Ambivalent Aggression Design Diary

Ambivalent Aggression Design Diary (This post is very long,  so you might prefer to read the document version I just included with the game files)

Ambivalent Aggression is a milestone for me. Most importantly, it dispels any excuse I could conjure that I’m not a *real* game designer.   It was unpleasantly easy to look at the small handful of game jam games I’ve made and comprehend the gap between my passion and production.  {To be clear this chasm is still unfathomably large. I want to devote myself to making games the same way other people can devote themselves to loving someone}

This is the 2nd game I’ve ever made that I’m proud of.  The game that takes the honor of being the 1st one happens to be a PowerPoint presentation which I made over the course of a weekend {I’m referring to https://zacharymorris.itch.io/magic-281-intro-to-alchemy ... a game where you attempt to understand an alchemy lecture that advances forward every 10 seconds}   Well I’m not here to talk about seconds, I’m here to talk about firsts.

Ambivalent Aggression is the first long-term project I’ve finished…I started designing the game in 2021 and released it on May 29^th ,2023.  That date sticks in my head, because my friend Maddison Baek released their excellent action video game magenta horizon on the same day.  {Check it out, it’s one of my favorite action games- https://store.steampowered.com/app/2109060}.   I remember it feeling so strange to be excited about being done with my own game instead of just Magenta Horizon’s release.  This was just a coincidence from when I happened to finish writing the rules, but it sure gave me a conflict of interests for what to shill with my discord status.  Well, my games are just grains of sand in the desert so it’s not like this release date collision was meaningful in any way except to me.

Ambivalent Aggression is also the first physical game I’ve made.  Personally, I see video/board games as two halves of one whole.  Despite this philosophy, until now I’ve only designed video games.  This is due to a playtesting bottleneck.  From my experience, the main difference between designing a board and video game is the difference between a discrete and continuous function. 

My process of designing a boardgame is a collection of design sessions which are separated by playtests.  Videogame design feels more like a period of time where I always have more work to do, and I sprinkle in playtests as I desire.   Playtesting is an absolutely vital part of both of these forms of design, but in the former it’s a blocking task. 

I consider playtesting the most cursed part of game design due to the logistics challenge.  First, I need to have people in my life that are willing to try a prototype. Next those people have schedules (that’s especially annoying since so much of my creativity is based off whims and moods). I’m also not a particularly social person so recruiting strangers is not something I’d consider easy.    Additionally, playtesting and playing other games is zero-sum: to playtest is to pay the expensive opportunity cost of not playing someone else’s awesome game.   

.

.

.

For most of my life, I’ve only had a handful of friends.  My love of boardgames has grown under the rations of games with family.  Luckily my dad and brother love to play games.   I started working on Ambivalent Aggression during this era of my life, for the first few months of its existence my dad & I were the only people to have played this game. 

A lot has changed in 2 years.  I could blow 18-year-old-me’s mind by saying that 40 people play tested Ambivalent Aggression.  That’s the most important first from this game, having a whole credits paragraph of playtesters.   So, to properly chronicle my history of design, I’ll be spending a lot of time talking about the people I played with.    

So, like I said, this game was born during an era where most of my multiplayer gaming was with my dad.  He prefers mastering specific games and I prefer trying a variety of games.   That’s not a great combination, but it actually works out well enough. There’s plenty of games which are so wonderful that I’m always eager to play more and he’s willing to try new games. My favorite games to play with him are Terraforming Mars, Dominion, Azul, Peggle and Cribbage.

Most relevant to ambivalent aggression is Cribbage, my favorite playing card game. When my dad taught me that game, I got to have the fantastic experience of being absolutely crushed… shown first-hand that there were entire levels of skill to Cribbage I didn’t understand.  I adore realizing that there’s much more to a game than I was giving it credit for. The process of learning how to be competitive against my dad’s Cribbage play was incredibly satisfying. I’ve always had an interest in playing card games, this journey of learning cribbage gave me a deeper love for them.  

I lied about the last paragraph being relevant   in fact, I can’t think of a single meaningful thing Ambivalent Aggression shares with Cribbage.  Some other playing card games that I love are Poker, Caravan, and quite a few flavors of solitaire.    AA lacks similarities with those games as well, to me this is a sign that I succeeded in one of my design intents:  to make a playing card game that is nothing like every other one I’ve played.

I mention these games because they still had an influence on my design.  It’s just a subtle hard to measure influence.  I play so many games that it’s pretty hard for me to list influences or inspirations generally, I just acknowledge that the vast collection of games I play constantly influences my designs in ways I don’t understand. Even as someone who prioritizes creating unique games, other games’ ideas are a constant fuel for my process.  In Race for the Galaxy terms, I need exploration and consumption in order to get the cards that I’ll discard to pay for placing my own. Even if the final product bears little similarities, you’ll notice a running theme of scrapped iterations of AA harvesting from other games. 

.

.

.

Anyway, the beginnings of Ambivalent Aggression. There are 2 ways a game can be born:  either from an idea or an intent.  This one came from an intent:  I wanted to design a playing card game.  I thought it would be an interesting challenge and set of constraints to work around.  Every single iteration of this game exclusively used a standard deck of playing cards as its components {although some versions required players to take notes to keep track of scores or deals}

Only using a deck of cards is a pretty interesting design constraint.  I could never solve a problem by just conjuring up more components and I often came up with clever ways to reuse other game elements to represent mechanics.      This style of design is actually an expression of how much I despise making components. It was actually a blessing to never have to do significant preparation for any of my prototypes {most I ever had to do for this game was write and print out reference sheets}.   

The other motivation for making a playing card game was making a game that I would want to play.  I used to be in the habit of always carrying a deck of cards with me, so if ever a random time needed to be filled with games I’d have a pocket full of possibilities.  Unfortunately, most playing card games don’t actually do a great job of fulfilling this niche for me. Poker only works as a gambling game, Cribbage has an arcane and hard to teach scoring system, I’ve gotten quite the fill already of playing Caravan, despite enjoying solitaire I don’t actually use it to pass time.  Then there’s the games most people would suggest playing with a deck of cards, and I find many of them quite boring. 

