Thursday Keynote
All The Reasons You Will Fail or Don't Even Think About It
Mike Wilson & Harry Miller (CEO & President of Game Cock Media and founders of the former Gathering of Developers)
The seminar opened with a video showing Wilson & Miller in a Mexican prison. The presenter (another
Game Cock employee) opened and gave the presentation in their stead.
The industry is led by brands, movie tie-ins, and very long-lived IP (Final Fantasy). Original IP is a staple of success in games. Unfortunately, far more projects fail than succeed in the market.
Within a week of their opening, Game Cock received over 100 submissions from independents. Quality is the key and is where many games fail. If your game lacks polish and doesn't compete with the AAA titles, then you should expect your game to fail.
A good example of entrenched IP was the "ESPN Games" football title versus the Madden-branded title. Even though reviewers considered the ESPN version to be superior, the Madden-branded title did better because of the entrenched IP and reputation of the product.
Speaking of reputation, you need to have a reputation as someone who can get a complete and polished product out the door.
Original IP is difficult to sell to big retailers. Large retailers have a limited number of spots available, and it's a safer bet to go with a mature IP or a media tie-in rather than a new IP.
In many ways, it's far easier to bring a title to market than before now that digital distribution is the norm.
The biggies like THQ, Activision, and the like have business teams who are gang-tackling every big new IP that's on the market. They're also always on the lookout for teams who can convert that IP into a good game reliably.
It's also important to keep your team mobile, because changing-of-the-guards do happen at high levels and big marketing decisions can be made that'll close your project.
Small projects still happen. One of the latest triple-A titles for the Nintendo DS was put together with a three-person team for a relatively small (under 2 million) budget.
It's also difficult to change your reputation. Not the one that you put out but the one that other companies impose upon you. Game Cock apparently has a reputation for making small boutique-style games, even though they've done some large-scale projects. Learn to work with (and sometimes around) your own outside perception.
There's really no such thing as a studio that's too small anymore. With things like XNA and Xbox Live Arcade, part-time development on a shoestring is still viable. It's not just a matter of getting together a huge budget as step one to your project.
The final bar is "did your last game sell well enough to justify making another one?"
Off the record notes
It's rather lame that Wilson and Miller didn't have time to speak at their own keynote. Strangely though, they had time to record a wacky video of themselves pretending to be in a Mexican prison. There are fewer than 100 people in attendance. I wonder if these guys just considered themselves to be too important for this scale of event. Oh well.
The guy next to me is editing code in Visual Studio on his laptop. Does he even know where he is?
Secure Your Game: Protecting Your Code
Steven Davis, CEO of SecurePlay
Developers are the biggest cheaters. Cheating should be built into the game design, not imposed later.
Think about security up front. "Think Security" should be something that pervades your development process.
There's no such thing as perfect security. Live with it.
BioShock was proud of their security because it worked for 13 days before it was cracked, so whatever they paid for their security/DRM system gave 'em two weeks of good security.
System security should be inherently simple. "If it's not simple, it's not secure".
Bad guys are smart. The key isn't to lock out the bad guys but to just make it difficult for 'em.
Completely software based security, by and large, doesn't work. That's mainly because it's running on the machine, so it can be examined.
For some Chinese MMORPGS, the second most important reason to select a game was "no hacking and cheating". The number one reason for leaving a game was "hacking and cheating destroyed the experience".
Webkinz (Google for it) - the physical item is itself the anti-piracy device. The security can exist as a physical device. An online card game, for example, could come with a really terrific manual that people will want.
Antipiracy in online games. Make things like credit-card storage part of an account, as that'll make people loath to share that account with their friends if they can your card through it. I didn't get a chance to press the presenter about this. Most credit card processing nowadays requires you to enter the three digits on the back of the card, and credit card companies absolutely forbid you to store that number on a server. I guess just the very fact that your card's stored with your account is enough to make someone avoid sharing the account.
Linking community into the game makes it more secure (helping to prevent cracks) but doesn't make it appear that you distrust the user.
"Gamestop is worse than pirates" - The retailer takes a bigger cut than pirates. You're better off with an online subscription model, and you'll likely make much more from subscription fees than from the miniscule amount you'll make from the retailer.
