The Daily Grind: Queuing or merging?
Provisioning an MMO launch is hard. All kinds of hard, actually. For starters it involves making guesses about how many of us -- you know, the gamers -- are actually going to stump up for the game and get in on launch day, and how many of us will be there over, say, the several weeks. There are just so many factors to consider. How will the reviews come out? What about word of mouth? What about an unexpected major bug? What if some aspect sucks badly and needs an overhaul? And what if we all actually love the game as much as the developers hope, and we pile on like mad?
So, the game operators do their best to set up the right number of servers and provision them with bandwidth. Each server setup costs money, even if that server isn't actually even switched on at launch, and it keeps costing money, whether or not it ever gets turned on until the operator gets rid of it. Also populations for a server can be too high or too low for the game design.
If the operator picks too low you have login queues, or overloaded servers. If they pick too high, then down the track we'll have some awkward server merges later on. Oftentimes both of these are considered to indicate a 'failed launch'. Which would you (as a gamer) prefer, considering it is likely to be impossible to actually get the number exactly right: Too many servers, with server merges to come later on? Or too few and potential congestion and login queues?
