Original Article
http://blog.hanfordlemoore.com/2007/01/13/how-one-man-made-an-mmo-an-interview-with-gene-endrody

Gene Endrody is the man behind the free MMORPG Sherwood Dungeon and other multi-player games on maidmarian.com. Despite the fact that Gene had little programming experience, he�s managed to do all the development himself using off-the-shelf tools and build a loyal following that often reaches 4000 simultaneous players.

So how did this one-man-team do it? I wanted to know, so I interviewed him. We talked about his background, the technical aspects of doing an MMO with off the shelf tools, and how he built up his user base. tools and build a loyal following that often reaches 4000 simultaneous players.

What is your background?

I started an Amiga Video Toaster dealership in the early 90s with a friend. The Toaster came with an early version of Lightwave 3D and that was my first exposure to 3D modeling. After Commodore went belly up, I became a 3D Instructor and Program Manager at the Center for Digital Imaging and Sound and later the British Columbia Institute of Technology. I joined Radical Games as a Technical Art Director in 2000. It was really the experience at Radical that put me in a position to pursue 3D web games. I got the opportunity to art direct on a few projects, but a Technical Art Director is mostly responsible for solving technical issues related to stuffing 3D art and special effects into games. It�s a great position because unlike many game industry jobs, you get exposed to the entire process of making games rather than just one specialty.

Tell us a bit about Maidmarian.com. When did you start it, and what were your goals for it at the time? What was motivating you to start it?

I�m a card carrying, Lord of the Rings loving, fantasy freak. I played Dungeon & Dragons as a teenager, read fantasy novels and made a few pathetic attempts at writing games on my Commodore 64. In 1996, I was planning to do some web delivered 3D animations based in Sherwood Forest and discovered that MaidMarian.com was still available. I reserved the URL without having a clear concept of what to do with it. (Steve Jobs had reserved RobinHood.com through Pixar just prior to that as well.) Working with properties in the public domain, like Robin Hood and King Arthur, seemed like a good strategy for a small independent developer. You get to work with a great property with lots of history and don�t have to worry about being beholden to the owner of the intellectual property. Even through the focus is now on games, MaidMarian.com has always allowed me to make a living with one foot in a fantasy world.

It looks like you�re focusing on Sherwood Dungeon at the moment. Is this the new vision for MaidMarian.com?

Sherwood Dungeon is the game I always wanted to make. The other games on MaidMarian.com taught me the lessons I needed to take on that challenge. That being said, the games work together to provide a variety of experiences with the same basic feel. I think of MaidMarian.com as one big experience or community with of lots of choice about where to hang out and what to do. Some of the games, particularly Marian�s World and MoonBase, aren�t really games at all�more like social holodecks to goof around in. Talk to your friends, dance, drive around�whatever you�re in the mood for. Ratinator is really the exception to this philosophy. It was more of an attempt to get on-board with the digital distribution craze in 2003 and make something that fit the business model of game portals like Shockwave.com.

 Maidmarian.com has evolved over the years. Has it been a learning experience?

 Absolutely insane learning experience! I was lucky enough to be nominated by Macromedia for a People�s Choice Award at UCON2001 after the Director 8.5 beta. I�d never won anything before and just being nominated gave me the confidence that I wasn�t wasting my time on this 3D web stuff. I wanted to do web based MMO�s but the industry seemed so excited about the digital distribution and casual games. After talking to a few game portal producers, I started working on Ratinator. That was my first revenue generating independent game, distributed through Shockwave.com. Although Ratinator was named by the Adrenaline Vault as one of the Top Shareware Games of 2003, it failed to have any real commercial success. After that experience, I tend to look at industry trends and hype with a great deal of skepticism. Ironically, now that I�m focused back on MMOs, there are signs that I inadvertently �caught the next wave�.

How has Sherwood Dungeon been accepted by players?

I get fan mail ! Can you believe that? Players seem to love the game and they seem to relate to the idea that I�m a just guy with a computer living his dream of making massive multiplayer games in his basement. MaidMarian.com has grown to 1.4 million unique visitors every month with Sherwood Dungeon taking the lion�s share. I don�t want MaidMarian.com to be some faceless corporation in the eyes of the players. I want it to be grass-roots and personal because it�s certainly personal to me. Trying to maintain that personal connection to the players is getting really hard, but I do my best to listen and keep them in the loop through my developer�s diary.

Tech Stuff

Your online games are developed with Adobe Director and run in a web browser using the Shockwave plug-in. When did you start using Director and Shockwave, and what was your programming experience at the time?

I started using it with the beta of Director 8.5 in 2000. That�s when Macromedia first introduced 3D into Shockwave and when I got really excited about the possibilities. My programming experiences was very limited�Commodore 64, Radio Shack TRS80 BASIC and some MEL scripting (Maya�s Embeded Language). I really learned to program in Director MX, as most of the veteran programmers on DirGames probably figured out from all the noob questions I ask. Honestly, I don�t think of myself as a programmer so don�t ask me about best practices for object oriented programming�you�ll get a confused blank stare.

Has Director/Shockwave made this possible for you?

I love Director and even with it�s challenges over the last few years, I�ve found no other product that can deliver hardware accelerated 3D multiplayer games to a large install base of users on a web-page.

