George Moromisato
28 December 2013

In 2013 Kronosaur Productions recorded revenue for the first time in its existence. Of course, revenue does not imply profit, and the latter is what we need for the company to be a going concern. Nevertheless, accruing revenue is an important milestone in the company's history, and for that I wish to thank all the players who supported us in 2013.

2013 was a good year of progress for us, and in this annual report I'll write about the current state of the company, the progress we've made on our games, and our plans for the coming year.

The Business Model

Now that the company has revenue I'd like to write a bit about the company's business model. To be sure, I haven't yet had enough experience to refine or even validate the current business model, but it's helpful for me to at least articulate how things might work in theory.

Kronosaur Productions sells software, so it is tempting to follow a software company's business model. The wonderful thing about software is that the marginal cost of production is close to 0. Once the software has been created, selling 100 copies costs about the same as selling a million. Each additional copy costs almost nothing to produce. That's why profit margins on software are so high and that's why Microsoft became so successful: It was the first software company to figure out that it was better to sell 10 copies of Excel for $99 each than one copy of Lotus 1-2-3 for $495.

But video games, though implemented in software, share many more traits with the movie industry. For one thing, there are thousands of different video games out there, all of varying levels of subject matter and quality. And you've never heard of most of them. The hard part in a video game business model is attracting customers, and it turns out that attracting customers costs money in proportion to the number of copies you want to sell. The marginal cost is, unfortunately, much greater than zero.

Given that reality, there are two viable strategies: The first is to attract as many people as possible for the least amount of money per player. One way to do that is to create a game that spreads virally through word of mouth. Minecraft is a great example of this. The second viable strategy is to keep the players in the game for as long as possible (so as to amortize the cost of acquiring them in the first place). World of Warcraft and EVE Online are examples of this second strategy: charge a subscription fee and keep players as long as possible by adding more content and enhancing the game.

The problem with the first strategy is that it is almost impossible to replicate. Everyone wants to be the next Minecraft, but none have managed it. What's worse, sometimes you can stumble on a viral hit (like Draw Something) only to see it fade in popularity after a few years (or even a few months). After that, you have to come up with another viral hit to survive. (This is why movie studios are always in crisis: they are constantly one flop away from bankruptcy.)

I very much prefer the second strategy of keeping players around. Unfortunately, the problem with the second strategy is that it requires a lot of investment. One doesn't simply create EVE Online overnight. My theory, however, is that I don't have to do it overnight. I believe the number of players we attract will be roughly proportional to the investment in the game. As the game improves over time, more and more players will be drawn to it. [Note that this gradual accretion model would never work for the first (viral) strategy. The first strategy only works at scale.]

In short, Kronosaur Production's business model is to create games that players can be passionate about, not just for a few months, but for years. As the games get better and more elaborate, the player community will grow and support a greater development investment, which in turn will attract even more players. The exact way of funding development is less important. Regardless of whether we sell subscriptions or game pieces or expansion sets, the important thing is to create games that players can be passionate about for years.

Will this business model work? In thinking about the kinds of games I like to create I came up with three adjectives:

  • enduring: games that grow and evolve over decades,
  • elaborate: games with richness, depth, and detail, and
  • extensible: games that can be enhanced by the player.

It's a fortunately coincidence that these are exactly the kinds of attributes that will (hopefully) create passionate players. But a lot hinges on marketing: spreading the word about the games and convincing people to try them. My plan is to spend much more time in 2014 marketing our games. Between the new trailer for Transcendence and a great PR plan for contacting various reviewers, I believe we're starting 2014 in the right trajectory. All I have right now are theories. By the end of 2014 I'll be able to report some hard numbers.

Corporate Command

The big news in 2013 was the release of Corporate Command, the first paid expansion and the first expansion available on the Multiverse. Finishing the expansion took longer than I originally thought, partly because of the need to implement new Multiverse features (payment processing, distribution, etc.) But I'm happy with the result and I believe that Corporate Command will serve as an example for future expansions.

Corporate Command went on sale on November 21st and by the end of the year had sold over a hundred copies. On its current trajectory it should sell between 300 and 500 copies in 2014, though I hope that a greater marketing effort will increase those numbers. While the company is still a long way from being profitable, I am very encouraged by these numbers.

My plan is to update Corporate Command sometime in the first quarter of 2014. There will be lots of bug fixes, of course, but hopefully also a few new features. The update will be automatically distributed to anyone who purchased a copy.


The release of Transcendence version 1.2 was another big milestone for 2013. Though most of the changes in 1.2 related to the Multiverse, there were several game-play changes that I'm happy with. The updated graphics were, I think, a big hit with players and I intend to add more such enhancements in the future. In addition, I believe people appreciated the changes to the Tinkers, communications UI, and the introduction of analyzers.

I am also really proud of Wolfy's Osaka expansion, the first third-party expansion on the Multiverse. In 2014 I expect to have many more such third-party expansions to populate the Multiverse catalog.