Ironically, near the end of this game’s development I stopped carrying a deck of cards everywhere I went. Despite that, I certainly succeeded in making something I’d find interesting to play with them. 

I also specifically wanted to make a game where you manipulate a grid of cards, that felt like underexplored design direction.  I’ve since played some games that use space and positioning in interesting ways {especially solitaires which are often known for their card structures}, but at the time it felt like I was going to invent a card game where space mattered. 

The first mechanic I came up with is one that’s still at the heart of the final game, which is quite impressive given that I did 8 complete redesigns. This idea was the concept of replacing cards on the board with lower valued cards.  Of course, this alone isn’t very interesting.  In order to make this truly work, you need players to have a reason to be invested in cards on the field having high values.   Many of the early versions involved summing up the values of certain cards on the table.   This makes for a pretty spicy dynamic:   high impact cards are more vulnerable to being replaced so you had to carefully consider the proper time to play them.  Similarly, the cards that are pretty stable didn’t contribute much, but could still add up.

The original version of this game is unfortunately not preserved in any of my files.  Every time I did a redesign I made a new rules document, but I only started this habit after the 2^nd version.  While designing this game, I thought it was part of my style as a game designer to have the game live on its rules document. Many designers don’t write the rules until the end, meanwhile I never play tested a version of this game that I hadn’t already written down rules for.   A year ago, I would have said that the process of writing is fundamental to how I approach game design… now I’m not so sure {This month I ran a playtest for a game where I hadn’t written down any of the rules which is something I’ve never done before.}

For this game, almost all of my design sessions were short bursts of activity.  I would constantly go weeks without thinking about the game at all… then spend hours working on it. I find writing one of the most powerful ways to channel my chaotic explosions of attention. 

Writing is useful because of the way it forces binding of ideas.  You have to make many decisions about what directions you want to commit to. As that starts to take form, you’ll find many details to resolve.  If you’re struggling with a game being too fuzzy in your head, often writing down rules or even just your ideas helps to give it more shape. 

To be clear, you shouldn’t feel any obligation towards any writing you do in the process of designing a game.  I started a new document for each redesign mostly for preservation reasons, but it’s also a helpful way to discourage blindly transferring ideas from the previous version.    Each set of rules I wrote only for myself, it was vital that I could simply ditch them once the game moved on. 

My favorite part of designing this game was the freedom I had to change course.  Now that I think about it, this is the real difference between video and board game design.  Videogames have so much other work that it’s incredibly easy to get invested into the design decisions you’ve already made.  Scrapping whole systems or completely redesigning everything is painful if there’s been programming work already done. 

Most board game desiginers will make a game and then join with a publisher to accomplish all the other work that a game requires. Most video game designers work concurrently with all the other work on the project, giving their decisions much higher costs.   

As someone who’s spent most of their life learning to make video games, wow boardgame design is liberating. That was especially true here since this game never required me to invest into making any components.  None of my decisions had a work cost.  it’s a joy to be able to try ideas I know are stupid. It’s a joy to wake up one day and change the entire identity of the game. It’s a joy to try a playtesters suggestion in the moment.

Major Version #1

 This was a 2 player game where one player represented red and one black.   It centered around a decorative card which had 8 spots for cards adjacent to it.    Players spent their turns placing new cards or replacing existing cards. After the endgame condition which I do not remember, players added up the values of all cards in their color as their score. The clever part to me was that throughout the game you’d be playing both black and red cards from your hand, the same gimmick at the heart of twilight struggle. 

I have an incredibly strong memory of being unable to sleep on the night preceding my high school graduation… not in anticipation of the next day but in ecstasy thinking about this game.   I spent the night listening to 2010s pop music and writing rules. 

Major Version #2- Card Succession

This version was an expansion of the previous one, and I started to conceptualize the game as a succession crisis where factions are trying to gain enough support to get a monarch on the throne. 

Instead of just a static 8 slots for placing cards, the central play area became an expandable zone of placement. Each Queen or king played would expand the possible places you could place cards, as they supported all 8 spots adjacent to them.

Each round ended once any royalty was surrounded by its color of card {all 4 of corners are its color or all 4 sides are its color}.  Then each player would score by adding up the values of all cards on the field of their color.  If your color of royalty was surrounded, you got 10 bonus points.

I also added a mechanic where you use a card from your hand at the beginning of the game to bid for which color you want.  This is probably where my fascination with agenda scoring cards (another staple of all these iterations)  originated from. 

This version was also inspired by cribbage, the scoring was a race to be the first person to score 100 points. If you ended a round, you’d get the privilege of scoring first (whether or not you surrounded your own royal)

Also jacks were a wild number that could go on top of any non-face card and have any number go on top of them.

One of the strongest quirks of this version was the way kings and queens were permanently placed with nothing being able to replace them. Each round was shaped around them.  Two opposite color royals next to each other would lock both of them from one of the ways they can be surrounded. Two same color royals next to each other made it much easier to surround one of the.   It was pretty interesting that these cards were so impactful but didn’t really score points.

.

.

.

These are the only 2 versions (me and) my dad played.  Before the next version I moved away and started college.  Now starts the game’s strange relationship with player count.   I started by making a 2-player game, now it becomes a 3-4 player game with a brain-melting “play 2 hands at once” 2 player mode. 