In China & Asia, networked games are moving towards a P2P infrastructure. Central servers are handling logins and lobbying, but actual gameplay is being done peer-to-peer, just because bandwidth costs are crippling, although bandwidth is cheaper there than in the US.
Protecting assets - keep 'em on a server where possible. Otherwise there's little that can be done about cracking and users exporting/changing game assets. Regarding exporting valuable assets (i.e. music), it's better to make it a value thing rather worrying about exporting and sharing the song. Make multiplayer or extra downloads or something like that tied to a serial-numbered song on the machine.
Guitar Hero's anti-piracy tool is the plastic guitar. The software itself just isn't as compelling without the plastic guitar.
Disney's anti-piracy tool in China (where piracy is rampant) is a little gold medallion-sticker-thingy that could be redeemed (online or by mail) for a contest entry. If someone sent in a bogus medallion, they'd find out the retailer selling 'em. The security budget comes out of the marketing budget.
Off the record notes
The speaker was really engaging, and I could've heard much more of his presentation. His powerpoint deck consisted of 140 slides and was intended to be a full-day tutorial, so we just got a taste of what he had to offer. I'll have to check this guy out if he does the full day sometime.
His main take-away seems to be that real security for non-online games just can't be done with any kind of effectiveness. I'll have to check out his website and blog to see what exactly he's selling :)
www.secureplay.com - his main site
www.playnoevil.com - his blog
Anatomy of a Perfect Pitch
Mike McShaffry
Pitching Your Game - there are five stages
Stage 0: Create a great concept. Try to ground it in things that the publisher will trust. If you like playing X, then you'll enjoy Y. You should have more information available than you'll actually be pitching to your publisher.
Stage 1: Research. Look at the games in the industry that are close to your game. Try to find a publisher that sells similar, but not directly competitive games to yours. Be sure to ask the business development department lots of questions. Cold calling can get you to the right people, but make sure you're talking to the business development folks from your area (US versus Europe). Tune your presentation to the size of the group with which you're dealing.
Stage 2: Rehearsal. Pitch your game to your friends and colleagues. It'll be a mite uncomfortable at first, but deal with it. Rehearse EVERYTHING, including setup and tear-down. Pitches have been lost due to lack of a video-cable.
Stage 3: The first meeting. This isn't the real pitch. You're meeting with some lower-level people who are just screening out the folks who don't have a serious chance. Don't be afraid to go with a fake box and other basic materials. Don't do this at the GDC. This can't be done cheaply. This will take time and money (travel, materials, etc).
Stage 4: The real pitch. Don't make demands during this stage. You're still interested in selling your idea to the business development folks. You need to be uber-professional at this point, because you won't have time to do this again if you screw up. Try to be flexible with flight and travel plans, and be prepared to spend an extra night, because bizdev folks are often working under the assumption of unlimited expense accounts.
Make sure the client knows what your AV needs are and work with it.
Communicate with passion, with authority, and with humility.
Use a powerpoint presentation, but don't use the stock "wizard" presentations that come with it. Your presentation should contain the following:
- Slide 0: Cool concept art, but no actual content.
- Slide 1: Introduce your team and studio.
- Slide 2: The game concept. It should be a single sentence that describes the "soul" of your game. Know your genre and avoid cross-genre issues (puzzle-racing-shooter).
- Slide 3: Why your game is fun. The repeated point was "how would you convince your friend to get this game".
- Slide 4: The fiction, characters, and setting. It should be interesting, but also marketable.
- Slide 5: The demo: It should be playable if at all possible. Even better if the bizdev folks are allowed to play it. Use a prerecorded demo if a playable one is not possible.
- Slide 6: The target market and platform for your game. Also your target rating (teen, mature, etc). This is important information for marketers.
- Slide 7: The technology you're using (the engine), and how doable your project is.
- Slide 8: Your team's credentials and past successes.
- Slide 9: The current status of the game (complete, production, preproduction, etc)
- Slide 10: Why you're here. Are you looking for a deal? Are you looking for feedback? Are you looking for funding?
