The Evolution of Gaia

A few months ago I mailed copies of Project Gaia to four groups for testing. The game was only about 6 months old, but it had gone through several iterations, felt "okay," and good enough to send to folks without my presence. The goal was to attain validation for the concept. Not to hear "this game is good, box it up!" but to gauge the pulse of folks and get a general thumbs up or down on the concept.

The tests have gone well, but the game had a lot of problems. Most glaringly were the issues with complexity. The game had a lot going on, and it took me a while to find out what was needed, why, and how to do it simpler. Thankfully, one test group in particular, Ruth and Jeff Ashton, stuck with me over 3 or 4 iterations now.

An in game photo taken by the Ashtons An in game photo taken by the Ashtons

I received a very positive test report from the Ashtons Friday, which was a really nice affirmation of the work that's gone into the game. It feels like the project is really turning a page, so it seems like a good moment for reflection.

I want to write about many of the changes that have been made. I'll try to cover it at a high level so it's both interesting and useful to folks who aren't intimately familiar with Project Gaia. Another way to see some of these changes are to watch two developer vlogs I recorded, showing the game at two points in time.

Here are the rules for the game. The Print and Play is linked at the top.

A quick explanation of the game: Project Gaia (name TBD) is a combination of my love of CCGs, like Netrunner, and tile games, like Carcassonne. I wanted to make a game where building a deck was core to the experience, but I knew I couldn't make a full CCG. I just don't have the testing resources, the financial resources, or the audience.

Therefore, in Gaia, players build or draft decks of 9 cards from a limited pool of 55 cards. There are 3 card types: powers (action cards you play then discard), creatures, which hold territory and attack, and monuments, which act as tiles that grant bonus actions. The cards that aren't used have Terrain on the back and they are placed as tiles to build a planet.

Players alternate taking actions on their turns to play cards, manipulate the planet, and battle. The result is a 30 minute card game with a nice amount of depth and interaction.

Card Costing: A key to games of this nature is cost. In Magic, Netrunner, and X-Wing Miniatures, cards have a mana cost, click and credit cost, and squad cost, respectively. You even see this in Star Realms and Dominion. If a card costs too little, it'll become overly dominant and can ruin the game. If it costs too much, players will avoid it in favor of something that's easier to get out.

In card games of this nature, tempo and pacing are crucial to manage.

I love multi-use cards. Borrowing the Magic mechanism from Summoner Wars, cards originally required you discard other cards to play them. The cost ranged from 1-3.

Going back to the very beginning of the game, this has caused all sorts of problems. Initially, the game was laboriously slow -- you were constantly left with no cards. To counter this, I gave you a free draw every turn, and you could spend your entire turn to draw back your discard pile (somewhat like the discard and draw action in Combat Commander).

But, the problem was that people would spend all their cheap cards to play the good cards. You actually saw this a bit in the Summoner Wars meta where players would completely forego commons to play only Champions. The game ground to a halt, as it would take multiple turns to fully draw up, get your big creature, then maybe lose him, forcing you to refresh.

I also noticed creatures didn't have a lot of purpose. I'll dive deeper into the iteration of creatures below, but I added a mechanism where every creature you had in play reduced your discard cost by 1. So, if I have 2 Creatures out, a 3 Discard card would cost 1. While this idea is simple, it was very complicated. Players constantly forgot it. They were overwhelmed by the math, as well. Think about it -- if you're evaluating 5 cards in your hand, and all of them have a simple math problem, that requires you work much harder to decide what card to play.

DiscardReminder The reminder card

Shortly after I added reference cards to help you remember that you had a creature discount, I took a step back, and admitted I was devoting a ton of mechanisms to making this single mechanism work. Basically, I was putting a lot of good money after bad.

I removed the discard cost entirely. It was immediately more fun. Players started with all 9 cards in their hand and they could just play them. One action, one card. Suddenly, the game was simpler, faster, and players were able to enjoy the full range of cards. I removed the free card draw at the beginning of the turn.

However, this too had a few problems. For one, nobody ever spent 1 Action to draw 1. Players would spend their full turn to draw their entire discard around turn 3 or 4, and would never need it again. This felt like there wasn't really an interesting economy of decisions related to card draw. I removed the 2 Action - Draw your entire discard option. Now, you spend 1 Action to get 2 cards (at random). Therefore, fewer actions, and one that's more consistently used throughout the game.

