I remember cracking the seal on my first COIN game – Colonial Twilight – and seeing that bot chart in the box. I was a relatively new wargamer, and Colonial Twilight was not only my first COIN game but also my first P500 (much to my wallet’s chagrin, it was not my last!). After playing a learning game I quickly took out the bot and began to play a Solitaire game. I was floored by the ability of the bot to play better than I could, teaching me the nuances of the game. While I recognized the limitations of the bot, I was quite enamored of the experience and satisfied with the playability, usability, and skill shown in the design of the bot.

Fast forward a year, and I am sitting at my gaming table playing Volko Ruhnke and Mark Herman’s Fire in the Lake and puzzling over the NVA bot’s March. Naively, I assumed that my experience with the Colonial Twilight FLN bot would translate well to Fire in the Lake. This is not the case – the FLN bot in Colonial Twilight plays one of the simplest factions in all of COIN (mechanically, not necessarily strategically) and therefore makes a most excellent automated opponent. Fire in the Lake, on the other hand, includes the most complex factions in all of COIN (again, speaking of their mechanics) which makes the bots significantly more complex.

As a piece of design, I admire the Fire in the Lake bot charts greatly. They owe much of their design to Orjan Ariander, to whom I will forever be in debt for blazing a design trail that we (Bruce Mansfield and I) are following. Indeed, the Fire in the Lake bots are formidable opponents! They appeal to a limited number of players due to their complexity and approach to automating the game. These bots rely on sequential and nested instructions, and keeping them all in your head can be difficult on the more complex operations. Yet, these instructions are what make the bot competitive and create those moments when, as a player, you yell “How did the bot know I was going to do that?!”

When developing Gandhi, the entire design team – designer Bruce Mansfield, lead playtester Scott Mansfield (yes, they’re brothers), and myself – spent a good deal of time wondering if there was a better way to design a bot. Better, in this case, meaning a bot that played as well (or nearly as well) as the existing flowchart-based bots, but with a more accessible form of instruction. From this seed we grew the Arjuna system, with a card-based operation selection and a visualized space selection table.

After Gandhi was released, we immediately saw requests for this system to be applied to other COIN games. Our first question was “Is that even possible?” Gandhi was purposely designed to be easier to play with the bot, while the other COIN titles were, for the most part, designed and then refitted with Solitaire bots. One thing I’ve learned is to never tell Bruce Mansfield what he can’t do; not a week after our conversation, Bruce called me and told me he had figured out how to make it work in broad strokes.

So the design team went to work. And work it was – we talked with Volko Ruhnke to understand the design choices that went into Fire in the Lake, played about 100 games of Fire in the Lake between us to internalize the strategic choices, read all the event cards carefully (multiple times) to determine how many specific event instructions we would need, and analyzed the adjacency and layout of the map to try to identify the right priorities for each type of action in the game.

We are currently deep into playtesting and expect to wrap up in the next month. The hardest part lies ahead – polishing the rules and making sure that they are clear and concise while still covering all the different situations that can arise in such an epic and sprawling game. We look forward to sharing Trưng with you soon!

---

Previous Article: A Sneak Peek at Tru’ng: The New Fire in the Lake Bot System