At this point, I purchased Tabletop Simulator for myself (and also my best friend who was subsequently subjected to the above-mentioned cursed 2 player mode of the new iteration).  Honestly considering my interest in games it’s strange that I didn’t have tabletop simulator already at this point in my life {I went through all of 2020 without considering it necessary}     

I’m not particularly fond of playing physical games online, especially dealing with the jank physics of Tabletop Simulator specifically, but its is a great tool that has had its time to shine every so often.  Most obviously, it allows me to circumvent circumstances that would otherwise prevent me from playing a game. It’s also quite useful for setting up/manipulating things that would be a pain in real life, doing on the fly experimentation with components, and creating save-states/objects for future replay or analysis.  The specific program doesn’t have to be tabletop simulator, I need to get around to familiarizing myself with other software {Tabletop Simulator has some questionable leadership and annoyances to deal with, but it has the unfortunate advantage of popularity and widespread adoption}

I think I only did 3 tabletop simulator playtests for this game over the whole development; the vast majority of playtests were in-person.   I think I’m generally overconfident in my ability to separate being annoyed at tabletop simulator friction and flaws with a game.  Each of these sessions were helpful, but this is definitely something I slowly stopped seeking out.

.

.

.

At my college’s game club, I play tested my game with strangers for the first time in my life.  This was an incredible emotional high for me. It felt so strange,wonderful and validating to see people getting joy from an experience I was responsible for.  The writer in me is sad I didn’t write any journal entries of that day… the best I can do ~1.75 years later is just this paragraph.  I don’t remember what I learned about my game from that session (I know there was plenty of stuff), but that night was an incredible highlight for me… probably the first time I’d experienced validation as a game designer. 

This is pretty amusing in hindsight, because the version they played feels like the most cursed iteration of this game.

Major version #3- Succession now with strange hand size economy

Rereading the rules for 2.0 and then 3.0 has been interesting.  There’s some pretty fascinating changes and unique design space I experimented with, I wish I remembered what I was cooking at the time.  

So the first significant change I already mentioned:  I was now targeting 3-4 players. Corollary to this change was the idea that players score a specific suit instead of a color.    I changed the bidding mechanic significantly.  Now the person who played the highest rank card would go first and people could share the same suit.   This system became less like bidding and more everyone collectively deciding their fate at the beginning of the game.  If you played the same suit as someone else, then great news-> you’re both on the same team this round.   In order to balance the power of getting to work with another player, any player on a team would also be penalized for any suits nobody is scoring {which by definition must exist since there’s 4 suits and at least 2 players are sharing the same one}.    [your scoring card was publicly visible]

Instead of surrounding a royalty ending a round and prompting scoring,   I made an entirely different scoring trigger (more on this later) and simply made any kings or queens that were activated worth 15 points.  For some reason, I decided to make the terminology of royalty being “activated” or not, it was the same mechanic of being surrounded by their color on either 4 all 4 sides or all 4 corners.

This version had 3 phases to every turn: Play, give, take.

The play phase was the core gameplay from the previous 2 versions.   You could play a card to empty spaces adjacent to royalty cards or replace an existing card with a lower value.  

Next is the optionial give phase:  Where at your discretion you could give cards from your hands to other players.    The best part of this was watching people’s reactions as I explain it, since it raises a rather obvious question -> So why would I want to give others my cards?    There’s a more interesting answer than cooperating with a teammate, look forward to it.

Finally, is the take phase, where you refilled your hand. Not sure if I’ve mentioned this yet, but the display of 4 cards where you take one whenever you play a card is in every iteration of the game.  This version was special though because there were conditions that allowed you to take multiple cards.  

•  Default: take 1 card
• For each royal you activated   this turn , take +1 card
• If you play a joker in phase 1, instead of placing a card on the table, you take all 4 cards during this phase

The scoring trigger is what made both this take and give phase matter.   A scoring was triggered whenever all players have 5 cards at the end of a turn. Players by default have a hand of 4 cards, so there had to be some royals activated or jokers used in order to trigger a scoring.  Strategically it was incredibly common to build up a surplus of cards, then distribute some to the table  to trigger a scoring.

After every scoring, players lost their agenda, meaning they had to use a card from their hand to make a new one for the next part of the game. This conveniently ensured that everybody was no longer at 5 cards, along with shaking up player’s goals and suits they cared about giving the in-progress  game an interesting strategic shift…. Just kidding -> almost everyone would always pile on the the suit(s) that were obviously scoring the most from cards already on the field.

I’m not sure if I had any rule to cover the edgecase of everyone choosing the same suit.  If it happened in a playtest I probably just told everyone to pick again. 

After the first scoring, your agenda choice pretty much decided if you were going to be competetive. If you didn’t have cards from the dominant suit in your hand after a scoring, then lol lmao skill issue.    The game also sure becomes a lot less interesting when most players are complacent with the dominant suit remaining that way.

.

.

.

Major Version #4 – untitled playing card game (I stopped thinking of the game in terms of being a royal succession crisis)

I diagnosed the previous version with 2 fundamental flaws: the way you were locked into scoring cards and the hand accumulation just being  out of place.   This design focused much more on timing, specifically avoiding being caught in a state where you can’t score much.  Of all the versions this one felt like the most work to play… so much counting. There was also an element of bluffing and reading opponents.

The basic turn flow of playing cards was the same. The difference is that taking a joker from the display of cards would immediately trigger a scoring.  {Taking the 2^nd joker ended the game as well}. 

So there was an element of tension because the jokers could show up at any time, while also making the decision of “Do I want to score right now” much more direct.   A solid change, especially along with cutting the hand size economy.  The idea of having 2 scoring events during the game would hang around for quite a while. 

It wouldn’t be one of my redesigns without something deranged.    Now each player’s scoring agenda cards were facedown, and you had to always switch your facedown card with one from your hand at the end of your turn.   {The evil part being that timing wise you switch your scoring card after you take a card and potentially trigger scoring beforehand. So, the agenda you score will have to be down for at least sometime}

In theory this meant that every turn ended with an agonizing analysis of what card you were comfortable scoring (and not having in your hand next turn). In practice most people just alternated between two cards.   The strain of considering multiple realities didn’t quite work out as intended in this version, but this design focus is both something that is essential to the final game and something I want to revisit in future games. 