But, finally, there was one more issue -- some cards were clearly better, but all had no cost, except the single action. My option here is to make all cards consistently powerful, which is both difficult and, in my opinion, boring, or implement a cost. I learned my lessons from the previous iteration, so I returned to the discard, but a much simpler version. Now, cards either cost 1 Action, or they cost 1 Action plus 1 Discard. The 1-3 is gone.

GaiaReference

In summary, there is now a nice way to balance better cards that isn't complicated and doesn't require supplemental mechanisms to make it work. There is a nice hand management layer to the game. Turns are simple. Take two actions, which basically means playing two cards, using two cards in play, drawing 4 cards, or some mixture of these.

Creature Evolution: Creatures needed a lot of love. When I decided to have a strong spatial element, I felt like it only made sense, thematically and mechanically, to have creatures on the worlds you're building. Summoner Wars is my primary inspiration for Creatures. I wanted them to move simply in a grid system and attack to protect your other creatures, protect monuments, and foil an opponent's plans.

All Creatures have either permanent passive bonuses, conditional bonuses (do this to get a thing), or Actions (like many Monuments). For a while, creatures were in the game just to be there. You didn't need them for anything and often, players wouldn't play them. Therefore, to help combat the card discard problem, I made it so that creatures reduced the cost by 1. If you had all 3 creatures out, your cards were free to play!

This is mechanically simple on paper, but as I noted above, was too complex and didn't work. Then, I made it such that Monuments could be used by players to complete objectives. I also made it such that Creatures could block regions from use. However, as combat began heating up, players noticed a few problems:

  • If someone just flooded their deck with Creatures, it made them all free to play, and an opponent would be unable to knock them off or score.
  • Creatures were so flimsy. They all died in one hit, which meant they were impossible to keep on the board.
  • Creatures could be played anywhere, which made it even easier to just hot drop a creature next to an opponent's and kill it. Whack a mole!

I added three fixes to address these. I added a Deck creature limit of three, I did a tuning pass of health and attack, and I made it so that creatures had to be played to a specific land type. This really improved things!

After listening to a fantastic Mark Rosewater Drive to Work podcast about worldbuilding, I decided to put his teaching to work and really think about how the creatures belong to the world. It was a really fun exercise! I thought about each terrain and what it meant for the creatures in that ecosystem. I tried to design key principles for each one. Then, I listed common and fantastic creatures one might find in such a region. Finally, I put the names to paper and tried to find a marriage of theme and mechanism. I think it was a really fruitful exercise. As the game tightens up, I plan to dig into the theme more strongly to find how I can add more of those touches to the experience.

You should read the scoring section for further details. Creatures went hand in hand with those changes. Mostly, creatures became simpler - no card discard cost, no range - but also became more integral to scoring. Players had to have at least one Creature or Monument involved in an Experiment to Score it. Then, only cards covered by your Creatures could be added to your Biosphere. This made Creatures integral and simple. You can take the ground you cover, essentially.

Finally, for now, I recently did a tuning pass on Creatures to further diversify their values per feedback. Now that I had the discard cost back (though only a single discard), I could make some creatures more powerful. I also began using the knobs more fully by experimenting with Movement speed, defense, and the bonuses provided by the creatures. The hope is that a player will choose 3 Creatures that have synergy with their other cards. Somewhat like how Ice or Programs will really dictate how your deck plays in Netrunner, Creatures will be the primary movers and shakers in Gaia.

Here are some creatures.

Creature_BoltMoth

Creature_Serpent

Card Design Evolution: I'm very proud of the work done here. I've put a great deal of thought into it and have seen great results come from investing time into the quality of the layout. Now, it's all placeholder. Obviously, a professional will do the work if/when it gets published. But, for now, I'm super proud of the results.

Initially, the cards had far more components, so more needed to be displayed. A creature, for example, had:

  • Name
  • Creature Symbol
  • Tile Affinity: Basically, play this to a Grassland
  • Discard Cost
  • Card Text
  • Creature Stats: Defense, Movement, Range, Attack
  • Creature Discard Reminder

Yowza! Naturally, some of these things were lopped off as the game simplified. Range was eliminated as a creature component. The discard reminder died with complex discard cost and was replaced with a simple discard symbol.

