Microcurrency on Puzzle Pirates has turned out more lucrative than a subscriber system (a pay-for-items model). They wanted to self-publish, despite their encounter with Ubisoft. They've 'sworn off publishers'. The goal was for a coherent experience for the new game, rather than a game that's never 'quite' finished.
The development process started with the idea of doing a game about Bugs. The gameplay was fun but the bugs were proving to be nonsensical. That became a more typical 'war' strategy game, with a mix of real-time and turn-based. The title "BANG! Howdy" is from the movie "The Party", a scene with Peter Sellers. They took the game to a 3D level (can't focus on midwest housewives all the time).
Their development tools consist mostly of their self-developed tools (now open-sourced), more traditional open-sourced resources (MySQL, etc.), and a little bit of commercial tool use. More than half of the tools used on BANG! Howdy are open-sourced. The game has an online multiplayer component, and they spent a considerable amount of time on the backend server stuff.
Release Plans: 1. Collect underpants. 2. ? 3. Profit
An alpha testing period will follow GDC, with a full-on public Beta test starting up sometime next month. The game will be free, with certain elements of the game available for-pay. There are two kinds of currency in the game: Soft and Hard currency. You can earn soft currency in game and transfer that to hard. They'll be selling the hard currency after the game 'launches'. The game is organized into a series of towns. Everything is modular, so they'll be adding new content all the time.
Demo of the game ensues. You can buy packs of collectable trading cards, and fancy hats. Lots of revenue for Puzzle Pirates comes from customizing avatars. Annd the demo crashes when they try to get to the actual game. They're running the client and the server on the same little Linux-running laptop, so it's understandable.
They switch over to talking about Puzzle Pirates and the virtual item sale model. They feel it's not only more lucrative per customer but more attractive as well. Make em' pay for stuff, not for playing. Their free-play/pay for stuff servers are their most popular. They are peaking at about 2000 people per server for concurrency, with about 4500 people across all the 'oceans' at at time.
Growth in the Booty! A revenue per day graph. They're bringing in about $7,000 / $8,000 per day with their pay-for-items business model. Their subscription growth has definitely leveled off, with the money brought in by the 'doubloon' model is continuing to grow in spades. They generalize their subscribers by taking their total profit and dividing by 8 (the amount they make from a subscriber). With that rule of thumb, they've got about 35,000 subscribers.
What's the money being spent on? While some money is spent on housing and basic stuff, most is spent on badges and items. Badges give you in-game privileges (like allowing you to be a captain), but have a limited timespan. Items wear out. Thus, they're mimicking the subscription model by having reoccurring fees.
There are a ton of people, given their graphs, who do not pay the company but still generate money for the company. They're making items, which paying customers fork over doubloons for, and thus pay the company. They estimate that players sell their time for about $.25 per hour. They have had some problems with fraud, but have put in fraud control. They had one kid wrack up $1200 in doubloons. He sees this as a more honest business than subscription: Subscriptions require no effort and thus trawl in money. If they're buying stuff, they're actively playing the game.
They try once again to fire up their new game while they answer some questions. One question is a blunt "How much are you making?" Cost of goods, server depreciation and server bandwidth ($7k-$10k per month) and customer service are their real expenses. Everything else is paying the employees. Their profit margin is between 75-80%. Three Rings has been profitable for about three months. (Congrats!) They've gone up to about 32 total employees between office workers and at home. They are editing code on the screen in order to make the demo work. One of their keys is allowing as many revenue streams as possible (including SMS in Europe). The mention Habbo Hotel as an inspiration.
The demo almost works, and then fails again. More editing of code, followed by the resurrection of the game. "Two Ponds" board, with some claims to jump. Each unit has a timer, and you move the unit across a set of squares. A 'semi-realtime' feel. The object here is to eliminate opponents and collect gold. (Claim Jumping) Bonuses within the gameworld try to balance the current match. Whoever is behind is most likely to get a bonus dropped by them. Trying to balance the title ala Mario Kart DS. Gameplay is quite fast, it seems. Besides the claim-jumping, there's also a 'cow-branding' mode. You brand cows, and your opponent can brand cows back. Sort of a 'node-holding' type of title. 'Gold Rush' is like claim jumping, with players going for gold nuggets, but is more of a race to get your loot back to your claim. Each new town will introduce new unit types and new gameplay types. You win soft money by winning games.
They round out the session with some more Q&A. They chose a downloadable client to go for a more tech-sophisticated audience. They're also working on a Flash game, to go for the other extreme of the market. The engineers are frustrated with ActionScript, but it's "almost" a real programming language. Always an interesting presentation from the swashbuckling crewe at Three Rings.