In a postmortem for Blizzard's Battle.net at Austin's GDC Online today, project director Greg Canessa warned developers about underestimating the difficulty of creating an integrated online gaming network.
"It's really hard designing and building this stuff," said Canessa, who was also a key leader in the development of Xbox Live at Microsoft. He said it's a "common and frequent" misconception that developing an integrated online gaming network is an easy task. "Design iteration is just as important here as it is with game titles," he said.
The new version of Battle.net launched alongside July's release of StarCraft II, which is tightly woven into the Battle.net system. Canessa revealed that Blizzard had been developing and iterating on the network since 2007, making changes right up to StarCraft II's launch. "Build enough time into your schedule to iterate on these services," he advised.
One of the numerous hurdles that Canessa said Blizzard encountered in developing Battle.net was aligning efforts with the game design team for StarCraft II. Battle.net is deeply integrated with the actual game, so features of the service could have a profound effect on the StarCraft II experience, and vice versa.
"Make sure you're partnering 50-50 with the game dev team," Canessa suggested to game network creators. "It can help out with time to market and stress level." He said it "took a long time" for the teams to get on the same page, but Blizzard gave itself enough time to work out the issues.
Before Stacraft II, Blizzard hadn't released a non-MMO game for several years -- the studio's business had been all about the massively-successful World of Warcraft since 2004. "Blizzard is a company that back in say 2002, 2003, when we were launching Warcraft III, that's the last time the company shipped a non-MMO boxed product. … It really was an adjustment for us, thinking beyond the MMO."
The original Battle.net, which launched alongside Diablo in 1996, claimed 12 million users, half of whom were from Korea, as of a year and a half ago. That matches the number of World of Warcraft players registered today.
The new Battle.net didn't only have to integrate with StarCraft II, but also with the pre-existing World of Warcraft. The Battle.net team had to listen closely the that huge fanbase, try out social features, and try to developer deep integration without disrupting the MMO's audience or the game itself.
"Integrating with a community of 12 million users and not screwing it up is a huge challenge," said Canessa. "…Integrating a game service and an MMO is just challenging from a technical and compatibility standpoint across the board."
Battle.net technical director Matthew Verslyus added, "This is easily the most complicated launch I've been a part of at Blizzard."
Since its inception, Blizzard's Battle.net team has expanded from one person to about 50 people. Blizzard will be expanding its team substantially in the coming years. As part of what Canessa called a 10-year roadmap for Battle.net, Blizzard plans to create a "living game service," tightly integrate the network with each Blizzard game and development team, and "build and scale a Battle.net team to pull it off."
Finding that talent can be difficult -- while many graduates and industry workers have skills related to art, programming and design, there are relatively few with real know-how about creating an integrated gaming network, Canessa said. And finding key talent is the difference between success and failure.
"The bar has gone way up from 10 years ago," he added -- customers expect a lot from these networks. "It's completely different today than it was in the year 2000." Canessa also warned that "it's really, really expensive to go this route." In fact, Battle.net required a full client rewrite for each iteration: "Just know it is very expensive and very time-consuming."
For Canessa and the Battle.net team, the key to success is to "think through this stuff early, really understand what is important to your customers in each market, build in the tracking mechanisms to evaluate success and failure and understand regional challenges."
Canessa said, "Launching the service is just the beginning. I can't emphasis this point enough."
While it's difficult, they've done better previously.
One issue was they made things hard for themselves trying to lock everything down with DRM which, despite being hideously awful to users, is also hideously awful to code.
They also decided to make a multi-purpose system to link their different games instead of a dedicated SC2 system (like all their successful gaming networks previously).
Also, they added on-line monitoring of single-player.
IMO they stumbled over their own makework.
"The bar has gone way up from 10 years ago," he added -- customers expect a lot from these networks. "It's completely different today than it was in the year 2000."
That's not what I've been hearing on all the forums. They pretty much want what WC3 has.
As products, Steam indeed feels superior to Battle.net. You have to understand though that the challenges are completely different. With Steam, the platform came first and the games are built on top (with the notable exception of CS:S achievements). With Battle.net, they had to build the platform on top of a 12M subscriptions MMO. I'm not entirely sure if WoW achievements are integrated with Battle.net, I have a feeling they aren't in the same way that SC2 achievements are. But regardless, dropping something like that on top of a huge live system seems an immense task!