However, the Creature symbol needed some thought. It seems obvious now, but it took a minute. I wanted to have symbols on the cards to represent the type. The idea being, players would see that symbol, think "this is a creature," then remember from the rules, "Creatures are played like this."

The problem was that with 4 card types (it's now 3), all of which played a little differently, the symbols didn't help. Players basically had to remember 4 rules that weren't reinforced on the cards. Lame! It was frustrating, because the rules were really simple.

  • Powers: Resolve the card, then discard it.
  • Creatures: Play to the tile type indicated.
  • Monuments: Play to a Desert.

But, people kept mixing things up. I took inspiration from Ashes from Plaid Hat Games. On their cards, they tell you exactly where cards are played. They use simple phrases like:

  • Play to your Spellboard
  • Play to your Battlefield
  • Play then Discard

I thought about it. Why can't I do the same thing? I removed the useless icon and at the bottom in tiny text, I just told players how to play the card. Here's an example:

Creature_GrandBeaver

Monument_Beacon

For each card, Creature and Monument respectively, the bottom tells you how to play them. Just a nice reminder that is driven home. I did a few other simple things to convey differences. Notice there is a bullet shaped frame on the Creature. The idea is to convey, you play it to this type of card. But, on the Monument, it's in a box. Now, a real designer will improve on this, but one conveys an action, they other conveys a permanent state.

There are other cues I can provide eventually using color and shapes. In fact, I found some art on the internet, just as an experiment, and put together some card mocks. Now, I did them. They look bad. But, it's fun to see what they could look like with more than white backgrounds.

Mock_BeaconoftheSteppe

Again, ignore my terrible choices. But, you can see a generic grasslands shape on the bottom, so players always know "this is a grasslands." That lets me remove the icon in the top left. I made a specific icon - the fence - so you can also know that it's an icon. Finally, more fun art!

I did a similar exercise with a creature.

BearMock

Here, you have a generic forest silhouette on the bottom to remind you where to play it. But, it isn't filled in, so that it doesn't state that it IS a forest. You then have the pertinent character stats in the top left corner, all chosen to represent a creature.

Gaia is a complex game with 55 unique cards. I've had to put a great deal of work into the layout of the cards sooner than typical to help facilitate that learning. But, every step has paid dividends. Tiny, subtle tweaks have noticeably improved the enjoyment and comprehension of my testers.

Planet Construction: Originally, the players built the planet at the start of the game. After building or drafting decks, they'd turn the cards over, shuffle them, and deal 3 to each player. Players would take turns placing these on the Planet, then drawing a new one.

There were two early problems. One, the planet was too big. It had a little too much of everything and there was no conflict or tension. Players would just build what they needed in their own corners. Secondly, the rules were too restrictive. I said you had to attach a card to one of its type. This meant you'd effectively have a Neopolitan planet of Grasslands, ocean, and forests, cleanly separated.

I shrunk the planet from 15 to 11 cards and added an initial seed - 3 random cards played diagonally. But, if one of each type wasn't played, this meant you could play a card anywhere. So, now it looked like a slightly melted Neopolitan. I removed the restriction entirely. Play a card wherever you want, as long as one of its sides matches the side of the same length of an existing card.

Fundamentally, though, this section wasn't interesting. It added another 3-5 minutes to setup, depending on the AP of the players. It also felt like a choice that I didn't want to be a choice. Players would try to build the planet "correctly." My fear in providing a pre-arranged layout was that the game would become static. But, I ripped off the band-aid to implement the following solution:

Players chose 1 of 3 pre-defined layouts shown in the rules. Cards would be shuffled and randomly dealt, so you'd have a different layout with a different assortment of cards every time. This was simple, and worked pretty well. Then, when I added the special tiles (Mountains, Fjord, Fissure), I had the final twist. After the layout is complete, players alternate placing the special tiles. They can place them in deserts (blank spaces), or displace another tile, shifting the row or column to make room.

Now, the planet is setup quickly, with randomization in the tiles, and a slight, quick player twist to really get it going. The key summary here is that I simplified it and focused on what the game really needed. Building a deck is the cool part. Not tediously building a planet!

Tile Evolution: Project Gaia was in my mind for months before I figured out how to make it and begin testing. I knew I wanted to make a tile game. In fact, the original original idea was to make a game where players would create tile sets, like decks, that they would then use. But, this had some weird product complications, and tiles were too small to give me the flexibility to make a broad, robust game. When I realized I could use the cards as tiles, and save cards by using the cards that weren't selected, it was a real eureka moment.

