I just finished reading The Icelandic Model of MMO development on Gamasutra about EVE Online and I can say that you can probably take the whole article and apply it to Golemizer. Well not the part about the “subscriber curve” and the “bare $2.6 million in funding” but everything else fit quite well.

When I decided that I was going to build a sandbox game I did so because first I fell in love with the concept while playing SWG (the initial version) and because I knew I wouldn’t be able to produce enough content to provide a nice linear progression. I could have decided to build all kind of games but I wanted to build an MMO so I went with a sandbox game.

I like to think of sandbox games as a games where there’s nothing and many things to do. I’ve been asked over and over what’s the goal of the game from people who were trying to understand why I was spending so much time in front of a computer. Since I had about 1 minute to explain the concept before feeling I was boring them I was just saying “it’s a place you live in”. If you start playing and expect to be told what to do (not talking about a tutorial here) then you are already gone looking for something else.

When it was released, EVE could barely be called a game at all. It was a pure sandbox design, with only a smattering of content. The learning curve was vertical, the tutorials a disaster

You can say the same thing about Golemizer. Well there was no tutorial at all. I think I had 2 or 3 NPCs telling players that they could learn skills, create golems and build houses. It remained like that for a long time and yes it did hurt the game by turning players almost immediately. But! … some players started playing anyway! That was really amazing. Those players were really hardcore explorers not afraid to dive deep in something that had no documentation at all. Yes I could have held the release for several months and improve everything but I was working on this project for 1 year alone and I seriously needed to release something before going mad or something.

So there was nothing in the world, no help but plenty to discover, build, try. It gave the game a chance. The initial plan to add some kind of “real” content was to have a permanent war between guild around the concept of “guild golems”. Those golems couldn’t be controlled or packed. The were always in the world accumulating points for their guild. Guilds would then try to find and destroy those golems to claim the top of the ranking chart. Well that concept never worked. First because of many bugs, exploits (first game remember) and because the game was so huge (and increasing each day) that such hunt was impossible.

Features Over Linear Content.

That’s what I did. I dropped the idea of having some kind of “end game” and went with a full sandbox approach. The year following the release of Golemizer I pushed nearly 30 updates adding many new features like quests created by players, clothing, weapons, blimps, new golems, new items to craft, improved city management and so on. That was a crazy year and I still think that it’s what Golemizer was meant for. More features over and over. That’s just a bit difficult to pull alone.

At one point I did built a tutorial (which I upgraded later 2 times) and some “special” zones that new players could play in while figuring out what this thing is about. I built some quests with complete custom AI before the quest system was implemented. It was so painful to build that I think there’s only a dozen of them (still active). I wasted a lot of time trying to build this mini content because I was in some kind of panic. People were creating accounts but were not playing at all so I somehow convinced myself that with some traditional content to start with that it would help. That was just a mistake as once players went through that content many were left to wonder where was the next part. I just didn’t dragged them to the sandbox mindset even though I thought that was the way to do so.

There were however 2 big “traditional content” release: Diliculo Island and Demon-Hammer. Both of these were completely built by GMs. I just added the new data required and assest to achieve this content. It was a huge task for them and one that was impossible for me to do. It was also a huge task for me as this content was a way to unlock new recipes for players (new items and golems to craft) so while GMs were working on the story I was working on the features.

The problem with features is that when they start to interact together it can become hard to predict what will be the result. For example when I added the blimps in the game I had to handle many problems like players being able to go to places they were not allowed to go, prevent some actions (like cutting wood from the air) and so on. And when you add new features too quickly one after the other there’s the danger of not being able to consider properly the problems with these features. A stupid example is when I added the possibility for players to drop signs and change their content for other players to read. I didn’t make sure to filter dangerous tags so it was a matter of hours before a player dropped a sign that was changing the password of other players without them knowing about it. It was such a simple feature to add that I must have mixed it with something bigger and didn’t thought much about it.

Unique, Engaging Environment.

Golemizer surely have a setting that’s unusual but maybe a bit too much. Not because the theme is too niche but because it’s more work. if I was to start working on a new sandbox MMO it would surely be a space game like EVE Online. Everything gets simpler because you don’t have to worry much about collisions and pathfinding, it limits the kind of environment you’ll have to face (water, underground, forest) and the assets required can be much more limited (did you know that in Golemizer you could craft windows to decorate your house with …). No housing, no cities … Of course I’m talking about a completely different game but as a solo developer I think it would be easier to come with.

I also think that a space setting would be more appealing than a game where you play a crazy scientist. Of course I’m not saying that current Golemizer players would like such game but it would be much easier to sell IMHO. It still remains a sandbox game so not for everyone but at least you start with something people can usually relate to as they have seen this over and over in movies and on TV.

Too many developers with great ideas are intimidated away from creating intriguing MMO titles by the daunting startup costs of following the dominant blockbuster model. It doesn’t have to be this way; just use a different business model!

Yep. That’s what I thought and still think. I always knew that a steampunk sandbox game would make Golemizer a niche game and I don’t blame this decision for not meeting my expectations. All I know is that if I would have tried to come up with a linear content game that it would have been worse. Not just because I would have never been able to come up with enough content to keep players busy but also because I’m not just the kind of guy that can write stories (even boring ones) in a few hours. I’m best at pushing new features and having a sense of what would be nice to add. That comes with the price of having a less engaging environment but as I’ve seen so far if you give the tools to players they will use it. I don’t mean here that ALL players will use them but if you build a game centered on that mindset you will attract the players that are interested to build content. All I know is that before GMs started to approve quests I was having a really hard time keeping up with quest approval because there were so many of them.

So did Golemizer burned me from sandbox games? Surely not! One day I’ll come back with more experience and build another one I’m pretty sure of that. You can bet some bucks on a space themed game.