OGDC Event Coverage - OGDC in Review

The Online Game Development Conference was held on May 10-11 in Seattle, Washington and was the first conference focused on technology, art, design, production, and business of games delivered over the internet. The two-day agenda was packed full of seminars hosted by leaders in the online games world, including NetDevil, Cryptic Studios, Microsoft, and more. There were 4 rooms consecutively serving up seminars, with three in the morning and 4 in the afternoon on Thursday, and three in the morning, three in the afternoon on Friday. A keynote each day was held after lunch, conveniently located in the same room. While the 10-min breaks in between sessions were rather small and led to some sessions running off schedule late in the day on Friday, the trade-off was necessary in order to pack in as much material as the conference could take. To further help with the scheduling, the OGDC organizers sent out a survey asking attendees what seminars they were interested in seeing, no doubt with the intentions of organizing the schedule as best as possible to allow people to get the most out of the conference and not be stuck too often having to decide between two or more seminars in the same time slot. I thought this was a nice touch, although I don't know how well it was actually pulled off. I didn't have any conflicts, but I wasn't interested in nearly as many sessions as I'm sure others were.

Speaking of the sessions themselves, I for one did not think I would be interested in even as many as I found myself attracted to. I'm not in any way invested in the online games space, but a lot of the talks being given also addressed issues outside of online games - in this case they were merely applying them to online games. So all the sessions I attended I found informative and enjoyable, and I'll be sharing some on the next page. The rooms allocated for the sessions were small, so serious audio wasn't usually necessary, in fact many speakers chose not to even use the provided lapel mics and instead simply projected their voices. The visual part of the presentations were all top-notch and there was never any trouble seeing the projection screen.

Though there wasn't much time for serious mingling between sessions, the fact that the session rooms were in such close proximity to one another allowed you to at least throw out a quick hello or a handshake and short conversation. Plentiful amounts of snacks were provided as well, which is a good excuse as any to stop and chat. The snacks changed for the morning and afternoon sessions, ranging from cookies and muffins to fruits and granola bars to yummy ice cream bars on the last day even. Although I only ate lunch at the event one day, the food was quite good and many of my fellow attendees were complimenting the food selection as well. So thumbs up for the eats.

As I said, there are only 10 minutes in between sessions and not much time for serious networking besides lunch. Luckily the conference organizers recognized this and provided a networking party on the evening of the first night. To be honest I was expecting a bit more in the way of food other than small finger snacks of the kind where you're never really sure exactly what you're eating, even if it does sometimes taste good. The party was held at dinner time, but other than the complaints from my hungry tummy it served its purpose to gather conference members together for a good time of schmoozing and boozing.

In addition to the conference sessions, OGDC also made an attempt at an expo, though I'm not so sure the attempt was completely successful. The main problem was the fact that the expo was located on the other side of the hotel from the session rooms, so traffic past the expo booths happened mainly in the morning when people arrived and before/after lunch since the ballroom hosting lunch was past the expo area. There were about ten booths hosting companies like Big World, IBM, +7 Systems, Comcast, etc. I never really stopped in there myself due to lack of time. It was just a case of bad layout to work with and not the fault of the event organizers.

In the end it was a very enjoyable event. I've never been to a small, focused conference before and it was very relaxing compared to GDC, which has like a bazillion things happening at once. In addition to the conference, the city of Seattle itself was very pleasant. My hotel in the downtown area was just a few blocks from the conference venue, and I was able to walk anywhere in the city, from the Space Needle to the famous Pike Place Market. If you attend this conference next year, which I most likely will, I'd definitely suggest sticking around the following weekend to do some sight-seeing.

OGDC Event Coverage - Select Sessions Coverage

Appealing to the MMOB: Building Massively Multiplayer Online Brands
Brian Robbins, Executive Producer / Gaming Evangelist - Fuel Industries, Inc.

Brian talked about a few key points involved in building a massively multiplayer online brand. What exactly is that you ask? Well it's just taking branding to a new arena, that of MMO games. No, we're not talking about in-game advertising and dressing up your Orc with a Pepsi banner in WoW, we're talking about building complete online games around an existing brand name. For example, there's Coke Studios, Virtual Magic Kingdom, Nicktropolis, and even MTV tying together 3 of its TV shows together into a seamless online world.

The best part about building these kinds of games is that they are almost exclusively funded by marketing, which is important to realize because that means you instantly have access to a lot of funding. The downside of course is that marketing is a pain in the ass. This means that all approvals on game milestones (for example) are based solely on screenshots and logo usage; the people who are making the decisions as to whether the game is good or not have little to no actual gaming experience; and because of all this the brand always wins over gameplay. Brian gave an example of a Chester Cheetah paintball game that was killed when the developers allowed players to actually shoot Chester Cheetah. Hey, it may have been a fun thing to do but as far as marketing was concerned you were harming the brand. Big no-no.

So why even bother with such a project? Well the obvious appeal is that whether or not these games generate revenue is not the main worry here. In this case all the focus is placed on driving the brand. So you don't need resource sinks or a balanced economy to fund your way through the project (remember, these are marketing dollars) and it also leaves room for lots of experimentation for several reasons. For one thing, these games are much easier to tie to offline thanks to the fact that the brands already exist offline. For example, Coke makes use of its coke codes, and Disney ties in the actual parks to its virtual park game. The fact that the game allows for offline interaction could also be the driving factor in its success. You can make full use of game mechanics that don't drive revenue, but do drive usage - like customizable characters and custom levels/maps. In addition, you can target audiences that don't monetize well, because your focus is pushing the brand, not making money. Still, you do need to see some return on the investment, which equates to the number of players (Coke Studios had over 7 million users in 2006), repeat visits, knowing how long your players spend time in the game world and tracking your players offline.

