Game Design Pattern: Stakes aka Triangle Strategy Scales of Conviction

Max here, not so subtly promoting that Maximum Recursion Depth, or Sometimes the Only Way to Win is to Stop Playing: The Karmapunk RPG is on sale for the drivethrurpg Christmas in July sale (also available on or in print on Exalted Funeral). I occasionally post "Game Design Patterns" on my own blog, but I think going forward they might be a better fit for Iconoclastic Flow. This Pattern references the videogame Triangle Strategy, but I primarily use a Play Report for MRD2, the quasi-sequel-slash-standalone-but-also-cross-compatible-game in the "MRD series" that I'm currently developing, as the primary point of reference (hyperlinked further below).

The videogame Triangle Strategy is a traditional-style Japanese Tactical RPG that mostly plays it pretty safe, albeit well made, but it does implement a few innovations, among them the Scales of Conviction.

SoC involves pivotal decisions in the game affecting the missions and narrative, but the Player does not get to choose directly, only influence the decision. Instead, they hold a vote, and the party members have preconceived stances which can be Swayed by the Player. By investigating the area, gathering information from NPCs, and other small actions like that, new conversation options open up which can be used to better convince the party members how to vote, and sometimes this information is crucial to winning the vote.

It's a really cool idea, but it makes even more sense in a TTRPG, where the GM has a lot more flexibility, and where the group have to Sway not just the various NPCs in their association, but also come to an agreement among themselves.

I implemented this in the most recent Session of my MRD2 Campaign for which I wrote a Play Report, and it was a lot of fun. My players barely rolled any dice or even had to use many of their character features, it almost entirely came down to roleplaying, yet it was still goal-oriented, making it feel like an Encounter all its own, and perfectly in keeping with my Social Intrigue Game Design Pattern.


Rather than calling them Convictions, we'll call them Stakes ;). These are the voting issues (I'd call them issues but I'm already using Issues as a different concept in MRD2, maybe a future Game Design Pattern post...), which we'll assume to have only two Positions each. Ideally these should be logical extensions of what the PCs have been doing for some time, where either Position is valid and moves the game forward, but where the PCs would be likely to have an opinion and feel strongly enough to want to win their Position i.e. actually have a stake in it. There should be clear and tangible consequences to one Position over another.

You could in theory have Stakes with more than two Positions, but my inclination is that that would get much more complicated, and I haven't given it much thought yet.

In that PR Session above I had three Stakes, which admittedly is a lot to manage but made sense in that context:

Pro- or Anti-Corporate Alliance
Fold or Crush the Upstartups
Consolidate Power in the Space Settlements or Appease them

In MRD2 the PCs are Nazarites, agents Contracted by Corporations, granting them special powers and Golem (mecha). It's meant to be "commentary", but if you're interested at all in the premise then go ahead and trace through the hyperlinks from the PR above and elsewhere, but for now that's all you need to know.

My PCs chose to vote in favor of the Alliance, paraphrasing their words, "this is typical cartel stuff". By forming an Alliance and working in unison on the other Stakes, they are better able to influence the Upstartups and the growing stellar-political powers on the Space Settlements. Without the Alliance, the subsequent Stakes could still be voted on, but there is less incentive for the Corporations to agree to the terms. In the case of whether to Fold the Upstartups (basically buy them out or develop some other mutually beneficial deal with them) vs. Crush them (tie them up in frivolous legal battles, poach their talent, try to out-market them into oblivion, etc.), the PCs had already made the decision to buy out one of the Upstartups and one of the PCs had Karmic Attachments with the Founders of that Startup, so they had a preconceived incentive to want to Fold the Upstartups. The Space Settlements Issue was separate from the Upstartups Issue (these "Issues" being basically their quests) and they had not pursued it yet, so they had much less information, and no preconceived notions, so their Position came from interactions with the other NPCs, and this gave them some insight into the situation and consequences of them not yet engaging with that Issue. So here, for me, having three Stakes (or at least the latter two in particular) actually worked out quite well.


These are the entities who get votes. It could be interesting to treat the PCs themselves each as individual Stakeholders who might inadvertently be pit against each other, maybe if you like that style of PC Conflict games like PbtA, but I prefer to have the PCs first come to agreement among themselves, and act as one Stakeholder.

So in this case, the PCs act as one Stakeholder representing their Corporation, and the other Stakeholders are each representatives from the other Corporations. If you don't have too many other factions, you could similarly flesh out each Stakeholder faction with multiple NPCs, but I thought it would be more fun and more manageable to just lump each Stakeholder faction with a single Stakeholder, the Manager of that Corporation's rival Nazarite Team (those Rival Teams might get fleshed out in the future).

