In the previous development diary, I talked a bit about how my desire for system expressiveness led me astray during the game’s design and development. Figuring this out took months of work. As I mentioned in that piece, I spent a lot of time reading playtest reports, teaching the game, and trying to figure out why on earth the many little systems that seemed so robust and interactive in isolation were proving to be less than the sum of their parts.
I think part of my trouble is that I didn’t really have the vocabulary I needed to describe the problem. If you think “expressivity” is an excellent design goal, then it’s hard to make the argument to cut it. That’s really the core of what made the first month or so of cuts so vexing. I was cutting out stuff I liked without any deeper understanding of where those cuts might get me in the end.
This changed only a few weeks ago. In March, I had the opportunity to speak at the Game Developer Conference. GDC is a funny event. The vast majority of the conference is directed to professionals working on video games, and the event mirrors all of its best and worst parts of that industry. You’ll find yourself routinely bumping anxiety ridden graduates hunting for their first job, the worst type of tech-finance bro, and your design heroes. There you all are, in beautiful San Francisco, standing in line to get an overpriced banana at a convention center.
I went to GDC this year to give a talk along with Josh Yearsley about Arcs. We were talking about the experience of working on big projects in an “early access” style, with long development timelines and a lot of user-feedback. Usually I try not to give a big stressful talk while working on an active project, but it so happened that this year I was coming to GDC with my mind full of Oath. I spent the evenings and early mornings fretting and cutting, and I tried to find my way through what was proving to be one of the most complex development problems of my career so far. Even though I was a month into pretty major cuts, I didn’t feel any sense of progress or have an idea of where the design was headed.
Then, one morning, late in the show, I attended a talk by Jon Perry where he talked about incorporating some of the principles of good tabletop design in video game design. Jon is best known to folks in tabletop for games like Spots and Air, Land, and Sea. In the world of video games, he’s probably best known for his recent work on UFO50. In his talk, he described applying some of his tabletop design practices in the design of the turn-based games in UFO50. As I listened to his talk, I realized that my work on Oath’s expansion had actually provided a counter-example. I had been abandoning some of the best aspects of tabletop design in the interest of seeing how expressive I could make the system.
The next day I headed back to Saint Paul. As I typed away on my laptop during my flight home, I found that his talk had given me the terms I needed to formalize my approach to Oath’s development. Over the past month, I had recognized something was wrong in the design. Intuitively, I had started pushing the design towards a simpler design, but I hadn’t quite articulated why I was doing it beyond the usual development buzzwords of “cleanup” and “streamlining.” I had also pushed against this design in an effort to protect the design expressive range. Jon’s talk had helped me realized that I wasn’t protecting the game’s expressive range, I was protecting its granularity and that granularity was not something that should be protected for its own sake.
One of the most granular elements of the game was the curse and trait system. In its most complex articulation, the system could support all sorts of character positions. Initially the traits were just one-off powers. Then, earlier this year, we moved to a unified trait system where a player’s traits were also yoked to the game’s foundations. For instance, if you could achieve 3 nomad influence, you would become Well-Connected. Then, additionally, if you achieved some other goal, such as getting the Darkest Secret by the end of the game, you could alter a linked rule in the game.
This meant that you could maintain traits without adjusting the matching foundations. Perhaps you were a brutal warlord who wanted to change the rules of the game. Well, if you can’t change those rules, you can still be a frustrated brutal warlord. This level of detail seemed really neat, but in practice players often just asked why they should bother changing the foundation if they already had the special power. I tried to argue that many of the powers were designed to harmonize better with the changed rule than the default rule, but the players shrugged at this. A power is a power. They didn’t have the strategic bandwidth to evaluate how a rules change might improve their chances in the next game, especially since the changes were often so marginal. Often, they viewed the big foundation alterations in the game as semi-unintended consequences to their actions. And, I think they were right to do this! Though there are some exceptions (like abolishing the Chancellor), it was okay for them to stay focused on the near-term. The system needed to explain why the world was the way it was—it didn’t need to make the players into masterminds. Causation was more important than agency.
With this in mind, I combined the maintenance conditions and foundation shifting conditions into one condition: a manifest condition. If you met the condition at the end of the game, the trait would be active. If you didn’t meet it, it would stay inactive. Inactive traits wouldn’t be lost, but you could only have so many at the start of each game. This preserved the feeling of lineage collapse without any extra rules and let me keep the possibility of a frustrated warlord. It’s amazing how often in Oath players feel their position expressed in a facedown card. If that was true for a denizen (I’m looking at you Bandit Chief!), it would certainly be true for a trait as well.
In retrospect, it was clear that this had been a place where the design had moved towards chunky design and away from granularity. Now, armed with those terms, I looked through other parts of the design. I’ve written a lot about how New Foundation's essential design strategy is to increase the game’s ability to remember. But, just because it could remember more didn’t mean the players were bothering to use that information to tell more fully realized stories.
The shadows and curses were a good example of this. Initially the shadows were a place where players could corrupt the sites with special effects that would be linked to a site in a semi-permanent way. However it always felt a little unhooked from the rest of the game’s design, so players had few opportunities to organically interact with them. I build all sorts of rewards (and penalties) into different articulations of the shadows, but players mostly just avoided them. Then, later on I generalized the site-memory rules to include locked cards and the new edifice system. This gave those locked site-only denizens many of the same properties as shadows, but with the benefit of also being denizens .Shadows started to feel obsolete. When I cut them from the design, their absence was hardly noticed.