It's important to always remember though that brand vision rules all, and marketers worry most about promoting and protecting (don't kill Chester Cheetah!) the brand value, which is not always conducive to game design. The best way to fail at this type of project is to ignore the brand team's feedback. Instead, find a way to compromise rather than focusing on gameplay over the brand. Also, never try and hide the brand's involvement. Be honest with users about the brand and remember that it's your job to promote the brand vision. Find out ahead of time what's important to the brand team so you can design a game without having to reach too many compromises in the game design.

Here's the final deal: You're getting paid to work on a really hot property that already comes with a brand following and almost guaranteed players, so you don't have to worry about milking money out of the player base. If the brand is a household name (or close to) then it's suitable for this type of online transition. It doesn't need to be huge, it just has to attract users.

Oh and never dismiss ideas put forth by marketing - sometimes you may be surprised.

Working in the Data Mines: Uncovering Gameplay Gold
Darius Kazemi - Orbus Gameworks

The main goal of Darius' talk wasn't about actual data mining - that is, using metadata to extract useful information from large data sets - but rather about metrics. Particularly, metrics relating to gameplay in an MMO, such as recording data about quest completion, character advancement, trade, social behavior, combat systems, AI behavior, etc.

Since Darius included very copious notes in his slideshow presentation, I'm going to refrain from copying and pasting much of his words here and simply provide the file. Please see the download link to the right. The slide show is literally an article in its own right.

Games for Windows - LIVE! Essentials
Zsolt Mathe, Software Developer - XNA Developer Connection

Overview

If you've been wondering about the upcoming integration of the LIVE service between PC and Xbox, then read carefully because the following information is for you. Firstly, this integration of LIVE is out to solve many problems that Microsoft claims to plague Windows games (and is mainly correct). These include user challenges such as network configuration, multiple 3rd party apps and per-game quirks to frustrated communities that have multiple identities, no cross-game communities, and cheating, griefing and piracy abound. Based on their success with Vista, I don't see how Microsoft is set to handle the piracy problem, but I cede all their other points. Finally, every online game in some way tends to reinvent the wheel in terms of technology.

Obviously, we have Windows LIVE to the rescue, with the full-featured LIVE service and cross-platform LIVE community fully functional on the PC platform, with no barriers for entry (well, at the basic level anyways). On the developer side we have XNA, which has already been proven to be an extremely accessible code base, rich game features such as unified friends lists, voice chat, leaderboards, matchmaking, achievement and gamerscore, and of course the fact that the API is cross-platform between Xbox and PC.

Microsoft touts the easy setup involved in server connections for Windows LIVE games, which should require one single configuration forever and will punch only a single hole for all communications through a firewall and include automatic NAT traversal and UpnP. This seamlessness will also extend to the user interface, which will feature a consistent UI and support the global community through single Xbox/PC accounts, a unified community website and support for fan sites and mashups.

Closer Look

But how about some nitty gritty? Well first there is the XSocket API, which replaces WinSock and provides for more private network communications. Microsoft made the smart choice of working to make the XSocket call-signature compatible with WinSock to ease developers into transition from one API to the other. For added security, addresses and ports are all virtualized (as an option), packets are sent through a VPN tunnel utilizing a single UDP port with strong encryption as well as multiple authentication layers (all highly abstracted of course, not to worry). XSockets also include support for TCP, UDP and VDP (voice and Data Protocol).

Windows LIVE targets hackers by including debugger detection, as well as a protected buffer API. The API randomizes and replaces data withing the memory buffer to prevent scans from picking up anything useful. This means you have to use the API's special functions to read and write data that is then obfuscated and stored in the buffer.

To help prevent against cheating runtime cheat checks are performed, including signature checks on load, periodic codepage and IAT checks, and challenge-response from LIVE routines that can be tailored to individual titles. All these runtime checks are done in the background with low overhead and assigned low task priority. User feedback integrated into LIVE can also help fight against cheaters, such as the reputation and community reporting mechanisms. While obviously no system is 100% secure, Microsoft has dedicated a full-time security team to the task of staying one step ahead of the hackers/cheaters.

Why Bother?

Well, the best reason is the fact that Microsoft has created a service that comes to developers at a minimal cost. Microsoft foots the bill for the basic service, premium goodness is left to the LIVE subscribers, and the developer is tagged with the development cost - of the game. No server costs, no community management software costs, no online integration tools cost. A very nominal fee when compared to building it all yourself. Porting between Console and PC is a very easy process thanks to the nearly identical APIs, even the socket APIs differ only by name. However Microsoft discourages straight ports and reminds developers to take advantage of platforms. There are a few gotchas to remember with Windows LIVE titles as well. For one, there can only be one person logged in per box, unlike Consoles which can (obviously) have up to four. In addition, the players must be logged in and online in order to earn achievements, and the PC presents different attack vectors that Console does.

So when is this all coming? Currently the first public release is set for summer of 2007 with the initial feature set currently found on LIVE today. Future features are to include LIVE Marketplace, LIVE Arcade and a standalone client. Microsoft also plans on release points every 6 months or so

All conference photos courtesy of the OGDC staff