Designer Diary: Intrigue
Guest Column by: Jay Treat
My oldest lasting game idea is a trick-taking card game where each suit belongs to a player. I've been working on variations of this game for over a decade. If that doesn't sound like a very long time, you’re probably old like me. It was originally named Lords of the Realm, until I made a game for which that title was clearly a better fit. Then it was Shatterpact. In both cases, the theme was an epic Fantasy battle featuring heroes with personalities.
Shatterpact was a card game inspired by more traditional trick-taking games with three significant twists. Firstly, each player is the Lord or Lady of one of the races—all the cards of that suit benefit that player directly. The deck is dealt out randomly so you’ll have a hand full of enemy heroes as well as allied ones. The core of the gameplay revolves around playing the cards that help you when they help you the most and the cards that hurt you when they hurt you the least.
Secondly, your hand changes over the course of the game. You play two cards each round, then after scoring, you pick back up any one card. You can pick up heroes of your realm that other players have tossed out and shape your hand for future victory. You can also claim your enemy’s cards (so they can’t) or your ally’s cards. Which leads us to the third new idea: Each player wants to help the player to her left, but you have no stake in the player to your right. These one-sided alliances lead to numerous interesting decisions about when to help your ally, when to push for your own victory, and when to accept defeat—and how to mitigate the results.
What Could Have Been
In 2012, AEG (Alderac Entertainment Group) announced their Tempest IP: A game universe centered in the city-state of the same name, effectively Renaissance Venice. They announced it with three games lined up (Dominare, Courtier & Mercante) and an open call for further submissions. They've since released Love Letter and Guildhall. The former is pretty innovative in both form and function and the latter is hands-down my favorite game of the year. It would be an honor to be published in the same line.
As I read the design documents for Tempest, I was struck by a listing of characteristics that some of the games had already used and new games were encouraged to use if appropriate: Power, Wealth, Guile, and Influence. These lined up surprisingly well to the traits of characters from Shatterpact and I suddenly began mentally recasting the game as one of subtle intrigue and hidden power battles, rather than straight-up violence in yet another meaningless war. The mechanics and gameplay of Shatterpact turn out to make more sense as a Machiavellian game of manipulation.
It was never clear in Shatterpact why you were sending heroes of enemy realms to the battlefield, but manipulating enemy agents is exactly what the shadowy figures of Tempest do. How does one win an invisible conflict? By pulling strings that no one can see: manipulation, blackmail, and deceit. I submitted the idea and by a giant stroke of luck I had the opportunity to demonstrate the game in its pre-Tempest-ized state to the team at GenCon. I wouldn't say it bombed, but the game we played did illuminate some of the game’s weaknesses:
The nature of the game makes it reasonable to imagine that many games will end in ties. While I know from hundreds of games of experience that’s not true, it doesn't change the appearance and games are sold on appearance. It also happened that the last game ended with a player in a king-making position. He couldn't win himself, but his play determined which of two other players would. I hate situations like that as much as anyone, and while the game doesn't push toward that state, it doesn't push away from it either.
What I Tried
I was determined to improve the game. On the flight home, I thought about the way allies and points worked. It was interesting and simple enough, but it wasn't intuitive. Explaining it usually raised eyebrows, which is a warning sign that something’s a little off. I didn't want to give up that aspect, because I am obsessed with games where players have to help each other in order to claim victory, but how could I make them care without hamfisting it the way I had been? What if each player needed to advance multiple agendas…Agendas shared with other players?
If a player cares about one side of three different conflicts, then the game becomes a series of ever-changing alliances because each player shares one agenda in common with each other player. I might work with Bob this turn to support the monarchy against the senate, but Anne next turn to fight for science over religion. After all, if I help any one player too much, they might score more than I do. This model lacks the novel asymmetry of Shatterpact, but instead sports a much more organic alliance mechanism, and one that has proven more dynamic as well.