Tiles were originally varied and complex.

P_FLeft

O_F

O_DMid

Those are ugly, but you can basically see there are forests (green), oceans (blue), grasslands (brown), and deserts (tan). The problem with this much variety was that it was impossible to line anything up. It also made the game very complex in a weird way that wasn't intended.

Immediately, I shifted to simple, solid tiles, and cut it down to 3: grasslands, ocean, and forest. One neat idea to deal with blank spaces that would inevitably emerge due to shifting and removing cards was that blank spaces were deserts. This meant fewer cards, but I still had 4 states.

At this point, the Tiles felt rather dry. It seemed like the game needed some punch. Therefore, I introduced Landmarks. On some tiles, there would be bonuses granted to the player on top of them. Due to the random nature of what was in the game, this would add some variety to what powers were out.

ForestGlyph_2

The landmarks would do things like:

  • Increase card draw
  • Destroy planet tiles
  • Add planet tiles
  • Let you move creatures
  • Let you add creatures at no discard cost

The problem with Landmarks was two-fold -

  1. It added more complexity. There were just too many variables to track. And, you had to discern the icons on them, which was lousy.
  2. They were tough to use. The benefit of them was often outweighed by the cost of not using creatures to complete scoring objectives.

The landmarks also exposed a more fundamental flaw with the game -- the Creatures weren't interesting enough (discussed above), and the Monuments weren't interesting enough. Instead of adding more stuff, I removed the landmarks, simplified my rule set again, and strengthened my core content.

I reverted back to plain, simple tiles. But, the itch scratched again. First, I added Mountains. These two special tiles were added after you setup the original map. Players couldn't shift Mountains, which meant they locked territory, but also provided a defensive bonus to creatures. Temporary safe havens.

The mountains worked. They were a nice spice. Therefore, I added two more. The Fjord and the Fissure. The Fjord was trying to solve the problem of players just adding Tiles anywhere on the map and immediately completing scoring objectives. Now, you could only add tiles next to the Fjord. The Fissure was the Omega to this Alpha. Any cards, creatures or tiles, that moved onto the Fissure were destroyed. Basically, a caution zone.

The Ashtons reacted negatively to the Fjord initially. It felt overly restrictive and annoyed them. This is one of those cases where you need to take the feedback, but really think on it. I knew the game needed a constraint on adding cards. But, perhaps a single card was too much of a limitation? I made the following change, and asked them to try again.

Mountains now could not be shifted, and took on the Fjord's power. They no longer provided a defensive bonus. Effectively, there were two Fjords. No change to the Fissure. The change was received well! It gave players some flexibility, while also establishing basic limitations. With these special tiles, the planet was overall very simple, with just enough spice, and the Monuments acted as the primary points of differentiation.

Mountain Mountain
Fissure Fissure

Now, players can focus on creating and shaping the planet, and choosing whether to let an opponent keep their Monument in play to focus on scoring, or take it out, in the hopes of gaining momentum.

Scoring Evolution: This section of the game has probably seen the most iteration, aside from changing every single card for wording, balance, or functionality probably 30 times each. No exaggeration!

When I started the game, I didn't want this to be a war game. I didn't want it to be about dealing X damage (like Magic) or killing an enemy base (like Summoner Wars). I wanted to have an open path so you could use the cards in a variety of ways and hopefully have tons of variety. Therefore, I was leaning more towards the Netrunner system of scoring points, which can be done in a variety of ways.

I wanted the spatial element to be front and center. You're creating and shaping the planet, so that should be how you score. Initially, I had scoring cards with very precise goals on them. There were 7 in play. Once somebody scored 3, the game won. There was also a fourth card type that were basically powers, but if you met their condition, you'd score the point, removing the card from your deck. It was like having a secret objective.

Score3b

Above, someone would score if they had 4 Forests in precisely that orientation on their turn. This had quite a few problems. Firstly, there is the complexity of the shape itself. If the board starts with 11 tiles, plus player Monuments and cards, plus this can be mirrored...wowza! It's tough to watch all that. If you recall, this system is similar to Tash-Kalar which also has that, and it's tough there too. But, the surrounding framework of Tash-Kalar is MUCH simpler. Vlaada is a genius, after all.