In 2014 I plan on releasing at least two minor version updates. Michael Tangent, who composed the current (and past) Transcendence music, has been working hard on an extended soundtrack for the game. I've heard some early cuts and I'm hugely excited to hear the finished version. Sometime in mid-2014 we'll release the full soundtrack as a free extension to the game. Moreover, we've been working on some new APIs to support your own soundtrack extensions. I'll have many more details on this next year.

Also, at the end of 2014 I plan on releasing a minor version to coincide with a new paid expansion. Tentatively called Near Stars, the new expansion will finally reveal Earth, Sol and the surrounding star systems. I'm still planning out the details, but I would like this expansion to include both a stand-alone adventure (with a new player character who is not a pilgrim) and an expansion for Stars of the Pilgrim (like Corporate Command). There is much to be designed yet, but I am excited about the plans. You'll hear more about this next year when we're closer to release.

There are two projects that I'll be working on in 2014 that will not be ready next year. The first is Part II of the Domina & Oracus saga, which I've titled The Vault of the Galaxy. You've all been patiently waiting for Part II and I'm sorry to say that it will not be released in 2014. However, I do plan on making substantial progress in 2014 for a tentatively planned release date of 2015.

The second project is a set of core engine changes to support an RTS. I would love to create a game based on the Transcendence universe (and with the same core engine) using RTS mechanics. The first game on this engine, tentatively called CSC America, puts the player in command of the eponymous Star Carrier with responsibility for deploying its complement of Centurions, Britannias, and other ships to complete various missions deep in Ares Space. Again, this game will not be released in 2014, but I hope to make enough progress to possibly release something in 2015.


I didn't get to work on Anacreon as much as I wanted in 2013. The year started with several good releases, but by the end of May it was clear to me that I needed to devote more time to Corporate Command and the Multiverse infrastructure. Still, the current game has held up remarkably well, and I am confident that we have a solid foundation to build on. Here are some statistics about Anacreon:

  • Right now there are about 30 to 40 active player empires in the current game (compared to about 20 at the beginning of the year).
  • The current game has been running for more than half a million turns (each turn takes one minute of real time).
  • The largest empire rules 860 worlds and a total population of almost 7 trillion souls.

In the coming year I plan to work on a few key initiatives:

First, I want to enhance and balance the ways in which empires interact with each other. Right now, military conflict is the only avenue for interaction, which becomes rapidly frustrating when a powerful empire attacks a weaker one. Adding trade and diplomacy mechanics could encourage different kinds of interactions and competition. Ideally, every empire in the game will fill a viable niche, allowing it to thrive and prosper.

Second, I would like to add more variety and activity to the galaxy. Right now there is not much to do in the game other than conquer and be conquered. There need to be more events to make the game interesting: AI empires that need to be stopped; cataclysms that require players to band together; quests to find ancient and powerful artifacts. I aim to create a living galaxy filled with adventure.

Lastly, I plan on experimenting with various payment models. The core game will of course remain free, but I would like to figure out how to sell either additional expansion sets or scenarios in order to fund development. My plan is to be patient—getting the game to be fun is more important than extracting revenue—so I expect these experiments to be small-scale.

We're still at the beginning of this incarnation of Anacreon, but I hope that, for those who have stuck with the game for so long, your patience will begin to pay off in 2014.


Last year I talked about the development of Hexarc, our home-grown server infrastructure. It's still too soon to tell whether creating our own server was a good idea, but I'm happy with its performance so far. The Multiverse is implemented entirely in about 10,000 lines of Hexarc scripting code (and about 5,000 lines of JavaScript). Though I've had to fix various bugs during 2013, the core server has handled the load remarkably well.

I am particularly happy with the database system, called Aeon, which has survived power-outages and drive failure without data loss or corruption. Just as importantly, I've been able to add specific features to Aeon to support the kinds of tasks required by the Multiverse.

Anyone who's programmed databases knows how hard it is to deal with record consistency when there are multiple people touching the database at the same time. For example, imagine that I need to increment a value in a record. The traditional way is to read the record, increment the value, and then write the record back. The problem is that some other process might modify the record after you read but before you write. If that were to happen you would clobber someone else's update. The traditional fix is for each process to lock the record before reading and unlock the record when done. Unfortunately, a in client-server system, in which your connections might be lost at any point, you can't always guarantee that the unlock will ever happen. This adds massive complexity both to the database (which needs to keep track of lock timers) and the calling process, which must handle lock failure.

Aeon deals with this problem differently: it exposes more sophisticated record update semantics, including increment. In a single call, a process can ask Aeon to increment a field and Aeon will be responsible for atomically modifying the record regardless of how many other processess are trying to update it. Pushing the problem to Aeon makes the task much easier for the calling process. And it also makes it easy for Aeon, since it has full control over when it needs to lock (and unlock) a record.

Aeon has special record update semantics for increment, decrement, min, max, and many others. For example, there is a way to update a record if and only if a given field is greater than the current stored value. This is how we implement the high score table, for example.

There is, of course, much work left on Hexarc. I need to add support for SSL/TLS (encrypted communications), multi-machine configurations (for scalability) and dozens of programming enhancements. I did not get a chance to release it as an open source project last year, but that's still my plan (once I clean up some of the configuration and installation code).