I did a bunch of math to figure out how many cards the deck would have, how many factions would be present, and how many agendas they’d iterate between. There are 8 permutations on three dichotomies, so I found 8 factions in the Tempest IP that could headline the game (not to mention the three pairs of agendas that could define them). While a 60 card deck is always nice (because you can deal it evenly between 1-6 players), 15-card hands seem a bit much to ask for a 4-player game, so I went with a 48 card deck to support 3-4 players (and our 8 factions).
I found a piece of art for each faction (all by Levente Peterffy) that fit the setting and mocked up some cards. Originally, each card had one symbol for each of the three agendas that agent’s faction cares about and nothing else (because removing the rank simplifies the card). We would count up all the A symbols, compare them to the B symbols and determine which half of the A/B conflict won that round. It only took one test game to show that was way too much adding. We were playing cards simultaneously, but we had to reorient the cards to ease counting such that it was hard to track who played what or why. The game was miserable.
Don’t Forget Your Roots
I tried a couple more small variations of Intrigue that were nearly as poor before bringing Shatterpact out of storage for a refresher. Playing the original reminded me what was great about it that the new game was lacking. While it’s slower than the simultaneous turns I’d tried for Intrigue, playing cards in turn order led to much more interesting card play. The ranks on the cards helped stratify them, making each choice more meaningful. Finally, I had simplified the play-two-keep-one play style Shatterpact uses (into a series of overlapping rounds) and in so doing lost the ability for players to craft their hands over the course of the game. I played Shatterpact as the Lord of the Dwarves and started without a single Dwarf yet managed to win the game by playing what I did have well and picking up Dwarves from other players along the way. It was awesome.
And so I set out to make Intrigue more like Shatterpact. Or, if you prefer, to re-theme Shatterpact, but using the new shared-agenda model in place of asymmetrical allies.
I gave each agent a rank and changed the scoring method from counting the symbols across all the cards to simply determining which faction has the highest total rank and then scoring the three agendas that faction supports. Much simpler. In Shatterpact, the winning player captures an enemy agent (representing her victory point) before everyone chooses a card to put back in their hand. Under this new model, it is a faction that wins a round—not a player—so that option was out. I tried a couple games and the revision was playing much better, but the endgame was populated by only the highest ranked cards.
With no predation in the game ecosystem, the game would inevitably end that way every time. I was slowly working toward a solution when one of my playtesters, Josh IIRC, suggested an awesome one: Each player must kill one agent and keep one during the resolution phase, but the order she does it is up to her. It takes a bit more thought, but this keeps the game dynamic and offers an abundance of interesting choices. Do you kill an enemy 8 or save a friendly 7? Do you count on the next player to do what’s best for the pair of you? Can you force an opponent to save an agent friendly to you? The starting player for the current round chooses first and that can factor into which cards you’re willing to risk as well as mitigating the last-play advantage the ending player has.
Spicing it Up
With the core gameplay looking much better, I took a look at the advanced game. In Shatterpact, the advanced layer comes in the form of quests. At the beginning of each battle, you reveal the top card of the quest deck and it details the conditions for winning that card as another point, separate from winning the round. Most quests key off of character traits like Brave, Cunning or Swift. The characters with lower ranks have more of these traits so that cards that were strictly worse in the base game become better for this alternate strategy.
For Intrigue, I knew we would need an additional layer to really hammer home the flavor of secret plots and schemes, of quiet power struggles rippling throughout the city, of back-alley deals, and deadly double-crosses. A set of scenarios you could choose for each game and a board/map for each scenario would offer a variety of play and allow me to tell stories.
In one scenario, the Queen has just been accused of treason and is trying to escape. Those sworn to protect her are trying to secret her away while those set on imprisoning the Queen must catch her. A faction that supports the monarchy places cubes on a map trying to reach the escape boat whenever they win, while the factions supporting the senate place cubes that block and eventually trap her.
Another scenario was more generic, but requires players to play their cards to specific spaces on the board, each of which have a unique ability. One space doubled its faction’s score if they win, while another increases your agent’s rank and yet another lets you swap cards that have been played. That was neat but foreshadowed a long road of balance testing.