This short-lived version definitely didn’t live up to the ideal of having to dynamically cycle what you score.  Mostly because cards on the table were slow enough in changing that your facedown card was always a reaction to what you could see in front of you. {Instead of a risky speculation or binding intent}. 

Here is an important “halfway” point in the game’s development.   Firstly, I concluded that your agenda card had to be an early binding point -> something stable which your play must pursue.  {Not a dynamic possibility space which you can use as a reaction to the game state.}   Additionally, this version was a crossroads, after it I decided that the game should have combat.  If I were to return to any iteration and make a different game, it would be this one.  I wonder how the game would have progressed had I chosen a different path.

.

.

.

I already mentioned that I play-tested this game precisely 3 times on Tabletop Simulator.  Interestingly, this isn’t the most significant impact that software had on Ambivalent Aggression.     At the beginning of 2022, I had the opportunity to try the classic Dune boardgame where 6 players each negotiate and scheme as one of the asymmetric factions.  This game blew my mind. There’s so many fascinating mechanics and the faction ecosystem is incredible.   I was thinking about it for weeks, and eager for the online group’s next session (even though the one game I had played had taken like 10 hours).     

{Unfortunately, I’ve actually soured on this Dune game quite a bit, but I’m still amazed by the design}

Dune living rent free in my head is important context for what comes next in the story of Ambivalent Aggression.

Major Version #5- Scheme

Like many redesigns of this game, It all started with an ambitious idea that brought with it hours of uncontrollable   crunch where the entire game reformed around that change.  The scary thing about interesting ideas is that I have no choice in pursuing them, if one arrives, I must accept that this fixation is my life now.

One of my defining traits as a designer is the way my head lives in possibility space.  It’s easy for me to look past what a game currently is and imagine a sweeping new version that is closer to my preference.   Prototypes to me are not steps along the way, but stars in the sky that guide my navigation towards a game’s potential.   I must be an annoying guy to have playtest your game, because it is so natural for me to suggest deranged changes that completely change a game.  {Consider it an implicit compliment if I ever play your game and don’t have any wild sweeping suggestions}. 

Anyway, as I’ve alluded to, Dune suddenly inspired me to add combat to Ambivalent Aggression.   I conceptualized this as an additional layer to the game, so the core of placing and replacing cards was not only left intact, but also was the mechanism that determined the battlefield. Now there were two types of turns: deploying and fighting. 

Deploying as you’d expect was a simple sequence of place, refill hand, then refill the display.   If you revealed a joker while refilling the display, then the game would shift into combat. 

In the last section I lied about the part where I said I learned my lesson about making the scoring static.  Here I added another card that you could switch at the end of your turn!  So, you still had the facedown scoring card and now you also had a faceup commander card {Worth noting that you had to choose only 1 of them to switch each turn}. 

Your current commander card was only relevant once combat started, during which you could exclusively move cards that matched it (shared a suit or number).  [Making the selection of cards you can control follow similar “matching” logic as the agendas do when scoring was a real clever idea that I’m proud off]

One of the fundamental ideas of combat from the beginning was the notion that combat strength and resistance to being replaced would be opposite concepts. {The higher rank a card, the more possible cards can replace it during deployment, but also the more possible cards it can kill and survive attacks from during combat}    

Card movement works pretty much the same in the final game as it did when it was first born.  Move in any of the cardinal directions,  if two opposite colored cards make contact then the lower rank dies.  Where did the restriction on not being able to attack cards of the same color come from? Great question!  I had a bunch of points throughout development where I considered cutting that, I’m honestly impressed something so arbitrary ended up in the final game.  [I’m of the opinion that it restricts the battlefield just enough to make it interesting, although that was definitely something I discovered along the way to justify its existence. I have absolutely no idea why it actually exists]

The only difference in the movement was the cards would *slide* (moving until they hit another card) instead of taking 1 step at a time. It was pretty silly, the story of why they slid was even sillier.

One of my interests is Ice Puzzles in videogames.  I developed this interest as a direct reaction to realizing that Ice Puzzles are awfully common, yet nobody talks about them.  Of course, this turned into a hyper fixation and among many of my online friends I’m known as the Ice Puzzle guy. {This is where the Ice in my username IceNinja comes from!}

The original draft of this AA version actually had 1 step at a time movement.  I was talking to some online friends about having done such an extensive redesign to this game, and they jokingly asked if I added Ice Puzzles.  After reading that message suddenly my brain said “Why not!” and I changed the movement to be sliding.  I laughed at that and convinced myself it was interesting, but this movement only lasted for 1 playtest.   In my head it elevated the tactical strategy of the game, in practice it was just so stupid.  {especially because you could slide a card off the table to remove it from the game}. 

Another spicy part of this version was that combat had dynamic turn order (while deploying was simply done in clockwise order}.    At the end of every combat turn, you’d choose the next player to take their turn.  Each person could only be chosen once per round, the benefit of being chosen last is that the new round starts during your turn…allowing you to pick yourself. 

One reason I introduced the concept of command cards was so that players could have a card to “tap” in order to keep track of who had gone during a round.  Pro-tip: Don’t assume that the magic-the-gathering term tap is intuitive, I had one playtester who interpreted that rule as physically tapping the card instead of turning it sideways. [Also tapping is a trademarked term so it should be avoided anyway]

The main reason I included command cards is to make the dynamic turn order an informed decision.  If you know which cards someone is capable of moving on their turn, then when they take their turn becomes much more meaningful.  

Naturally following this was the idea that players could request binding promises as part of their turn order choice.  Stuff like offering someone “If you kill a specific card I want (or avoid one I want to keep safe), I will choose you to go next”. 

