We base our projections off of what we've seen in our own games, where we have perfect visibility into the data, as well as others'. We then scale up our capacity to match expectations, and purchase a reasonable buffer of overflow room for extra. As people saw during alpha and beta, our forecasts held up 100% correctly through the beginning of head start.
Can the team discuss relative launch numbers?
Sure thing. We'll be announcing that second million users a lot sooner than we would have expected.
Why didn't the team have extra servers in place ahead of time?
We did. The extra overflow hardware were all put into service the instant we saw what was happening. At that point we immediately pulled up the "too much success" plan and jumped onto it with our hardware vendors. We had three backup plans for "What if we need even more hardware," and we went with the one that would get us the most new hardware, the fastest.
The amount of extra overflow capacity is the real issue here. The purchases we make to scale up for game launches are multiple millions of dollars. As much as I wish otherwise, it's not realistic to multiply those purchases into the tens of millions of dollars, on the off chance that it was needed.
I also wish that cloud computing was performance- and cost-efficient for games like this, so we had more ability to be more elastic for launch loads. Unfortunately, cloud services really aren't there yet for MMOs. Cloud does storage, web, and distributed processing exceedingly well but doesn't do persistent MMO at large scale very well yet. We did evaluate that as a potential overflow plan for ArcheAge, but the performance just wasn't there for it to be a good game experience. We're greatly looking forward to the day it can. It's an area we're keeping a close eye on for the future.
Why the need to ship more servers across Europe post-launch?
As has happened in other MMOs that had huge interest beyond anything they could have rationally expected, their owners have had to purchase more hardware to hold the load. You don't hear a lot of huge success story cases in MMOs anymore; that's very true. As you'll probably remember, the last time this happened in an MMO, new hardware took weeks/months to arrive, and servers were frequently unplayable in the interim. Neither of those has been the case with us, fortunately.
Our teams got North American hardware set up in three days, and EU hardware should be set up this week. If someone has done it faster, I'd invite them to let me know about it -- seriously. In all of the MMOs I've been involved with in the last 15 years, this is the fastest I've ever seen it done.
How exactly has Trion changed the queues to favor patrons -- weren't they already favoring patrons?
Yes: It's ratio-based tuning. The ratios by which servers admit patrons vs. non patrons was tuned to further favor patrons. While we want a world where everyone can play instantly, for the time being, we felt the need to take extra care of the founders who have been supporting AA all along.Can you debunk, once and for all, the ugly rumor swirling around the NA servers that free players are being booted from the game to make space for patrons?
This question is the first time I've heard this rumor, and I spent the entire weekend interacting with our players in-game and on twitter while working. Happy to confirm: That's definitely not happening.(Thank you for that. It was a tin-foil hat rumor we'd seen spreading in-game in chat.)
Isn't Trion concerned that new servers added now will be dead servers in a few months after the usual post-launch slump that hits nearly all MMOs? Won't those additions necessitate problematic merges and the loss of land and character names for players who switch to low-queue servers now to help out?
Of course we are being extremely careful to not go overboard. :) That's why we've been careful about not opening too many servers, too rapidly. For us, this is an act of finding the optimal balance of healthy servers in the long term vs. wait times in the short term. That is precisely why this is difficult. We've heard just as many calls for "no more servers" as we have for "more servers." The view from in-game couldn't be more different than the view from the some places on the internet.And what about guilds whose players are now stuck on different servers because of character creation blocks on the worst-off servers?
We do expect these blocks to be temporary, and you bet everyone is working extremely hard to make sure that it's as short of a time as possible.What sort of compensation to players is Trion considering for these launch problems? Patron time? Cash shop cash? How can the studio adequately compensate players who missed out on the land rush thanks to double-digit queues?
Our update last night addressed the compensation to date: It's important to us to make good by people. At the same time, competition for, and trade of, contested land is a core part of this game, which is vastly different than what's gone on in AAA MMO releases lately.What would Trion say to players who are patiently waiting on the sidelines for this madness to end before they try to play?
Over the past weekend, with the new servers in place in NA, we're hearing a lot more stories of: "I queued for a couple minutes, and now I'm in," when new players are rolling up on the newer servers. Our goal is to bring that type of experience to as many people as possible, as fast as we possibly can.Thanks very much to Scott Hartsman for his candid (and timely) answers.When readers want the scoop on a launch or a patch (or even a brewing fiasco), Massively goes right to the source to interview the developers themselves. Be they John Smedley or Chris Roberts or anyone in between, we ask the devs the hard questions. Of course, whether they tell us the truth or not is up to them!