Here is the first part of two-part development diary that is long overdue. Be warned, I’m going to really get into the weeds of New Foundation’s development and try to describe what we’ve been tangling with. This post is going to get pretty detailed and most of it is going to be about a system I ended throwing away. So, be warned: a darling is going to be killed at the end of this book.
I often boil down my design practice to two distinct phases. In the first, I work on developing a core syntax and vocabulary. In the second phase, I try to imagine what what sorts of stories players can generate. Often, I grade my success in terms of how expressive that game-language is. Does the game have the necessary nouns and verbs to tell the kinds of stories I want it to tell? Usually expressiveness is a good thing, especially for the sorts of interactive designs I like to work with.
Optimizing a design for this kind of expressiveness is a central part of my creative practice. Its been the guiding light across many years of design and development. And yet, with New Foundations, I found that some of my assumptions about expression actually undermined the things that made Oath work in the first place.

Over the past two months, I’ve been guiding New Foundations through a very difficult phase of development. While many parts of the game are finishing basically on schedule, certain specific elements were not working. And, the harder I worked on them the further from completion they seemed to get. I believe my essential problem was that I had failed to adjust my own design and development sensibility to match the demands of the project. Techniques that had proved extremely useful in Arcs and Molly House simply weren’t well suited to Oath.
At the same time, these many long weeks in the wilderness were critical to the design. By letting the design balloon, I learned critical things that showed me what should be cut. I can’t imagine I could have gotten the design to where it is without this period of expansion and ambition. So many otherwise promising designs suffer from premature development. If you’re a designer just starting out, know that often it’s best to shoot very high and then start fresh than to preemptively clip your wings for fear of a development liability. The key thing, at least for me, is to never get too precious with one articulation of your ideas. The core promise and ambition of a game is always more valuable than whatever you happen to have on your table. Never be afraid to start over.
I’ve wanted to write that last paragraph for weeks now, but it didn’t feel earned. I was caught deep in the tumble of iteration and I wanted to make sure I had something to show for all of that rough-and-tumble work. The same things that I wanted to write about were proving most elusive. I didn’t want to write anything before I understood why I was having so much trouble with the design. I think I have those answers now.
---
One of my major goals with New Foundations was to make each and every empire feel unique. With only six edifices in the base game, strong empires were usually defined by the same sort of powers. It’s true of course that the selection of denizens in the game could go a long way in providing texture, I still felt like something was missing. If players didn’t put in the work to interact with the denizen powers, one big group of denizens could feel like any other.
To address this problem, I went down a long, long design journey. At the center of this journey were the govern tiles. The idea is that a stable empire would slowly generate a pool of tiles that they would carry over from game to game. If the denizens couldn’t give the empire a distinct character, maybe these tiles could? Additionally, by keeping the same pool of tiles from one game to the next, I hoped to give empires a greater sense of continuity.
But, Oath is already a complex game. At this level of complexity, the expressiveness of a design becomes a zero-sum game. If you add neat features to one part of the game, it’s going to pull attention from other (maybe neater!) parts of the design. So, as I went through various iterations, I did my best to make these tiles as simple as possible. By the time we got to the Kickstarter, I had a system that was working and was fairly simple. I suspected it would get simpler as we continued through development, and, indeed, that’s exactly what happened.
Eventually we had just six unique govern actions, one for each of the six suits. The empire still had character, but there were still too many rules in play. I had cut about everything I could, but it wasn’t enough. So, I looked for ways to deepen the implications of the govern tiles. In other words, I was trying to hide the complexity.
At one point during development, govern tiles gained an important property—when used at the start of the game, setup the Chancellor would assign them to cards to use their linked govern actions. This would also alter a that card’s suit. There were all kinds of neat implications here, and players could give the same old cards new resonances. But, if these implications were neat, they didn’t really guide the players or help them tell better stories. So, I put in an additional implication. At the start of the game, players needed to stock the initial favor banks, so it wasn’t too much extra work to have those favor banks reflect the cards in play. That way the placement of govern tiles would also inform the character of the game.

Without realizing it at the time, my design goals had started to shift. Now, rather than preserving and highlighting imperial differences, I found myself interested in the question of building an expressive core system that could organically create many different types of setup situations. With this in mind, I looked back at the 6 govern actions and thought about grafting them on top of the other setup phases. Maybe the six different favor banks could be linked to the essential elements of a game’s setup. Base Oath features a slightly asymmetric setup—that is, the Chancellor has their reliquary and gets an extra favor—maybe I could use this new system to explain why that was and give players the power to alter it.
The result was a revised setup system called Imperial Character. After govern tiles were placed, favor would be “sowed” on the various cards of the game. Then that favor would go to the banks along with whatever other starting favor started there.
Then, during the rest of setup, players would resolve each favor bank in order and the size of each bank would determine the setup features. If, for instance, the order bank was high, then the empire would have additional warbands. A high hearth bank might mean extra favor for the Chancellor? High discord might translate to extra exile warbands. It was neat!
The game was essentially treating the initial favor distribution like a set of attributes in a role-playing game. I could also add favor to the game or pull favor out, which meant that folks could find themselves steering a “low stat” empire through troubled times. Some of these attributes would benefit all of the players and others would just benefit the empire. This balance was critical. I wanted certain setups to favor the empire and others to favor the exiles. After a few weeks of testing, I had boiled these attributes down to a chart and cleaned up the setup procedure.