This opened the door to making this game even more like Dune!  I added free negotiations, where players could trade promises, information, points, and cards from their hand.    Don’t have points yet since there’s only 2 scoring phases in the game? No problem, I made a convoluted system of debt so that all points you traded would be noted on a central accounting paper and you only had to pay debts once you had all your points at the end of the game.  {I cannot emphasize enough how this debt system was more effort than the value it gave.  I tried a lot of tinkering with it, but was very happy to exterminate it once I realized negation didn’t actually belong in this game.  For a few months the game required a central ledger to record player transactions because of this system.} 

[Here’s some fun facts: There was a significant amount of time where all agendas scored 2 points per match, to make it more reasonable to trade points with other players.  I also had a time where players bid points for turn order.   

Now that I’ve introduced the random negotiation focus, I can return to talking about the combat system.  The display of 4 cards served a dual purpose this version, during deployment it determined what you could add to your hand and during combat it was a timer.   If a player took an “alternate fight turn”, they’d take one of the display cards to their hand instead of moving something.  Once those were depleted the game would return to a 2^nd deploy phase or end. 

Using these cards as a combat timer was a clever idea that caused me a lot of pain.   It coupled the mechanism that determined which cards you could take with the length of each fight.   This required me to make its size dependent on player count, but more importantly I eventually realized that I had no easy way to adjust combat length without having unintended side effects. 

Also, this system was a return of the funky hand-size economy from 2 versions ago.   If you take an alternate fight turn, you have an additional card in your hand for the rest of the game. This was pretty powerful, so there were conditions that needed to be met to take an alternate fight turn (and thus decrease the battle timer).  Firstly, it was available if you had no legal move with your commander. More interestingly, you could take an alternate turn if you had the approval of other players. This was the heart of what shifted the game into negotiation…  deals where approval was traded for some promise or points were essential to the game. 

I’m not done describing the nonsense I added!   Apparently, I thought that the special face cards should be made more interesting to go along with this new world they existed in.  Kings and Queens had a scaling strength which was determined by each matching color card on one of their corners (and then multiplied by 3). Jacks had no loyalty; they could attack and be attacked by their own color.  In case you haven’t noticed, it’s incredibly easy for me to embrace stupid ideas while I’m overwhelmed by the momentum of doing a redesign.  These rules for royalty added such an absurd overhead to the game, and there’s already plenty of other stuff in this version that made the teach unnecessarily confusing.

This iteration stands out, it’s such a jarring shift in the game’s intent and I tinkered with it for 5 months.   

Like I said, the sliding cards only lasted 1 playtest.  Leaving with it was the dynamic strength of Kings and Queens. Next it became obvious that the scoring agendas should be static, so only the command card was swapped every round. 

Eventually the command card swapping was axed as well (because it complicated teaching the game).  Instead you played a card from your hand at the start of combat to determine your command card (an excellent change for game flow,  it also makes sense to only have to make the decision when it’s relevant). 

I also added more conditions for ending the fight, including having only 1 color left on the battlefield (meaning  nothing else could die) and premature ending as long as there was a unanimous vote from all the players.  This latter idea was pretty interesting, it was a memorable part of the game.

One of the things this version succeeded at was unique group dynamics, especially in the conflict between those who want a battle to stop and keep going.  Notably the 2 player count rules died again, and my target player count went back up to 3-5.  {After a bit I killed 5 player as well, since player’s didn’t have enough influence on the game and had too much downtime}. 

Overall, I consider my foray into negotiation and group scheming a distraction from the true form of the game, although I can’t deny that something was lost when I cut it.   One of the big things that helped me kill it was how much I hated teaching the game. Eradicating all the traces of negotiation made the game two magnitudes easier to teach. 

I also realized that the negotiation was something I was pushing as a player.  Without my encouragement it wasn’t something that people naturally found themselves interacting with.      The final straw that killed it was one playtester who made me uncomfortable with the out-of-game offers they jokingly made {I’m grateful to them for two reasons,  firstly because I’m glad for the indirect  prompting that ensured negotiation was cut, but also in a later version they helped me understand a fundamental flaw with the design} 

I actually got close to thinking the game was done during this iteration.  Lol…lmao even.  One of fun parts about game design is that you can have your entire sense of progress shaken up.  I spent quite a lot of time happy with the late stage of this iteration, my changes kept getting smaller and smaller… until one day all of the data from my various experiences playtesting coalesced to tell me that something was wrong.  Oh also a playtester (whose opinion I value) told me the game had no strategy, the painful part was realizing I agreed with him. 

This was pretty jarring, given that I thought I was at the point where I all I had to do was discover a good name & theme. 

.

.

.

A few weeks before Major version #5 was born, I discovered the Kansas City Game Designers.  This group deserves mountains of shoutout, both for contribution to the development of this game and immense personal value.  {Thanks to all my friends there… I’m incredibly grateful to have you in my life. I legitimately look forward to these bi-weekly meetings more than anything else}

It was life changing to get consistent access to playtesters.  Just by joining this group my entire relationship with game design has changed.  Before I was like a circuit designer who relied on static electricity and lightning strikes to power my work.      {It is also magnitudes more helpful to get feedback from other game designers instead of random gamers.  Additionally, I’m glad I have an opportunity to playtest other people’s games, I consider it a vital part of my creative process}

Most of the playtesters credited for this game are members of the Kansas City Game Designers.  Without them Ambivalent Aggression would simply not exist.  These people were essential to discovering what the game wanted to be, fixing problems, and finally gifted me with the confidence and momentum I’d need to finish. 

.

.

.

Major Version #6- Skirmish

I just read through the rules for this, and I’m not sure I even ever tested this one!    The idea here was to make the game more strategic by focusing just on the combat.  That’s pretty funny since so much of this game to me is about balancing the two halves.

We should give some respect to this version, even without it actually seeing the light of day, here many foundational ideas were born.

 I introduced the concept of a left and right agenda. A fantastic idea that’s one of my favorite parts of the final game.  Firstly, having 2 scoring agendas is 3 times as interesting as having 1. Especially since there are only 4 suits, each person having 2 random ones creates much more interesting player dynamics and relationships.  Speaking of relationships, the best part is how this mechanic gives you a knowledge-relationship with your neighbors. It changes the strategy a lot to have a partial idea of what some other players want, especially the ones who also are adjacent in turn order. 