This also happened when I moved to the curse system. One of the last reasons Shadows were in the game was to offer curses. Curses were a central element of the New Foundations project and I hoped they would offer players a way to “borrow” from future game positions. It took a long time to get the right balance of benefit to penalty, but, even when they were well calibrated, they lacked the kind of thematic and narrative heft they needed. After cutting shadows, I decided to drop curses as well. Once again, their absence was hardly felt. In retrospect this was obvious. The game already had curses, after a fashion. No player-debt system I could conceive would be as compelling as cards like A Small Favor or Vow of Silence.
I now turned my attention to what is probably the most central aspect of New Foundations: the revised Chronicle Phase. This element of the expansion had been quite stable for months. Generally, players seemed to enjoy it and it was certainly interesting. But, as my development goals clarified, the Chronicle Tasks began to feel like a loose fitting shirt. They were clever, but they were no longer fully in sync with the rest of the design.
For a long time, the player with the People’s Favor handled the introduction of new cards from the archive and the dispossession of other cards through a “Judgment” system. Based on the strength of the People’s Favor at the end of the game, that player would perform a number of trials. In each, two cards would get pulled from the game’s discards. One would be revealed. The player could then either protect that card—making it a potential starting adviser for next game—or dispossess it. Their decision would influence their choice of starting adviser next game and impact the cards that were introduced to the game. If you protected, the unrevealed card would become dispossessed and a new card, matching the one you protected, would be added to the archive. If they chose to dispossess the face up card, both would go to the dispossessed pile and two cards of the “opposite” suit would be added. The opposite suits were determined thematically (Order and Discord, etc).
This was a really neat system! It made the amount of turnover in the deck variable, from 1 to potentially 10 cards being introduced each game. It also gave players a lot of control over what could be added and showcased it all in a very visible phase where players could react in real-time to decisions about the composition of the deck. Players liked it. I was happy with it. And, for the past 4 months or so, it remained stable.
However, as the design matured in other areas, this system started to stick out. While experienced players enjoyed it, it was a little hard to teach. And, even once players got the hang of it, it still took awhile to resolve. There were bigger problems as well. The Chronicle Phase happens at the end of a game of Oath. That means players are already going to be tired and looking for a break. The phase needs to be propulsive. It could never match the excitement of an end-game situation, but it could try to at least look forward to the next game. At its best, the trial system had players swapping stories about the game they had just played. This is usually a good thing, but, if it starts happening while the game is still going on, it can slow the game’s forward momentum to a crawl.
With this in mind, I looked to apply some of the lessons I had learned elsewhere in the design. The only game-state requirement for this phase was that it needed to remove about 6 cards from the game. Okay, easy. The neatest part of the trial system was that the player who had the People’s Favor could defend cards they liked or banish cards they hated. I boiled all of this down to a simple system. They would draw 7 cards from the shared discard. They would choose one card to protect—this card would count as their starting adviser for the next game. The rest would be dispossessed. This also gave the player special knowledge of which cards were exiting the game and the minor thrill of knowing that they had been responsible for dooming those cards.
Critically, these choices were made after the player with the Darkest Secret had determined the next game's Oath. This meant that they could go into their choice to protect a card knowing in advance what the Oath would be next game. In essence, their chronicle task gave them a benefit for the next game. (I should mention too that there are foundations that adjust this, and give players additional controls over what gets discarded and what they might protect.)
With all of those responsibilities, I moved the task of introducing new cards to the game from the Archive to the player with the Darkest Secret. Now free from the constraints of building an ultra-granular design, I followed the same basic framework as base Oath, but with fewer restrictions. This player would pick any three different suits and introduce 3 cards of the first, 2 of the second, and 1 of the third. They could pick one card to keep and the rest would be added to the world deck. It was easy, and it had impact.
Finally, I revised the relic pack-up system to respond more cleanly to the game that was just played. I had put in rules that allowed site’s to remember relics shortly after the crowdfunding campaign. However, in practice it wasn’t working. Generally what happened was the only relics that got stuck to sites were those that players didn’t want and didn’t use. You can’t remember something that you never experienced, so it didn’t really matter that the sites could remember them. To address this, I reworked the relic restocking system to draw on cards held by players. If a dramatic game revolved around the bandit crown, that relic could now face two fates. If held by the winner it might end up in the reliquary. If held by anyone else it would get shuffled together and then reseeded at the sites in the game. In short, the relics would get linked to their corner of the map.
The map also now benefited from a more fixed geography. Once you’ve set up your chronicle for the first time, players would never need to shuffle their sites. During packup, the sites would be returned to the atlas box in a fixed order, allowing each copy of Oath to have an almost unique geography. Think of the sites as forming a line, a little like a side-scroller. The part of the empire would get remembered and then the governing player could choose to scroll their world forward or back at the start of the game. Looking for that lost Bandit Crown—you need to head east, to the ancient ruins of past empires!
Collectively, these changes have created a much deeper chronicle experience that I think lives up to and, in some respects, exceeds the promise of the project. We still have at least a few months of work ahead of us in the office. There are a lot of cards that need tuning and editorial work that will require the usual care to get through. But, I think this is what the game is going to look like.
Find all of Cole's Design and Development Diaries for Oath: New Foundations, and more, on Board Game Geek.