Also, initially, you just got this if it was in the world. There was this agonizing problem of wanting to set yourself up to score, but not get close to it, then your opponent would score. Players would play chicken and have a staring contest. It would grind the game to a halt as players tried to setup the multi-turn setup to create the pattern and keep an opponent from getting it.

It was also tough to focus on all seven goals at once. Therefore, I put in 3 at a time. But, sometimes they were difficult to execute, or painfully simple to execute, based on the random start of the board. I created more to create more variety, but the problem still remained. Sometimes the game didn't jive well and it was tough to get the precise shapes.

I was worried about players just auto-completing them. I started putting in back pressures. To score,  you had to discard cards. Or, take an action. I tried several things, but the fundamental problem still remained, and adding yet another reason to discard cards exacerbated the discard card cost I discussed at the top. Plus, people were really struggling with the shapes.

I shifted to a much simpler system. Simply have a defined number of tiles of a type touching each other. No patterns, just assortments. Now, players merely had to create a pocket of 3 Grasslands, for example. This is round the time I started involving Creatures more into the scoring framework. An opponent couldn't use a tile covered by your creature to complete one of these goals. Players could use their Monuments to complete these goals, but not their opponents. Now, there's a layer of board control which started to create a more cohesive whole.

Here are two of the Experiments. One is a simple Planet one, the other a simple combat one.

Experiment_TheSeas

Experiment_Predators

There was still was the problem of the cost. Eventually, I kicked the framework and made it such that you would complete the goal at the end of your turn if the conditions were met. Three of these Experiments are in play at all times. At the end of your turn, you can score the three if you qualify for them. Then, any new ones are drawn. You would then get a reward instead of a single point.

These rewards led to the creation of the Biospheres.

Biosphere_DiverseA

Biosphere_StrictA

Above are the two possible Biosphere cards players would get at the beginning of the game. One flexible, one linear. When you met the condition, you'd get a reward. You could score a Tile on the planet to your Biosphere, completing that slot on the track, or you could do other things. This idea sorta worked, except it didn't.

There were too many confusing rules on what you could take, and why, and when. Players were allowed to force their opponents to take cards for their Biosphere, which would cost them points if it was the wrong tile type. So, putting an Ocean where you need a Grassland would hurt. There were also just too many symbols.

I tried again with a new iteration.

Biosphere_Ocean

There were 3 Biospheres, each associated with a tile type. You had a great deal of flexibility going down the track, choosing one available card in each row. If you chose the highlighted tiles, you'd get a bonus. This was better, but still too rigid, and players hated shoving a card in their opponent's Biosphere. It felt wrong for the game.

I tried to simplify it. To complete an Experiment, at least one of your Creatures or Monuments must be involved. This means you need to maneuver and have presence on the board. This is effectively "the cost." Instead of discarding cards, which is lame, you have to effectively do fun stuff. You know, moving creatures, attacking your opponent, and setting up your Monuments. I used a strength of the game as a cost.

When you complete an Experiment, you get 1 or 2 Rewards, but never the same one twice. These include:

  • Adding one Tile covered by one of your Creatures to your Biosphere. This was a nice simple solution. If you have creatures, and they are in position, they allow you to score good tiles.
  • Add a tile to the Planet from the Supply
  • Remove a tile from the Planet
  • Add cards back to your hand (if you want to move a Monument, for example)

Adding Tiles to your Biosphere Scores points. Looking to Coloretto, I tried something dead simple. The deeper you can go in a single color, the more points you get. If you have 4 Forests, your forests are worth 7 Points total. If you have 2, they're worth 2 Points.

This allows for flexibility and it's very simple. As a final tweak, you also get an immediate bonus based on what Tile type you add to your Biosphere. These help move the game forward.

GaiaReference_Back

On the horribly designed card above, you can see the 3 simple bonuses, as well as the card to point distribution on the bottom.

In a nutshell, the scoring is about manipulating the planet and marshaling your forces to control a sector. Then, you add tiles to your Biosphere for one time bonuses, but also, hopefully, focusing on 1 or 2 tiles to maximize your points. The game ends when the 6th tile is added to a Biosphere.

Conclusion: I think at 4500 words this has gone on far too long! If you have any questions or thoughts, just ask. I'd love to talk about Gaia and where it's going. Thanks for reading!