I used my Character Formula from the previously hyperlinked Social Intrigue post to create the Managers for the other Corporations, and used a similar Pattern for creating the Corporations, and could also use this Pattern to create the other Cyblessed Liaisons or Rival Nazarites for the Rival Corporations. I tried to make their positions on the Stakes make intuitive sense given their personalities, or if the PCs had strong prior interactions with these Stakeholders, then as a logical extension of those previous interactions. For MRD Karmic Attachments were useful, or you could use FATE Aspects, or anything like that, but it's not necessary.

(Over-)Balancing the Scales

You need to have an odd number of Stakeholders or else you can run into a tie. However, if the PCs are themselves the tiebreaker, then there's no real stake. Likewise, if you weight the NPC Stakeholders in favor of one Position and the PCs agree with this Position, it's even more unsatisfying. A few options for "(Over-)Balancing the Scales" are:
  • Subtly Seed the Stakes ahead of time and get a sense of what your PCs' positions might be, and Overbalance the Scales against that Position.
  • Occlude the NPC Stakeholders' Positions so the PCs aren't certain about the vote and feel compelled to investigate further.
  • Have one or more Antagonistic Stakeholders. These are Stakeholders who hold opposing Positions to the PCs, and feel strongly enough about them to try to Sway other more ambivalent Stakeholders if the PCs don't also attempt to do so.
  • Have multiple Stakes. This requires more work and can get complicated, but it decreases the likelihood that for all Stakes the PCs' Positions will be the majority Position.
Some of these can be used together, and I'm sure there are other ways. It's not the end of the world if it does end up being too straightforward, just like anything else with TTRPGs, be ready to pivot if they take things off your "rails".

In the PR Session, some of these Stakes were Seeded via how they engaged with the Issues or Karmic Attachments; in other words, the events of previous sessions.

I did not use Occlusion, instead stating the other Stakeholders' Positions outright, but in retrospect I think Occlusion is a better approach, where in some cases some Stakeholders might be forthcoming about their Positions, while others might play their cards closer to the vest.

I actually improvised the idea of Antagonistic Stakeholders during the session, so it was a threat hanging over them if I needed it, but since I hadn't prepped it, and I Seeded the Stakes well enough apparently, it didn't come into play. If I use this Stakes Design Pattern again in a future session, which I definitely intend to do, I will also probably develop this idea further.

I don't regret in this case having multiple Stakes, but in future cases I might just stick with one and see if I can't develop that one Stake out further instead, possibly by having more than two Positions.

Make them Sway

So you've got a sandbox environment with a lot of cool stuff in it, and you want the PCs to explore it? Well, now they have a reason to do so. Track down the Stakeholders, and Make them Sway to your Position on the Stakes!

In the PR Session, I had the various Stakeholders shuffle off wherever they please after the initial pre-meeting, and the PCs could investigate the different Locations in Venus City for a bit before the Meeting where the Stakes would be decided. I chose not to tell them where the Stakeholders went off, instead encouraging them to either guess the Locations of the various Stakeholders, or just go to whichever Locations sounded interesting and see what happens. I was satisfied with that; it still wasn't arbitrary because they had details about each Location, and with a stricter time limit the guessing of where each Stakeholder would go might have been an interesting minigame in itself, but if you want to tell the PCs outright the Locations of the Stakeholders I think that's fine too.

There were four other Stakeholders, with one each at one of the three Locations in Venus City, and one of them as a random Encounter (from a shortlist) between each Location. By doing it deterministically, I could craft the Encounters a bit more, but you could also randomize them.

Another note is that Venus City is unique in that physical violence is physically impossible (it's a metaphysical extradimensional space), and I think that worked in my favor for how I like to do things and in forcing the PCs to be creative in how they would Make the other Stakeholders Sway, but you could also interweave violent Encounters or physical coercion as well if it would make sense and be more fun to do so.

Using The Asherah Gardens from that Session as the example, one of the Stakeholders was fascinated by the five-dimensional psychedelic fruits which supposedly could be used to derive prophecies, but he was himself incapable of interpreting them, and so it was straightforward enough for the PCs to offer to eat the fruit and make prophecies to Make him Sway to their Positions on the Stakes, and these prophecies will have other longer term consequences for the Campaign. Most of the others were a little more involved, but this gives you a sense of how you can integrate the whacky and interesting sandbox Encounters with these Social Intrigue Encounters for the sake of the Stakes.

Plugging MRD one more time ;)

Wrap up on the Stakes Game Design Pattern aka Triangle Strategy Scales of Conviction

This Design Pattern is system agnostic and mostly non-mechanical. It's based on the intuitive nature of democracy and personal investment, and it gamifies Social Intrigue to create what to me is a seemingly pretty novel kind of Encounter for a TTRPG. I think there's a lot of potential here for this Pattern, especially for "higher level play" if you're coming at it more so from a D&D/Picaresque direction, but you could even use it for a Fellowship of the Rings or a Thieves Guild or something of a "Lower World".

Also I still haven't beaten Triangle Strategy so no spoilers!

Also also, buy Maximum Recursion Depth!