Launch Day: How to deal with day one in the post-WoW MMO space

From an outsider’s perspective looking in, Aion’s launch seems to be having the same issue Warhammer Online had a year ago; dealing with the huge initial rush of players and tourists flooding the servers on day one. In the post-WoW MMO space, it’s more or less a given now that any majorly advertised AAA MMO is going to have to deal with this issue, and the current solution of server queues and character creation limitations seems like a rough deal for those who matter most; paying customers.

DarkFall, while certainly not on the user scale of Aion, dealt with the problem by limiting sales of the game for the first month or so, allowing new characters to enter the world at a set pace and re-distribute themselves before opening the doors once more. Current customers benefited at the expense of potential future customers. The luxury that Aventurine has over NCSoft is they don’t have to answer to other retailers, and so it’s up to them when to allow additional sales to open up. NCSoft can’t call up Gamestop or BestBuy and ask them to pull all copies of Aion off for the week because the servers are full. I’m also guessing, due to retailer contracts, they can’t start selling the game online-only, and then doing a ‘full’ release to brick-and-mortar stores later, but perhaps they could.

The problem itself is very easy to understand; the first day of any MMO is the day it will see the most users trying to get online at one time (unless the game is EVE or WoW and grows considerable after launch). You can’t just open enough servers to handle the first-day crowd because on day two it’s smaller, on week two it’s much smaller, and on month two it’s likely down 30%+ as the tourists move on. In any MMO, but especially one where population is critical (WAR and Aion), opening and closing servers left and right is asking for disaster. So what can be done?

For starters, we might as well end the ‘open beta’ charade and just call it paid pre-release for most games now, and since people in the beta have paid, why not get them on live servers? Unless you MMO gets a miracle patch on launch day (and if you do, you have more issues than just server pop), the end of beta is basically the launch-day game, so it might as well count. Allow players who pre-ordered before a given date two weeks or a month of play time on live servers, so that you mitigate the impact they will have once the boxed copy customers come online. Aion had 400k preorders, which is more than enough for 10 server (give or take), servers that can easily be avoided by boxed copy users looking for a truly fresh start. Guilds won’t have issues with surprise queues or character creation, members buying the boxed copy will still be able to play with their guild or friends, and overall you hopefully mitigate 50% of the launch-day crowd from all jamming in to see your game for the first time live.

The other solution would be to accept this trend, and have a plan ready for it. Go live with enough servers to handle all the day-one traffic, and then have tools in place to automatically merge servers as their population drops. The only real issue here would be with character and guild name overlap, as everything else in most of these games is static (this solution would obviously not work in DarkFall, as different guilds own different cities in different states of construction). Continue merging servers (quickly) until your population settles. The bonus here is you can merge one imbalanced server with (hopefully) the opposite imbalanced server, creating a better environment overall.

I prefer option one over option two, especially because the hardware to support option two might not be cheap depending on how your server clusters are arranged. The bonus with option one is that it further encourages pre-orders (perhaps you could even stagger that, so the earlier you pre-order the earlier you get in), which will give the company a better picture on what their games population might look like. Regardless of the choices made going forward, one thing is clear, and that’s that the current go-live model for AAA MMO’s is not working. It causes confusion and frustration among players and guilds, it creates imbalanced or over/under-populated server, and it leaves a poor first impression on everyone involved.

6 Responses to Launch Day: How to deal with day one in the post-WoW MMO space

  1. Snafzg says:

    Someone else said it really well:

    They could have launched with a hundred servers; the same top 10 would be packed and bogged by login queues, while the others languished.

    There really are a lot of low population servers for you to play Aion on. The problem lies with coordination.

    I don’t think Aion gave people enough warning or informed them properly for what took place. Had they done both, people (and legions) would have had backup plans in place.

  2. I like both of your ideas and I think either would work. I wouldn’t be bothered about queues if I wasn’t paying for it :) Either that or the devs should implement tools to allow for quick server deployment and merges etc.

  3. Malakili says:

    I know you aren’t a fan of Champions Online or heavy instancing in any game for that matter, but their server struture scales with the number of people and was actually really stable on launch. (it did have 1 almost full day of downtime, but that was due to a fuck up on their end with some guy messing something up with the patch files on the server and had nothing to do with the numbers of people.).

    The downside of instanced zones are obvious and I agree with the criticisms of it mostly, but it does deal with this particular problem rather well. Though, I honestly don’t know what their pre-order numbers of launch numbers worth, and if they were dealing with the same types of numbers of people that were flooding the Aion servers.

  4. Anecdotally, I have had no queues on Vaizel which is a West coast timed server. I haven’t been prowling the forums to see if the queues were all on East coast timed servers or not.

    It could be overall low population. There are plenty of Asmodian on Vaizel, but since I’m not yet at lvl 25, I’m not sure if there are any Elyos. ;)

  5. Saylah says:

    Vaizel is in deed low population but no longer accepting Asmo characters. The problem lies in poor communication/expectations and no way to move groups across servers. Our guild tried to bail on the high pop server to find that only a small fraction were able to create toons on the low pop server of the right fraction. Where does that leave us now? Uh, your guess is as good as mine. People on the high pop don’t want to continue with those toons if they’re going to leave them yet, they’re unable to join the guild on a low pop server. What a mess. It is the East coast servers that are all high pop. But as I’ve mentioned, low pop doesn’t mean you’ll be allowed to create the faction of your choice either.

  6. skarbd says:

    The problem with rapid expansion/contraction, is the negative PR that goes with the contraction.

    With Aion, I would have probably switched off the private stores for the 1st month, to help the capacity issues.

    I am sure that people would have been understanding, especially when they had queued for a while.

Follow

Get every new post delivered to your Inbox.

Join 167 other followers

%d bloggers like this: