I've been posting short development planning and sprint review notes on the community forum. So far those notes don't have many views, and may have been too brief and cryptic to make much sense to anyone but myself. It's been I while since I posted any substantial development updates, and as I'm getting questions and feature request it's time for a longer dev-log post.
First, a few words about my background:
I grew up in Stockholm, Sweden, and have always been interested in games and computers. Learning programming early on, I enjoyed writing small simple games that I could play myself or try out on my friends. I started working on a real game (Mission Starlight) with a friend while studing Computer Science at Stockholm University.
This was in the ancient pre-internet era, so we had to send out paper-based emails (aka letters) and physical demo disks to publishers in order to get it released commersially. Although retaining my interest in game development, I've been working professionaly as a software developer and architect on many different kinds of non-game related projects. But now I'm back to games again, working full time on the Battle Match - Samurai Wars project. Being a strong believer in lifelong learning, I've also gone back being a part-time student of mathematics at the university. In addition to the fascinating beauty of mathematics, I find it very useful for game development: Linear algebra and analytic geometry for 3D graphics, calculus, numerical analysis, and statistics for battle simulations, graph theory for network programming, game theory for balancing, the list goes on.
Having played all sorts of games, I eventually developed a taste for games where there is a clear rock-scissor-paper (RSP) relationship between the pieces and where intelligence and tactical decisions matter. Where a relative simple set of rules produce a complex and rich game play. There are many abstract games in this category, chess being one of the most popular. Then there are more realistic tactical wargames such as Squad Leader, which I enjoyed playing.
These board games are based on the mechanics and the tactical aspects of abstract games, only with far more detail and realism. But the complexity of the rulesets, which you have to learn and apply my hand by throwing dices and calculating the results, makes them somewhat inaccessible to anyone but the hardcore wargamer. I wanted to dive deeper into accessible realism and turned to computer games.
There are two computer games in particular that have inspired the development of Samurai Wars. The first one is Myth: The Fallen Lords (and its sequel Myth II: Soulblighter) developed by Bungie and released 1997.
It had several innovations that set it apart from other strategy games. There is no base building or resource management. Instead you start each level with your army already assembled and ready for combat. In single player, you have to use tactical skill to fight your way through a story driven series of engagements. It has a great fantasy-based narrative and the graphics, music and sound effects really sets the mood. Visually, its landscape and buildings are made up of irregular polygons, giving a non-artificial and natural look (not common to wargames at the time), enhancing the feel of realism and suspension of disbelief. At the core there is a real-time battle simulation with hundreds of free-moving individuals, realistic physics engine, and units with understandable RSP relationships. It also had a multiplayer mode and a devoted community. Despite the smallish scale of the battles, it was doing so many things right.
Then there is the original Shogun: Total War which was released in 2000. In contrast to the Myth series, it had the right scale: thousands of individually simulated soldies on the battle field (not mere tens or hundreds). Realistic physics engine, and it modelled psychological aspects such as fatigue and morale too. Yet, commanding the units was no more difficult than playing chess.
This was it! This was the game I had been dreaming of. Winning the battles were difficult at first, but after completing the campaign a couple of times, having learned all the units, I had become an invincible tactical master. Or so I thought... So far I had only played the single player mode, and it was time to try my superior skills against other humans. My first online battle was a complete failure and a total humiliation! After a hundred or so battles I wasn't totally losing. After a thousand battles I was doing ok, or at least above average.
The eye-opener with STW was the sharp contrast between playing solo and playing against other humans. Not only the difference in tactics, and the human ability (or inability, depending on skill) of reading the battle vs the computer. And the sometimes annoying unpredictability and irrationality of human allies and opponents. Even more important are the social aspects, the community of players, the in-game chat and taunts, the clans, and the competitions. Solo play is has its place and is important too, but for any game with symmetric gameplay, multiplayer should be developed as the primary mode of playing, not as an afterthought.
As I continued to play games in the real-time tactics genre, I more and more felt that they didn't take advantage of the opportinities of online play. It was obvious to me that you could do so much better. So I started experimenting and prototyping such a game myself. After some failed initial attempts at a Java-based game, I managed to put together a first rudimentary 2D version for iPad.
