesn’t mean that we should just declare that there is no right answer to the problem and fall back to the old tried and true server structure.  I feel like traditional servers are going to be a thing of the past, but we need to find ways of carrying the best aspects of that intimate server community forward with us.  I am interested in seeing how people solve the problem moving forward.  Games like Landmark and eventually Everquest Next will be tasked to solve it, and with Landmark they are already moving in what feels like the right direction.  The ability to move between servers freely gives you the ability to play with anyone also playing the game.  The permanence of an individual island however gives you a localized community feel.  Do I think it is perfect?  No, but I think it is a step in the right direction, and I hope more games try doing something new instead of falling back on the pitfalls of the server.

6 thoughts on “Of Game Servers”

  1. Im not sure why anyone thinks Megaservers make it difficult to build community? Do you mean that it’s too many people to build intimate connections? Because if thats the case, then we’re just talking about social tools, not megaserver tech. Players just need a reason to interact and that doesnt change because of megaservers.

    AA’s current situation is embarrassing. Theres not any good excuses for their current situation. This isn’t the first MMO launch, not even the first MMO with land and other features that complicate server flexibility. Theyre simply unprepared for deliberate reasons. There’s just no way they didn’t know what they needed for a smooth launch.

    It’s worse that people who shelled out hundreds of dollars to support development are reporting not getting that 4 day advantage they were promised. That’s a serious charge.

    Should AA have gone Megaserver? I don’t see how this wasn’t mandatory given the kind of features it has. You need a vast server community that’s STABLE. And you can’t have that when your system is as inflexible as the one they’ve adopted. I think they’re sinking their own ship right now.

  2. I agree. Megaservers are a good approach, but they become problematic when it comes to building community. I think they’d work better if most MMORPGs weren’t designed with so many anti-community (or at least not pro-community) features.

    For example, matchmaking as your primary way to play with others doesn’t necessarily generate any friends. I’d like to see more games promote keeping the same group through matchmade content, perhaps with a bonus. Also, more systems designed around reputation and making matchmade friends.

    For example, take the hybrid approach you mention, but let the system do some of the work. People you like? Rate them positively and they’ll be more likely to show up in your world. “Favorite” them and they will definitely show up in your world, as well as show up in a special tab of ‘not quite’ friends, but still people you could start a group with at any time. It could even take into account their friends, favorites, and people they’ve rated highly to keep a large web going.

    From there, maybe guilds could mark one another as allies and rivals? I dunno. There’s room to play with it, at least, and it is at least some attempt to bridge those gaps and create those communities!

  3. I’m not sure ESO by itself is enough of a sample size to determine whether community “feels” are possible in a mega-server setup. TSW seems to prove the opposite, though I’m sure no MMO at launch is hoping to be as narrowly focused on a niche playerbase as TSW has become. Megaservers seem to be what WOW is inching towards, and is an example of the odd situation WOW often finds itself in, with its lingering success actually working against it. WOW seems more and more like a giant beta for Blizzard’s next MMO, but I digress. Community seems like something that can be built on a megaserver with a little more time. True, many players wont have the patience required, but if you’re relying on them to build your community you’re in trouble no matter how much time is involved.

Comments are closed.