The Penultimate Version
The last major flaw in the design was that having 6 cards per suit across 8 suits meant that it was often hard to keep the cards most relevant in the endgame. At the same time, I was trying to figure out how to remove the agenda icons from the agent cards so that there would be room to reintroduce traits. Even if the game didn't need traits, having played with generic agents versus fleshed-out heroes showed that a big part of the fun of the game is learning about the characters and creating memorable stories about them through play. Getting individual card art as well as helping define their personalities through game-relevant traits was paramount to making that happen. (Sorry, I won’t share the art the prototype is currently using, out of respect for AEG’s IP guidelines.)
As is often the case in design, the best solutions answer multiple questions. Eliminating agendas gave the agents room for traits. Without agendas, the players have to key off of factions instead and the math that lets each player care about one faction that each other player cares about in a four-player game results in six factions. A few playtests convinced me this is a great setup to force the players to work together while making their choices simpler and more relevant. The traits inspired a set of scheme cards, not entirely unlike the old quest cards.
Each player chooses one of two secret schemes at the start of each trick and if the scheme’s condition is met, she can reveal it for an effect. Many trigger at the end of the trick and score you an extra point, but some trigger when agents of a particular faction or with a particular trait are deployed, letting you actively mess with the game-state. It feels very spy-ish without adding much time to the game and players seem delighted to scheme and gamble and keep secrets from each other. I may still introduce scenarios/maps as an add-on at some point, but I’m quite happy with the schemes for now. Ideally, schemes are one layer and maps could be another, letting you use either or both as you please.
I took the game to Metatopia where at least 12 players gave me more positive feedback than I've had before, to the extent that one of them told an attending publisher to get a demo of the game. I submitted the prototype to AEG at BGG Con, but they decided to pass on it.
The Game Today
I've played hundreds of games of Intrigue and I’m quite pleased with the depth of play, and the hilarious moments that can arise when players reveal their schemes. The core game has been stable for some time, but I iterated on the schemes quite a bit; I've printed up six different sets of ~50 scheme cards by now, removing the ones that don’t enhance the play experience and keeping the one that do. I want cards that are fun, flavorful, and interesting or that shore up weaknesses some players feel the core game has. For instance, there are a tons of schemes that score points, allowing players who have fallen behind in the very tight core game a chance to come back. A few cards offer as many as three points, but are much harder to pull off. That’s one place where having a flip side really shines—You can always choose a smaller, more reliably alternative if you’re not in the mood to gamble. Some cards offer rewards for long-term strategies and some shake up players’ hands to help you get out of a rut you may have slipped into.
There are a couple modes of play I still think have potential: Playing with hidden affiliations is really interesting; and I might still bring back maps/scenarios, but they’re clearly not necessary for the initial game. I think I’ll save them as bonus content once the game is out there.
I've sent Intrigue out for two blind playtests so far, and I’m really glad I did because they exposed a very serious problem with the rules presentation. What I've learned is that while casual players enjoy the core game, hardcore gamers dismiss it as trivial because the simple rules don’t seem interesting enough. They don’t even try the advanced game because they've already decided to hate the core game. I've never had this problem when teaching the game in person, and the difference is that I explain the game differently than the rules, highlighting the subtle strategies that make it play so differently from traditional trick-taking card games. I've done a few things to combat this:
Firstly, I added a strategy section to the back of the rulebook, plainly explaining the top layer of strategies that players tend to miss initially when left to their own devices. Next, I recommend that hardcore gamers add the scheme cards after just a few tricks. I find that the schemes are quirky and flavorful enough that they keep players engaged long enough to start noticing the intricacies of the core game. Finally, because I don’t expect every group to read the strategy section in the back of the rules, I’ve made an Explainer; a sheet separate from the rules that is meant to introduce the gameplay enough to get players started quickly as well as informally hinting at the sideways play that will help you to succeed in this game of manipulation. I just sent out a new copy with these tweaks, so I’m eager to see if that’ll do the trick.
I’ve never been so confident that I will find a publisher for this game. After over a decade, dozens of versions and hundreds of games, that’ll be nice.