After that, it's time for Q&A. Be flexible but not too flexible to suggestions and questions. Anticipate common questions and have visuals to back that up. Be ready to be blindsided by stupid and/or unexpected questions.
Stage 5: Follow Up. Have a postmortem meeting, but save it until you're out of the building. Be prepared to wait for feedback, and don't come off as desperate by demanding it immediately. If they want to visit your studio, that's a good sign.
Off the record notes
The speaker was pretty animated and interesting and clearly knew his stuff. He had more stuff done in the game industry than I could list, so just Google for him and you'll find his credentials.
Unlike the previous speaker, Mike was able to cover his subject comprehensively and in the allotted time. This is good and bad. It's good in that I got the "whole package". It's bad in that I didn't feel like I wanted more. Such is life.
3D Game ENGINES - What to look for, how to choose
Mike Duggan3D GameStudio, Torque are the two big players that'll be discussed here.
Game Engine - a total framework and complete software development kits that are programming frameworks used for games.
Game Engines settle physics, AI Bots, collision detection, and graphics rendering for you so you don't have to.
Commercial versus free. The free or low-cost ones tend to target a specific game genre. Most of the open-source ones are code-libraries that require you to own a compiler.
Commercial and cheap
Torque - Windows, OSX, Linux, C++ and C# and its own scripting language. Works with loads of 3D formats. $150 per programmer for <$250k sales
3D GameStudio - uses a C-ish script language or can talk to C++ or Delphi. Windows only, but very stable. $99-$199
C4 Engine - seems to have similar specs to Torque. $200
Most will require some kind of branding for their product unless you want to pay a higher price.
Non-Commercial
Blender - all open source and OpenGL. Wide platform support. Little to no actual published support for games created entirely in Blender OGRE (www.ogre3d.org) - Windows, Linux, OSX.
Irrlicht - lots of late-breaking features. Free. Integrates well with 3D GameStudio.
Cube 2/Sauerbraten - open source and ships with SUSE, so there are plenty of games that use it. Built-in shaders and multitexturing.
Wintermute Engine - Available since 2003. Competes primarily with Adventure Game Studio. Designed with point-n-click adventures in mind.
Adventure Game Studio - For creating Sierra/LucasArts style games. Completely free.
RPG Maker - RPG's only. Scripted in Ruby.
Check 'em all out and figure out which one best supports your needs.
Off the record notes
This one was pretty loose, and the audience provided as much input as the speaker. I did hear of quite a few engines that I'd never heard of before. Always choose your engine carefully. Some look quite nice but are as well built as a Yugo, and changing engines midstream is not an easy thing to do.
Protecting Your IP - Keeping Your Rights and Protecting Your Future
Zachry Bishop - Hunton & Williams LLPEmployees & Consultants
Use of third-party materials
Copyrights, trademarks, and trade secrets
Publishing agreements
The downside of Success
Under copyright law, if you create a work and put it in a "fixed" form, then you own it. There's a "work for hire" exception that makes the work the property of the employer if done under the scope of employment. This determination is difficult sometimes. The work for hire concept applies only to copyrights. There's also a problem with works created prior to hiring.
The solution is to enter into written employment/consulting agreements with all employees and contractors that explicitly spell out. This should ideally be done before employment, but it can also be done after-the-fact. Agreement should cover what should be the scope of the code (employees cannot take code home, for example). It should also have exceptions for non work-related content. It should clearly state what is "company" IP. Company property must be returned upon termination. Any obligations to former employees must be spelled out. Non-compete during and after employment should be spelled out. New employer should be notified about prior agreement.
Something new is "personal likeness license" that allows the company to use your likeness in any photos or videos.
Artistic content. Permission is necessary for all third party artistic content included in a game. Includes artwork, voices, music, sound effects, motion capture, and fonts.
License agreements for tools and middleware. Can be flat-fee or royalty-based (or both). The scope of the license can cover one or multiple games, platforms, and rights to third-party developers. There can be attribution requirements ("powered by. . ."), as well as licensor warranties, indemnity obligations, and limitation of liability. You should define the IP ownership line between the middleware and the game.
Open Source Software - Software that allows for its use and modification, sometimes subject to further licensing.