The next banger idea was having multiple commanders and a restriction that prevented you from using the same one two turns in a row {like scythe} . This vastly expanded the potential for combat choices in players, without sacrificing the point of even having commander cards.

Finally, was the idea of exterminating everything that matched your agenda.  This was a suggestion from a playtester, inspired by shooting the moon in hearts {which I still have not managed to play yet… someday}.  One of the largest problems with my game was the way it could become pretty obvious that your agenda has no chance of scoring any points due to low presence on the board.  It was common for a player’s fate to simply be decided by the agenda they were dealt, and not be a result of their strategy.  {to say the obvious,that’s a pretty terrible flaw for a game to have! It’s still hard to believe that I thought version #5 was close to finished}

Now this  is what we in the business call a fantastic idea.  Not only does it fix the aforementioned glaring flaw, but it makes the whole game way more dynamic.  It introduces a fundamental layer of choice and evaluation to each player,  it adds a vital element of uncertainty to other player’s motives,  it adds more weight to what it means for a card to be killed, and it forced me to give each player 2 agendas instead of 1 (so that it wouldn’t be so easy to go for extermination).  If you’ve played the final version of Ambivalent Aggression, you’ll know that it’s all about balancing the 2 possible game ends. I was only able to find that identity because of this mechanic.  

Oh, also I tried to add sliding movement back (lol), but this time it was a choice where you could move 1 step or slide in a direction. There was also a mechanic for accumulating more commanders, and a mechanic where you burned commanders to take an extra turn.

.

.

.

Before I describe the next version, I’ll step back to #5 to talk about a design consideration I spent a lot of attention on.    The game had a prescribed arc of stages. First deployment, first battle,  second deployment, then final battle. 

This was arbitrarily chosen from the fact that there are 2 jokers in a deck of cards. So I spent a lot of time questioning whether there should be more or less battles. 

To me there was a strong argument for not having 1 battle, due to the new-player experience. The idea being that scoring a battle being the halfway point means that players will actually know what they’re doing by the time the 2^nd pair of deploy and battle comes around. {where before many people would be fumbling around}. 

Yet it’s also much more intuitive for one round of the game to simply be a deploy stage than a battle stage.   With two stages, the definition of a round of the game would always feel arbitrary and inelegant.  Also, the way the battlefield persisted from one stage to the next created a handful of problems. Most egregious was having to transition to a different scoring agenda. 

The question of which phases of the game were the most interesting started to haunt me.  If the first and 2^nd stages didn’t feel different then it’d be hard to justify putting both of them in the arc of the game.  What consumed more of my attention, was the split between deploy and fighting.   If one of those wasn’t interesting enough than I’d have a game where you just go through tedious motions in order to get to the actual interesting part.    It was essential that they both felt strategic, and that the play in the deploy phase created a meaningful game state for the combat.

You can tell from the fact that I drafted a combat only version #6 that a part of me considered deployment unnecessary, that I should have more focus in exploring the potential of moving cards.  I think it’s also worth mentioning that version #6 probably didn’t get tested because I couldn’t truly get on board with removing deployment.

There were also timing problems for when combat started. Where should the jokers be in the deck? Should it be possible for them to be drawn together (obviously not because then the 2^nd deploy can be skipped).  How should I reduce the randomness of their deck position to ensure they aren’t dealt to close to each other?  Forcing them to be in a certain section of the deck complicates setup in an annoying way.  

Furthermore,  If you had no cards to move around and cards you wanted dead on the current battlefield, then you were just sitting around waiting for the battle to end.  In order to prevent this state from occurring immediately, I introduced an incredibly inelegant rule which required all 4 suits to be on the battlefield before a joker was taken.   This reduced a lot of the tension from “when is a joker going to show up”, since this secondary condition could often be manipulated.   

To complicate matters, I knew that after the first battle there should still be a sizable portion of the board left, since that’s the seed for the 2^nd stage. This means that I had a large interest in the duration of each battle, something that the rest of the design made incredibly difficult to control.

All of these issues had to sit with me for months before I suddenly had the epiphany which would spawn the next redesign.

.

.

.

Major version #7- Start and Stop

Here is the answer to all the concerns I raised in the last section:   It’s not my problem! In other words, why not give complete control of the game phase to the players.   Let’s see how everyone else likes balancing the duration and frequency of these two gameplay modes themself.

Finally, I had discovered the true identity of this game:  balancing these two distinct yet highly connected halves.  In order to make this even more interesting, I expanded the game to have 2 possible ways of scoring.  Either all the surviving matches would count, or all the matches that you had personally killed would count.    The evil part is that the scoring method depends on groups actions, everyone has a huge investment in whether someone manages to achieve an extermination (which triggers the latter scoring method}. 

Many of the great ideas from version #6 got to shine here.  Left and Right agendas added a great layer of understanding what outcome your neighbors are trying to push for.

I also realized an elegant way to implement commanders, whenever the game leaves peace your hand becomes your 3 available commanders.   I especially loved this because of the great side effect of tracking the game state. If hands are private then its peace, if hands are faceup then its war.

Now that the game had the concept of trophies and using up commanders, I could finally fix the “alternate fight turn” {which I concluded had to be in the game, both as a fail-safe against no possible moves and to expand the ways combat can be productive for someone}.  First, due to the premise of this redesign, it no longer needed to be responsible for ticking the fight timer which was a huge relief.  Next, I could replace gaining a new card to hand with taking a trophy, which fits so much better in the flow of the game.