In my eyes steam isn't all that super but still a bit better then battlenet, still I don't think steam has the potential left to keep it up being the top player for 5 more years unless some stuff changes.
Yet if you compare Battlenet to for example EA's systems, hehehe now there is a big difference!
"Make sure you're partnering 50-50 with the game dev team,"
... I have been working on a RTS with an integrated network and I felt like designers were a pain to the process. I'd say you'd better give the lead to the dev team and let the designers simply give the trend and advices.
This is especialy true if your game has to run on both Xbox Live and PS Live which add a lot of constraints to server code architecture. To sum up, you really need to have a very strong communication between designers and devs, a two-way communication I mean.
A good point has been raised here. It is a massive disadvantage for Battle.Net to not have separated Friends List -style support.
I have Steam load on startup, and it just sits in the background throughout the day. Friends can invite me to a game, chat, etc, and I can also see what they're up to with a glance. I can also access the store to easily check deals and even update my favourite servers at a whim to see if I want to hop in.
Battle.Net, on the other hand, only runs from within the game, or at least as far as I can tell as a recent registrant, having never played WoW but having gotten into SC2 quite seriously since its release. I often wonder if my friends are online in SC2, as it is a real incentive to play, but a load-up and sign-in is a long process. Granted, Xbox Live suffers the same issue, but I can still sign in on the webpage to see what Friends are online.
It's a shame that we have to turn to a third-party program like XFire to get the features we really want, but even that is lacking. It's odd that, with so much work and polish put into Battle.Net, they leave out such critical features that are already available and proven on competitive platforms.
The new battle.net is a disgrace: is way worse not only than Steam, but than the old battle.net. I'd never have changed RealID, facebook integration and a bunch of achievements for LAN support and cross region play, not to mention the restrictions in space and freedom when creating custom maps.
I am impressed by the fact that Steam is getting such props in this thread. I can remember when it was hated and reviled and nobody could understand why Valve wasted their time with it.
Give Blizzard time to work out the kinks. I love Steam and I am sure I'll come to love Battle.net. Can't wait for Diablo III.
One issue was they made things hard for themselves trying to lock everything down with DRM which, despite being hideously awful to users, is also hideously awful to code.
They also decided to make a multi-purpose system to link their different games instead of a dedicated SC2 system (like all their successful gaming networks previously).
Also, they added on-line monitoring of single-player.
IMO they stumbled over their own makework.
"The bar has gone way up from 10 years ago," he added -- customers expect a lot from these networks. "It's completely different today than it was in the year 2000."
That's not what I've been hearing on all the forums. They pretty much want what WC3 has.
Steam still has something that Battle.NET doesn't: An IM program one can use when not running a game.
Yet if you compare Battlenet to for example EA's systems, hehehe now there is a big difference!
... I have been working on a RTS with an integrated network and I felt like designers were a pain to the process. I'd say you'd better give the lead to the dev team and let the designers simply give the trend and advices.
This is especialy true if your game has to run on both Xbox Live and PS Live which add a lot of constraints to server code architecture. To sum up, you really need to have a very strong communication between designers and devs, a two-way communication I mean.
PS: I was in the design team.
I have Steam load on startup, and it just sits in the background throughout the day. Friends can invite me to a game, chat, etc, and I can also see what they're up to with a glance. I can also access the store to easily check deals and even update my favourite servers at a whim to see if I want to hop in.
Battle.Net, on the other hand, only runs from within the game, or at least as far as I can tell as a recent registrant, having never played WoW but having gotten into SC2 quite seriously since its release. I often wonder if my friends are online in SC2, as it is a real incentive to play, but a load-up and sign-in is a long process. Granted, Xbox Live suffers the same issue, but I can still sign in on the webpage to see what Friends are online.
It's a shame that we have to turn to a third-party program like XFire to get the features we really want, but even that is lacking. It's odd that, with so much work and polish put into Battle.Net, they leave out such critical features that are already available and proven on competitive platforms.
Blizzard failed BIG with SC2.
Give Blizzard time to work out the kinks. I love Steam and I am sure I'll come to love Battle.net. Can't wait for Diablo III.