GPL - "viral" nature. Code is attached with a GPL license, requiring code containing GPL to similarly be licensed under the GPL. To avoid becoming subject to the GPL, don't include any GPL software internal to your project.
BSD license - Redistribution is allowed as long as license is included and credit is given. Non-viral.
Risks of open-source - licenses can viral, no recourse or support for problems. Licenses tend to be complicated and contradictory. There's little legal precedent regarding it.
Conclusion - know what products your users are using, open source or otherwise. Enact a policy regarding what software's being used. Have a review committee to look over prospective software for your project. If used, abide by all attribution requirements and the EULA.
Copyrights & trademarks
Copyright is created automatically upon creating the fixed copy for the first time. Copyright is strengthened by copyright notice, but copyright itself is implicit upon creation. Copyright registration allows for a public record of the copyright and allows for more damages in cases of infringement. Copyrights can be pre-registered, which is intended for works that have a pre-release release. It allows infringement actions to be made before the actual release is made, mainly for the case of leaked content.
Other copyrights include game characters, objects, and music. These can be copyrighted separately from the work itself for the case of derivative works.
Trademarks - words names, or symbols. Trademarks can be registered or can be common-law. Trademark registration is implicit upon sale. Federal trademark registration (R) indicates that the trademark has been registered with the federal government. In addition to stronger ownership, registration allows for easy lookup in a trademark search at www.uspedia.gov.
Where to register your copyrights. US, of course. Also the EU. It's a pain to do, because you basically need to hire someone in each location to take care of your trademark for that locale.
Policing your content is important, especially if you have mod-able content that might allow for creation of something that infringes (example: licensed characters being used in City of Heroes). Easiest solution is to have users to agree not to infringe, then watch out for infringement in the derivative content.
Intellectual Property (IP) includes, game code, content contributed by the publisher for inclusion in the game or used to market the game, ancillary merchandise rights (action figures, movies, books, comics), and sequel rights.
Research and Game Design - Spending the Time to Get it Right
Richard Garriott
Most developers take the easy and obvious route. First thoughts are rarely inspired, fresh, and
compelling. Just fixing the problems in your favorite games isn't enough. A truly worthwhile game demands study and effort.
Look for unique game mechanics. Tetris, Thief, and Orange Box are examples of that.
Intellectual Property is underdeveloped. Characters (Lara Croft), Settings (Myst, Oddworld, Alice), Stories (Ultima Virtues), can all be developed with depth.
New ways to have fun cannot be specified in a design document. Great ideas on paper often fail. Schedules often fall short of the development and tuning necessary to make a game "compellingly fun".
A relative approach is often better. Plan for plenty of cycles of new ideas, but the budget and schedule can still constrain potential new ideas.
- Iconography - key objects should be simple enough to be memorable (spaceship shapes). Important physical plot-devices should be recognizable.
- Terminology - must be readable, understandable, and meaningful (weapons, characters, spells, and places should be easy to pronounce and remember).
- Numerology - 3 principles, 8 virtues. Make things well-ordered. Well ordered fiction seems more real than reality (elements should be discrete and measurable).
- Symbols, Ankhs, codex - Humans simplify the world around them (symbols can be drawn from numerology).
- Parables, virtues, and ethics - makes stories relevant and current (not just kill, collect, and repeat/ You can recast contemporary issues into game context).
- Culture, languages, histories - makes IP real and complete and "not just a game" (Garriott's a big believer in "Tolkien" style development where the world's larger and richer than the one presented. There's a strong cultural history, including languages, beliefs, and characters)
- Familiar, yet fresh - well grounded discovery (base things in familiar cultural and historical background so that it's recognizable)
Tabula Rasa successes - gameplay featuring tactical combat vs. MMO inventory management. Use of dynamic battlefields versus static re-spawns.
The biggest take away from the seminar should be "research research research". You must know your topic very well and go beyond the easy answer to make a truly unique and rich design.
Off the record notes
I was surprised. Garriott's a compelling speaker, and he had some good "take away" ideas. If you get a chance to hear him in a conference, take the time to do it.