{As an aside, one interesting thing I discovered along the journey of designing Ambivalent Aggression is that 3 card hands are a perfect sweet spot to make the game as strategically interesting as possible. If your instinct is that more cards = more options = more interesting decisions, then I encourage you to try to derive my argument yourself. [It’ll be an enjoyable game design exercise] I think Slay the Spire, Calico and Reef are great games to study concerning this question}

This version was an immediate success from the first playtest.  In fact, this generation is the one the final game uses…I just had to spend a couple months studying it and making changes.   During these months, I simultaneously felt confident in being close to finishing the game and terrified of the possibility that I’d “discover” a brand-new redesign. 

Over this period, I became very critical of the game.  I would go into playtests expecting to learn that the experience was fundamentally flawed, and   I gave much more weight to any criticism I received.  There were multiple times where I would finish playing a test session thinking that the game sucked, only for everyone else to say it was great.  

 I also became quite annoyed with the difficulty of teaching the game, so started an inquisition against fiddly complexity.  Every aspect of the game was repeatedly put on trial, where I evaluated if it was interesting enough to justify its weight in rules.  Finding unnecessary complications to remove was quite a difficult task since this design has such heavy coupling.

To be clear, this was an unhealthy relationship with my own game.  Near the end of development, I lost a large portion of my ability to be objective in evaluating the design and had to build my confidence purely on other people’s opinions and statements.  {Thanks again to all my designer friends who reminded me that they liked this game}

Yet during this period, I also made several incredibly high impact changes.  This turmoil was definitely worthwhile and passing through it improved my confidence that what I had was worth locking in. I put a lot of pressure on getting this game right, especially since it was my first long-term project.  Doing so has given a great gift to my present self, now that I have some distance, I’m incredibly happy with Ambivalent Aggression.  I suspect this will be a project I remain proud of, even as I grow as a designer.  My brain can’t conjure up any doubts about it that can hurt me now, I already went through everything it had while trying to finish the game.

That would sure be a great ending to this retrospective.  Too bad I have more to say :)

Now I’m going to go through some of the interesting problems I battled in this final stretch.

• What should be the limitations on switching the game mode?
• Will the game be playable at 2 players???
• Agenda Distribution
• Maintaining the battlefield grid

It would be quite silly if there was no friction in switching the game mode, each player would simply start each of their turns in the mode they desire. The idea of game state would have much less weight and in turn would hurt the strategic dynamic between players.  Yet it’s also silly to put severe costs on people’s ability to switch modes, since that’s the core gimmick of the game and it deserves to be an essential part of the play.

I think the solution I came up with while drafting version #7 is quite clever.  As far as I can tell, it’s also quite unique. I’ve played quite a few games and can’t think of anything like the joker/marked system in Ambivalent Aggression.  The only cost to switching the game-mode is taking a joker, which binds your ability to switch the game until that Joker has been taken from you.    In a 3-player game there’s only 1 joker in circulation, so the moment anyone else performs a switch you regain your freedom to do so. In a 4-player game there’s 2 jokers, so it introduces an element of considering who you want to gift this freedom too. 

I quite like this system, but I was bothered by doubts if it was a tight enough constraint.  It’s possible for the game to constantly switch modes like 2 divorced parents arguing over custody.  I eventually made peace with this fact, but before then I spent a lot of time trying to add extra penalties to possessing a joker. 

Originally, owning a joker also prevented you from replacing cards during peace.  It’s hard to remember why I thought this was necessary because of how stupid I think that rule is now.   I remember originally having the opinion that taking a joker should be a strong cost, and it felt natural to double-down on how switching binds yourself.   The glaring flaw with this idea is that if you’re ever forced to switch the game from war to peace to prevent a threat of extermination, then you yourself lose your ability to acquire more points that would score if an extermination happens.  This is like a kick in the face to the intent of the game, a game that’s supposed to be about balancing two possible outcomes becomes egregiously simplified into tedious motions of just bolstering your presence on the board.   

Also not being able to replace cards just generally makes the game less fun.  It’s easy (in hindsight) to see why this limitation was stupid, but I had to engage in hours of internal arguments until I decided to remove it.

Then after that I still had a gut feeling that there wasn’t enough penalty for switching.  I considered adding a victory point penalty for possessing jokers or even at one point brainstormed additional actions to add to the game that would be locked if you had a joker.   All this effort and I think I’ve squarely learned the lesson that I was overly worried about adding a cost to switching game-mode.   Surprisingly, a game about adapting works much better when you don’t have to sacrifice your firstborn to switch things up.

As I’ve already mentioned, this game has had a cursed relationship with player count.  Sometimes it was a dedicated 2 player game, sometimes it was a group game that fundamentally couldn’t work at 2 players.  I adore 2 player games, it is my favorite player count by far {this feels like something of an unpopular opinion among boardgamers.}

It felt like such a shame to drop 2 player support each time, especially because it made the game more logistically annoying to playtest.  This version I tried developing a dedicated 2 player variant, with an entire set of joker actions to encourage players to keep passing them between each other.    After a few playtests I had to make the hard decision to ditch it completely.   Firstly it was taking attention away from the 3-4 player game which was obviously the main game,  and many of the dynamics that make Ambivalent Aggression interesting just simply don’t work in a duel format (despite how hard I tried,  I *really* wanted to figure out a way to make it work}.

The unfortunate cost of this decision is that it’s harder to show this game to my friends, since we’d need at least 1 other person to play it.   Two of my best friends have only played incredibly cursed 2 player versions that were scrapped soon after those sessions.  I wasn’t able to show my dad this game when I visited, which feels like a massive shame given that it started with just us two.

One of the original intents for this game was to make a game that utilized the fact that I carried around a deck of playing cards… so I could have an interesting game for any annoying pocket of time that I need to kill.   Since Ambivalent Aggression doesn’t support 2 players, I’d say I’ve failed in this intent, as that’s the most common situation I’d want a card game for.

One of the most important series of improvements I made to the game during this period concerned agendas.  