Though I had boiled things down as much as I could, it was still a place where the design had gained weight rather than lost it. At the time, I justified it by pointing to what we had purchased: the static elements of Oath’s setup had now been made dynamic. What’s more, they would respond to the world the players were creating. It gave consequence to the Chancellor’s governing. The system was responsive and expressive. I was pleased with myself and thought that just a little more experience and playtesting would allow folks to recognize all of the good work that the system was doing for the game.
So we started testing. At this point, about a year into the project (I started working on this project towards the end of January last year), a lot of the content was shaping up really nicely. We had a pile of new card powers, relics, edifices, and the like. We also had some new system-level adjustments that were, to all appearances, doing their job. And yet, despite the relative maturity of the project, I kept finding myself not engaging with the game’s world as much as I had engaged with the original Oath project. This was mirrored in my players as well. There was plenty to do, but players routinely reported not knowing what to care about or aim for in the game.
This was a very damning observation. The expansion was built on increasing player engagement, and yet, it was failing on its very first test. This was so upsetting that I initially disregarded these feelings. How could playtesters not care about the consequences of their actions?! There were so many new systems and the game had become so much more responsive than base Oath. And yet, I had players consistently suggesting new systems and ideas that would help them feel like they had an impact on the world. It seemed a variant of the classic developer’s nightmare: players find the game too complex, but would like you to add just one more thing to fix it.”
At this point, I decided to step back for a bit and think about how players engage with Oath. The goal of this whole project was to get players to more deeply engage with the game—to give players tools that would allow them to care about and care for the world they were creating together. I went to the white board and started sketching out the “layers” of engagement of players in an Oath chronicle.
The bottom layer was the game itself. How do you win? This layer was fine in base Oath and was fine in the expansion. Nothing I had done with New Foundations had compromised it. Though players now had other things to do rather than just win, there was still enough tension in the design. (I’ll write a bit about this more in the next diary).
The second layer was the immediate future. What did you do that would shape your next game? This was a place where base Oath was weakest. The new govern, edifice, shadow, curse, and setup system were designed to address this layer.
The third layer was the longer future of Oath. This was a place where things like the deck construction of the base game had supported a lot of weight. In New Foundations, there were lots of systems that engaged with this layer and they seemed to be working really well.
The final layer was the meta layer. How could players talk about the game when they weren’t playing it. This layer always more-or-less takes care of itself and is bounded only by the creativity of the players. Still, I had introduced a few new elements to the expansion to help players have more to talk about.
Writing this all out, it was very clear that the second layer was the most troublesome. The game didn’t need more rules or more multi-game incentive structures. It needed direct player feedback on that short-term horizon. And each of the systems at this layer was causing issues. In fact, during this same period, I had actually started to remove some of these systems from the design since they simply were not doing anything worthwhile.
The essential issue was that all of the work I did to improve system expressiveness had actually backfired. I had built a very responsive system, but it was simply too granular to provide the kinds of immediate feedback that board games excel and communicating. I needed a design that was as chunky as it was expressive.
So, over the course of about three weeks, I started going through the design and looked for places where the design was trying to protect the subtly of its systems. The first thing to go were the shadows and the curses. The more interesting shadow effects could be handled by the new edifice system and the curses were never as interesting as the base game’s vows. I was able to make some trait card adjustments to capture those dynamics.
But these cuts weren’t enough. Here I was spurred on by a new set of playtesters, both in office and on our discord, who were recent arrivals to New Foundations and almost immediately recognized the nature of the problem. As is typically the case, it took quite a bit of debates in our discord chat to clarify the problems, but it was well worth it. They helped me realize how much I was focusing my efforts protecting some clever and expressive systems rather than designing for impact.
Now, with a better sense of the problem, I turned my eye to the two parts of the game that had become almost sacrosanct: setup and the chronicle.
The setup system went first. What started as this:

Became this:

Here are the highlights. Favor banks still get to mirror the world, but they no longer inform the rest of setup and players don’t have to waste time counting and re-counting them. The govern tiles are now dead simple: the chancellor can discard them or assign them on cards they rule. If assigned to a card, they alter the suit. This has all of those usual expressive consequences (changing the bank, mixing up powers, etc), but the Chancellor doesn’t have to worry about them in the short term, because their other choice has immediate impact.
After all of the govern tiles are placed, the Chancellor draws 1 edifice for each tile that matches its card’s printed suit. There are 36 edifice cards, 6 of each suit. These can be kept in the box during the game, but, at setup, the chancellor will draw from them. Then, they will be assigned to imperial players. Each imperial player can have up to two (the remaining edifices are discarded). Edifices are always site-based cards with a lock. Let’s take a look at a couple:

As you can see, these cards let me fold in some of the expressiveness of the previous setup system but in a way that was fundamentally more chunky. This also provided me a place to put a lot more character into the game that could be linked to specific sites.
The previous adjustments to the site system had enabled things like locked cards to remain stuck to their site. So, I could use edifices to “link” persistent effects to a site that could stay there for the whole duration of the chronicle.
These adjustments dove-tailed perfectly with some of the continued development of the world’s geography and some of my feelings about the role of shadows in the game. The core theory of agency in Oath is that the players are the drivers of change. Where does evil come from in Oath? It comes from the players. No event or corruption system I could imagine would be as responsive to the game state as the decisions players made when they played it.
The new setup system had nearly all of the advantages of the most complex previous system but without mucking up the works with newly introduced granularity. Oath expressiveness itself in its denizens. If I wanted players to care about the game, I needed them to be able to speak to each other using the language they already had.
Next time, I’ll talk about what happened when I took that same realization to the chronicle phase, the traits, and the foundation system. Alongside that, we'll be releasing the next major playtesting kit.
Find all of Cole's Design and Development Diaries for Oath: New Foundations, and more, on Board Game Geek.