As a programmer, I enjoy challenging myself with difficuly technical problems. Combining the complexities of multiplayer networking, 3D graphics, and a real-time tactical battle game with semi-realistic simulation was the perfect fit. I did not have a clear goal, but I continued to research technologies and design solutions, and kept adding to the game on my spare time because it was a fun and relaxing activity.
The community has been very important to the development of the game. I'm very thankful to all the feedback that helped shape the game to what it is. But with any online community there will be troublemakers. I did not realize how important and how difficult it is to properly moderate a community. And that you need to have right tools to be able to do that.
So I learn a lot, but also made a lot of mistakes. The architecture was in some respects good and sound, but in other respects too inflexible or too complex and difficult to maintain and develop. This is a common problem in software development. On the first attempt you may produce something that solves the problem. Along the way you gain deeper and deeper insights into the problem, and eventually you realize that the solution you have is naive and doesn't account for the more subtle aspects and relations between the concepts.
The system I had developed didn't scale and eventually it simply became too expensive to host the servers. Moving to a new job with little spare time for hobby projects, it came to a point where I needed to take a step back and reconsider what to do with Samurai Wars.
There was a period of collecting my thoughts and ideas. The available technology stack today is considerable more mature. Many of the crazy ideas I've had, that were difficult or impossible back then, have now become easy to implement. I now had a clear picture of how to reassemble and rewrite the code base. Encouraged by players who contacted me about relaunching the game, and having secured financing, I eventually decided to embark on a new and ambitious project: Battle Match.
Battle Match is the name of a project and a platform consisting of (1) a battle simulation engine and (2) an online match service.
The battle engine is designed to be a general physics-like simulator. The unit types, such as yari ashigaru or katana samurai, are defined in what is called a "module" (mod), separate from the engine itself. Samurai Wars is the first such module, reimplementing the units and rules from the good old game we all know. There will be several other such official modules covering various time periods. And there will be tools for creating your own mods and maps.
The match service, like the battle engine, is designed in a general and agnostic fashion. All the rules for how matches are set up and resolved are defined in what's called a "scenario". The Samurai Wars scenario will provide the familiar lobby with last-man-standing matches. However, anyone with a little web programming knowledge will be able to write custom scenarios and victory conditions. It will be possible to do competitions, ladders, campaigns, etc, as well as single player scenarios with AI opponents.
It's important that the project is financially successful to support the development and operation of the servers. So while the battle simulation will remain free to play in sandbox mode, the match service will require the purchase of a Battle Match pass.
Discord will be used for the in-game chat, and will be seemlessly integrated with the game. This way I dont need to reinvent the wheel regarding community functionality, and it nicely solves the moderation aspect.
Planning software development is hard. I'm getting better at it, but things still often take twice the time you expect. Anyway, moving forward, the general plan is as follows:
The basic feature set of battle engine has been in development for almost two years and is now getting in shape. There are many little details in need of attention, and this will be an ongoing activity during the Early Access period.
Early Access will start once the online match service is up and running with functionally equivalent to the old version of Samurai Wars. This will be the focus for the coming sprints. Maybe not done before the end of this year, but definitely early next year. The EA will simultaneously be released on Steam, App Store, and Play Store.
During Early Access, I'll work on the map and mod tools, and the module and scenario support. It will stay in EA until this functionality is complete, too early to say when, but current estimate would be end of 2026.
Development of new features for the battle engine, like new weapon systems, new terrain features, improved algorithms, etc., will be an ongoing activity, both during and after early access. I have no plan for when to do what, but will select appropriate features to implement at the start of each new development sprint.
As mentioned, there will be other official modules in addition to Samurai Wars. This will be towards the end of EA, and the preliminary plan is something WWII.
This short summary is an overview explanation of what this project is. Maybe it raises more questions than it answers, I don't know. In time, further down the road, I hope things will start making sense.
[EDIT] When I started this project, the world looked very different. The recent changes with agent-based development has turned everying upside-down, and old development processes are no longer relevant. I've temporarily paused the project in order to evaluate the possible paths forward.
/felix