Originally, players choose two agendas for their left and right neighbors from their starting hand.   I eventually realized that this wasn’t a terribly interesting decision and therefore didn’t justify it’s weight cost {which was non-trivial, especially considering it asked players to make an important decision at the beginning of the game.}  This method of agenda distribution also made the game less interesting in a subtle way, because it made it more likely for players to have similar agendas. [There’s two ways this occurred. First there was a clever strategy where you could give both neighbors the same suit, creating forced competition between them while making it more efficient for you to mess with other people.  Next, many players would independently come to the same conclusion on what to give people based off the available information.]

Interestingly, I only discovered these flaws as a result of investigating  2 other agenda related problems. 

1. The double agenda edgecase
2. Too many agendas

What happens when both your left and right agenda are the same suit?  Should you score that suit once or twice?  Is it fair that it’s much easier for you to trigger an extermination?  Originally, I thought this edgecase was fine, I justified a higher scoring potential from double counting with the idea that having only one suit was much swingier and risky… so therefore it was surely balanced.  

One of the playtesters convinced me otherwise.  First of all, it didn’t matter if it was balanced, the focus and lack of options from being dealt a double suit was simply not fun {especially since it’s not something you signed up for.}  Then it wasn’t even balanced, it gave a significant advantage.  Worst of all it gave a player an inherent incentive to go for aggression, which is a total disrespect to the intent of the game

My first instinct to fix this was to put some kind of setup fail safe to redeal agenda cards if anyone had two of the same.  Unfortunately this is incredibly clunky/tedious and it’s hard to guarantee this process wouldn’t have to be repeated.  The better solution I eventually came up with was the “virtual agenda rule”, which transforms 2 agendas of the same suit into the 2 suits of that color.   This rule is still probably quite confusing to read in the rulebook, but compared to what could have been it’s an incredibly elegant answer. 

 When I introduced left and right agendas, they were incredibly fiddly.  This is because each player had their own set of 2 agendas, meaning a 4-player game would have 8 agendas in play.   Between each player was a pair of agendas, one for each player, that both of them could peek at anytime. 

This was INCREDIBLY annoying. It was too much cognitive effort to keep track of 4 different agendas, 2 for you and the 1 you knew from both of your neighbors. It led to people constantly checking the facedown agendas, a physical action that slows the game down a surprising amount.

I was searching for a way to reduce the amount of times people had to check their agendas and then had a playtest where someone misinterpreted the rules, thinking that all 4 agendas they could see were what they were scoring with.  In order to prevent this from happening, I decided to investigate the idea of shared agendas between players, so that you weren’t seeing agendas you couldn’t score.

I was convinced this change would destroy the strategy of the game, because of the unfair relationship it would create with your left neighbor who went before you in turn order.  It turns out I was utterly wrong!  This is one of the best changes that I ever made to the game.  It did wonders for simplifying the unnecessary cognitive load and agenda checking.  Not only that but it created even more interesting strategic play and relationships between players. 

In hindsight, the previous way of having player agendas was just deranged.  Having shared agendas between neighbors is so much better, I only needed 1 playtest to know this should have been what I was doing the whole time.

Interestingly, this was one of the last changes I made to the game.  That’s pretty funny to me considering how obvious it feels in hindsight.

This has been an issue ever since I introduced the concept of moving cards in the version that added combat.   Since the game is just played with playing cards, “the grid” is an abstract layout that must be eyeballed by players.  This works completely fine, until cards start moving and enough variation adds up to make it unclear whether something is 1 or 2 steps away.  One of the advantages of pure sliding movement is that distance doesn’t matter, but unfortunately, I decided to use movement mechanics where it does matter. 

So obviously it’s my burden as a designer to figure out how to ensure the grid gets maintained, it would be silly to expect this off the players.   For a while there was a “desert” mechanic, where any cards not connected to other cards would disappear at the end of the battle.   This still allowed multiple islands to form, which was not ideal. It also made it too easy to kill cards.  

During the first playtest session of version #7, some card grid issues came up.  By combining built up frustration from this problem and playtester suggestions,  we decided that if a card lost adjacency to other cards then you should just move stuff to make it all one block of cards again. This way we could just avoid the problem!  The neat part about this fix is that it adds more tactics potential to the combat, since now you have a new way to move multiple cards at once.    It’s super cool how this “lazy” patch actually made the game more interesting.  Although now that I think about it again, I have no idea why I didn’t simply introduce a rule that prevented you from moving a card away from the battlefield in the first place. 

.

.

.

After my final playtest I realized I had to be done with the game because of how sick of working on it I was.  Conveniently, I was very happy with that final session, although amusingly there’s some rule changes that were only tested that one time (such as shared agendas).  I don’t think I would  have actually let myself feel finished if I was unsatisfied with the game, although had that happened it probably would have been a miserable experience.

Then all that was left to do was write up the rules.   Throughout development I had felt smug against my fellow designers for having a living rules document as I worked on the game.  Turned out that was useless as a consumer way of learning the game and was written only for myself…. so I had to start from scratch.

This was the worst part of my experience making Ambivalent Aggression by far.   I knew the game was *done* and was eager to release it but had to take the time to get this right.   I spent a week crunching on writing rules.

It was incredibly hard to focus, yet my immense desire to be free from this game forced me to keep coming back to writing.  I often wrote for hours and then got frustrated as I suddenly lost steam.  A more intense version of the experience I had finishing this design diary actually.  {I really need to find healthier ways to treat writing! Learning how to properly handle hyper fixations and ADHD is quite difficult}  

It's hard to say I learned a specific lesson from this experience.  You could say I should put a lot of effort into writing rules earlier in the development process, but is that really worth it when games are so malleable?  At the very least I’ll experiment with some other way to treat rules writing in my next long-term board game project.

.

.

.

Thanks for reading my design retrospective for Ambivalent Aggression, I hope it was interesting!  This sure took a lot of time to write. 

I’ll end with the sad thought that this game probably hasn’t been played once since I did my final playtest 2 months ago.  That’s depressing to think about, but I still found this game worth making.