You use Shockwave and Director for the client side, but what is the server side like? What technologies are you running there? I have three game servers running the Shockwave Multiuser Server under Windows 2003 on a 100 Mbit connection.

An application called Always Up runs the SMUS server as a Windows Service and restarts it automatically if a crash is detected. At times I�ve run very close to 2000 simultaneous players per server and been very happy with the latency. The SMUS is an underrated, disrespected piece of old software that just seems to crank along and do the job.

Sherwood Dungeon�s stats are impressive, and despite that I�m a huge fan of Director frankly it�s beyond what I thought Shockwave could handle. How did you go about doing an MMORPG in Director? Did you ever stop and think �this might now be possible�?

�This might now be possible� implies that I had some sort of grand vision or clear concept of what was possible when I started. I just dove in, crossed my fingers and hoped for the best. Sherwood Dungeon is a bit of an ongoing experiment, testing my own abilities and the limitations of Shockwave and Director. Because I work in such a weird, organic way, I can�t tell you what the game is going to look like next month or next year. There can be a risk of ending up with a spaghetti working this way, but the basic architecture of the game code has evolved to become quite robust. Often by doing creative things to the underlying systems, new features become not only possible, but also fast and elegant�much better than using brute force to shoehorn new features into the game.

How much of your time is working around speed/performance and download limitations, and how much of your time is true �game development�?

I can�t really separate the two. Sherwood Dungeon gets performance and download efficiency from the fact that much of the game content, including dungeons, forests and islands are procedurally generated. This is how I can have an infinitely deep dungeon in a game that�s almost small enough to fit on a floppy disk. The design process differs from more traditional narrative driven MMOs because you�re tweaking the parameters of a procedural dungeon generator rather than manually placing trees, chests, monsters or other game content.

You�ve talked a bit in the past about download sizes and times being a restriction. Seeing as how Director can make binary executables for both Mac and PCs out of Shockwave files. Have you thought about taking the game in the direction of an installable client application?

This is one of those questions where tech and business collide. Web games are inherently limiting because of download size. That�s not a bad thing for an independent developer because it effectively levels the playing field between you and the larger developers. Having an army of artists and vast cash reserves doesn�t give a large developer much of an advantage when the game needs to be small enough to run within a web page. By making an installable client application I would effectively be trying to compete with large companies, like Vivendi or Sony, on their terms�and that�s insane. By sticking to the web gaming space, being small and nimble with low overheads gives me the advantage. That makes a goal of developing Sherwood Dungeon into the most popular web-based 3D MMO achievable for an independent game developer.

Business stuff

How did you go about promoting maidmarian.com and Sherwood Dungeon?

Sherwood Dungeon is both an MMO and a web game, with one foot in each world. Web games have an uncanny ability to market themselves virally through word of mouth. Blogs, portals, email, forums, instant messenger, and social networks � anyone who can provide a link and say, �Try this cool game� just became your distribution partner. Players can go from discovering the game exists to playing it in literally 20 seconds. The trick is just making sure you don�t do anything to mess that up. Sherwood Dungeon and MaidMarian.com are designed to get you into the game quickly with an absolute minimum number of mouse clicks. Other than to maintain a very liberal linking policy for portals and websites interested in the game, I do very little to actively promote Sherwood Dungeon. This is actually the first interview I�ve done.

Your business model is to give the game away, and make money off of the ads, vs selling memberships. Has that been successful, and how did you come to that decision? Any major plans to change or augment the business model?

The viral distribution and ad based revenue model work well together. Ads were intended to provide revenue while I developed the games to a reasonable level of completion. I received plenty of feedback from players that the games needed to stay free. Once Sherwood Dungeon hit 4000 simultaneous players, it was clear that MaidMarian.com would be very successful on ad revenue alone. I think about other business models a great deal but as always, the devil�s in the details. Whether developers like to admit it or not, the way a game make money is a major influence on it�s design.

Earlier last year you quit your job as technical art Director for Radical Entertainment in order to work on Maidmarian.com full time. Having quit my job myself several years ago, I know it�s not an easy decision. What made you decide to go for it?

If you�re going to spend half your life working, why not do it on your own terms with the opportunity to profit by your own ideas? MaidMarian.com gave me an opportunity to follow a dream and determine of my own future. That�s the truth, but not the whole story. By the time I left, MaidMarian.com was already more profitable than my day job, so there wasn�t much of a risk involved. Radical Entertainment was such a great place to work that that I had to do some soul searching before making a decision that should have been obvious.

Has making the full-time leap changed priorities of the game for you? Has it changed your development style at all?

The only major difference was a shift in attitude away from trying to do everything myself. I still do the majority of art and code in the game, but I�m much more willing to hire specialists who are more talented in specific areas. When you have a �jack of all trades� mentality, it�s sometimes hard to recognize times when you are not the right person for the job. This shift will become more obvious in the game and on the website in the coming months. I should also take this opportunity to thank Jeff, who helped me with the development and modeling of the new player characters; and James, who has provided programming advice over the last few years.

What�s the future of Sherwood Dungeon? Where are you taking it?

Right now I�m focused on improving the visual quality of Sherwood Dungeon and MaidMarian.com so that it really feels like you�ve stepped into a fantasy world from the moment you get to the website.

Thanks for your time in chatting with me. You�re an inspiration to indie developers all over the world!