Network Game Design Best Practices: Delivering the User Experience, Hiding the Complexity
Harlan Beverly, CEO of Bigfoot Networks (maker of high end network cards)
Developer of a Linux-based network card (i.e. the card itself is a little computer running Linux) that
replaces the Winsock stack. They also develop some network communication software.
Online games that suck: Star Wars Galaxies. Bad network code and lag has pretty-much killed the experience. There were also some seriously bad design decisions, but bad networking design is a big problem for it.
Online games that suck: ShadowRun for PC. Integrated with Games for Windows Live. Is Vista-only for no apparent reason. Game is intended to give a "console experience" on a PC, which was a bad decision.
Online games that suck: PartyPoker.com. Matchmaking is very difficult. Casual gaming, but not a good casual gaming experience. No real social interaction.
Online games that rule: World of Warcraft. No FPS or FPS expectation. Spell timing and speed and flow were designed to hide latency.
Online games that rule: CounterStrike: Source. Not a great license like other games, but is well designed and succeeds because of that.
Online games that rule: Yahoo Internet Poker. Matchmaking is very easy and is integrated with friends from Yahoo Messenger. Easy to get into a game.
Meet the customers' expectation regarding the genre. Casual games should be appropriate for the audience. Ditto for FPS or RPG.
Figure out your network topology. Will it be based on network servers? Or a web server, or peer-to-peer? Is it going to be a MMO cluster or shared? You should know the typical latencies and response times for your given topology. Are your customers typically connecting via wi-fi or a wired card?
Build in a "budget" for your networking packets and be able to detect if you're going over-budget and are hence giving the user a degraded networking experience.
Match up expectations with reality. Don't run an FPS on a web-server or an MMO on a single dedicated server.
Plan first, then code to the plan. Plan your game based on the topology and its expectations. What happens if the user disconnects?
Project DarkStar from Sun - awesome Java-based MMO platform that's completely free.
Hero Engine - very good but quite expensive.
Multiverse - very high performance network MMO engine.
Microsoft's now giving out the Torque network library as the basis of their casual games portal. Good and free for certain classes of games.
Free network libraries - RakNet, HawkNL, Torque Network Library.
Off the record notes
This one was a bit of a hodgepodge. I was hoping to hear what the "bad" games were doing technically that made their networking work so poorly, but it seemed like his biggest problems stemmed from bad game design and not bad networking. I'd like to hear how improved networking can help a mediocre game improve, but I didn't get that.
Ads in Games - Enhancing Your Revenue Without Losing Your Customer
C Gordon Bellamy of Double Fusion
Double Fusion - a technology-enabled in-game advertising network. The model itself is flipping
around from old days where titles (usually sports) would pay advertisers in order to make games that are more realistic looking. Now advertisers pay for in-game display.
The market itself is still relatively small. On the console space, you must work with certain ad-networks (Massive for Xbox).
Casual gaming is a new and growing demographic, and PC's are the broadest market because it's the most open to competition.
They predict the in-game advertising market (static and dynamic, all platforms) will grow to a $1 billion business.
Despite early complaints, the market has accommodated advertising, and some folks have embraced it as it makes the game experience more realistic.
People are more forgiving of ads if they perceive they're getting some value for it. For example, UbiSoft released some of their old titles as ad-supported, and people were cool with the ads because they knew they were getting something in exchange for the ads.
Free to play games with embedded micro-transactions will grow in the US in the same way that it grew in Asia, and advertising will dovetail into that.
Off the record notes
Maybe it's just day-two-itis, but this one also turned out to be more of an extended Q&A than a one-way presentation. That's not necessarily a bad thing as long as the questions are of reasonable quality, and for the most part they were.
The Demo Room
The main event of the conference other than the seminars was a room for demo-ing projects. Most were noncommercial student projects, but a couple were quite impressive.
Dell was on-hand with a dozen of their huge impressive monster tower machines, but by and large, people ran demos on machines they brought with 'em.
Most impressive (both for your humble reporter and the attendees) was a Guiter-Hero style game that could be played with a conventional electric guitar and a little USB convertor-cable. The software for it is based off the open source StepMania game, so you should be able to put together a great guitar-god-wannabe setup fairly cheaply. . .provided you can find a decent secondhand guitar.
