Wednesday, December 21, 2011

Population Distribution Impact of SWTOR Staggered Access

I was debating whether or not to post about the SWTOR staggered launch plan, since I personally chose to sit this one out due to computer performance issues (more my fault than Bioware's).  MMO-Gamer Chick's questions convinced me to go ahead and post anyway, mostly because the comment I was writing on her blog was starting to look more like a blog post. 

To lead off with my main point, I wonder if Bioware has taken a long-term risk to alleviate a short term problem. 

The Background
For any readers who have not heard about this controversy, Bioware elected to admit players into the "head start" based on when they pre-ordered the game.  Historically, MMO's have gotten absolutely hammered on launch day when everyone shows up at once, and I suppose that dedication to the game - as demonstrated by pre-ordering early - is as good a criteria as any if you are going to limit access in the hopes of improving the experience. 

(It's entirely possible that they intentionally did not go out of their way to let "late" pre-orders know that they might be sitting out the better part of the week, but I don't know that this diminishes the fairness of the system.) 

One consistent complaint players have expressed was the uncertainty about when in the week of head-start time their personal invite would arrive.  To ensure maximum flexibility to adjust the rate of invites based on server performance, players had very little indication on when they could plan to start playing.  Personally, my guess is that people were more displeased that they personally weren't in the earliest invite cohorts than about the uncertainty - how many folks really would have been happy if Bioware had said on day 1 "fine, you're not getting in til day 5 of 7, but here's the precise time your invite will arrive"? 

Population balance going forward
All of that said, I don't think that any of the above will matter in a week, and I'm already seeing people who were very displeased with the roll-out while they were awaiting their invite happily posting about the game.  (Indeed, reaction amongst people who were on the fence about the game, myself included, who actually tried it has been almost universally pleasantly surprised.)  The real issue which addresses MMOGC's question about whether other companies will attempt the same approach is the effect on populations going forward.

Figuring out how many servers to offer at launch is probably the biggest challenge a MMO studio has at launch.  Launch with too many servers and you will end up with low populations, potentially forcing server merges (which are nigh universally regarded as a devastating vote of non-confidence for a new game).  The opposite issue is less obvious but harder to fix.  Because MMO's are social games, many players don't want any random server (and the ones who do are less likely to have social ties that will keep them around longterm to form a community).  Players want to be on the server their friends are already playing on, and no amount of additional servers added after launch can fix that population issue if the game launched with too few servers. 

In some ways, Bioware seems well prepared to know what to expect - the aggressive push for pre-orders, combined with a social networking guild setup system, gave Bioware some numbers. However, I wonder if the staggered access cut them off from the real-time data about how many people who for whatever reason did not pre-register with a guild were eventually going to join up with that guild once they got in.  Under a free-for-all system, Bioware might have seen on day 1 that certain servers would have been a problem.  Because of the staggering, it might not have been apparent until day 4-5 that specific servers were overcrowded.

As of this writing, the official server list has 72 US (EST/PST) servers that are not labeled "new" and 51 servers that are labeled "new" as of either 12/16 or 12/20.  And, sure enough, we have dire warnings that just a handful of servers are so overcrowded that players can expect queues for the indefinite future.  Unlike the pre-order access, which is already old news, this balance issue may remain with the game for months or even years to come (much as it did on my WoW server, which was one of the original 40 pre-announced for WoW's launch day and which ended up needing almost three years to recover).  Other studios will be watching, and their conclusion may be that it's better to have some short term pain. 

1 comment:

Anonymous said...

US Servers are full, heavy, medium and light in roughly equal numbers. Why?

* Servers were brought up in waves.
* People created characters off peak, when they couldn't view server capacity. So they chose randomly.

A GenerationA Server has 100% GenA population, 50% of GenB, 33% of GenC, 25% of GenD. While a GenerationD server has 25% of GenD.

If the Generations were equal size, the GenA server has 8 times the players of a GenD server.