Gamasutra - Features - "Postmortem: Nayantara's Star Chamber"

Major tech licensing issues - I had to switch back-end technology ten months into the project. This hurt badly, and was the biggest impediment in trying to reach my one-year estimate of getting the game out. The back-end technology I was planning on using was the one powering Digital Addiction's Sanctum. When Digital Addiction went bankrupt, it was forced to sell its assets, and the Sanctum game was bought by a group of players who formed a company called NIOGA (Non-profit International Online Gaming Association). When I first started work on Star Chamber, I had discussions with NIOGA about using the Sanctum technology, and it seemed likely, at the time, that they would be willing to let me license the technology for a relatively small fee and a small percentage of future sales. Somehow, though, over the months, after I was already building hooks into the client to use their back-end (with their blessing, assuming a deal would be worked out), the numbers that were verbally discussed changed into numbers that weren't palatable to me. So not only had I wasted some time writing code to hook into their server, but now I was on my own in terms of coming up with appropriate substitute technology. However, there were several silver linings: instead of forcing my game to conform to their more limited tech, I could go on to design a server that would have more and better features, my new server would be written in a language (C++) that I could maintain and upgrade more easily, and I would own it without having restrictions.

Trial play is a double-edged sword - While try-before-you-buy is a great opportunity for certain consumers and for the company when those trial players convert into paying customers, it is arguably a bad business model, for a couple reasons. With a small player population and a multiplayer game, new trial players are likely to play not against other trial players, but against veterans. The veteran will usually defeat the newbie, and many newbies will be turned off from the experience, because they feel they're diving into a pool of sharks. They would simply prefer to play against other newbies, or learn by themselves. I talked to a veteran producer at a games publisher after we launched, and he suggested that demos and trial play can actually dramatically lower the sales of your game. With regard to the above problem, it may be that novice players who have already paid for something will be much more likely to "stick it out" in a situation where they have some skill ramping to do, in order to learn how to actually play the game.

Single-player issues - I didn't think enough about single player, and this is, perhaps, where my focused vision backfired. I didn't think a single player element was necessary at all. Indeed, the model of giving away the client for free and only selling digital cards doesn't really support putting effort into a solid single player experience. But when we went into the beta test, it became apparent that the players wanted a single player mode, even if it didn't have a good AI. This way, the players would be able to learn the game and practice on their own, without the nuisance of having to find an online opponent, and perhaps because of pride and embarrassment issues. So while we did end up launching with a Practice Mode, allowing players to test their decks against a dumb computer opponent, not having a better single player mode probably hurt sales of the game significantly.

Not enough eCommerce options at launch - We launched with PayPal support only, and without credit card support. We were in that state for about two months, and were in the unfavorable position of losing potential customers who didn't want to use PayPal. We did have card resellers who would buy packs of cards with PayPal and then resell them to players, but many casual players don't want to be bothered going through that kind of effort in order to buy our product. When our resellers lost their ability to accept credit cards, it became critical that we implement credit card acceptance.

Insufficient load testing before launch - We didn't get enough of a chance to load test before Star Chamber's official release, and as a result, we didn't find some crash bugs with the server when certain unlikely conditions were met. In late December 2003 and early January 2004, after we received some good press, the server started crashing erratically as the number of players online had reached an all-time high. This resulted in a lot of lost sleep and a frantic search to find and eliminate the bugs.

While we did write a test client that would connect to the server, chat randomly, create games, and then quit out, our tests weren't thorough enough to find all the bugs with the server - a fairly complex multithreaded application. We should have put more thought and energy into this, as the instability was a big annoyance to our players.

Unexpectedly long development time - Morale and motivation dropped dangerously low because of the working environment and the longer-than-expected development time. For more than two years, I was working alone in the backroom of my house. While I worked with contractors for a good year, they worked from their homes hundreds to thousands of miles away from me. Without a spirit-boosting buddy system available, I went through some dark and lonely times, and was tempted to give up several times during the long course of development. Thankfully, with the support of family and friends, I was able to finish the game.

The making of Star Chamber was a brutal, draining ordeal; I was silly and naïve to undertake such a project by myself. If I had known it would take more than two years from start to finish, I would never have started. I can recall reaching the alpha test point in September 2002, and how it felt like I had been working on the game forever. Even then, I was overambitious in my expectations of getting to and through beta test in order to release the game. I imagine this doesn't apply to certain categories of games, but for a CCG/strategy game, do not underestimate the amount of testing that your title will need to go through. The Star Chamber beta test began in late December 2002 and continued until the end of May 2003. Even when the beta test concluded, I wasn't ready to release the game. I had been working so hard during the beta test, making game balance changes as well as interface and production value improvements that I was completely burned out mentally, and I hadn't even finished all the preparatory work to actually sell the product online. After the beta test ended, I had to take a mental health break, and I put the finishing touches on the game during the summer. In November 2003, the game was finally ready for release.

While I'm quite proud of the game, and it has achieved some critical success (Star Chamber is currently rated #74 on GameSpot's Top Rated Games list), Star Chamber cannot be called a commercial success, at least not up to this point. While it has attracted a small following of players, it hasn't come close to the potential that I thought, and still believe, it had. Whereas high-budget games count subscribers in the tens or even hundreds of thousands, Star Chamber has not yet reached 2,000 paying customers, and has never had more than 250 simultaneous monthly subscribers (subscriptions are optional, and you can play forever without ever purchasing a subscription).

However, what we have achieved has been with an extremely limited marketing budget (online advertisements only), and no distribution assistance - we're still looking for publishing partners for the game, if interested parties would like to contact us. However, we continue to support the product with tournaments, updates, and releases of new card sets - the first expansion, Incursions, was released in July of 2004 and we are currently working on a second expansion due out this winter. Overall, the team still believes in Star Chamber, especially as it offers a unique gaming experience with little competition in its field, and, despite the ups and downs of development, are delighted with the final results.

______________________________________________________

[back to] Introduction