Palette Swaps: when variety underwhelms

Last week, I discussed why I am a big fan of variable setups, but acknowledged its shortcomings. The main one, I think, is variety that isn’t meaningful. Today, I dig into that a bit more and talk about palette swaps.

In video games (and I’m assuming animation in general), a palette swaps is when you create a new character by taking an old one and just changing a few colours here and there: that’s how you get Luigi from Mario, Subzero from Scorpion, and 18 different types of goblins in Diablo. The only difference between them is the colour: they use the same animations, the same code. It started as a way to allow two players to play the same character, yet still tell them apart if they were on the same screen, and it filled that role perfectly. Since, it’s mostly been used to generate multitudes of different characters and items and attacks and enemies, but which all feel the same. Many board games go down that road and offer a lot of empty variability, where things change in ways that don’t matter.

There are three things to keep in mind when trying to build variability in a game: the inherent properties, the scope, and the focus.

Inherent properties

I’ve talked about interchangeability twice before: once in my grammar-as-theme series, and once in Avoiding non-decisions. Inherent properties are the opposite of interchangeability. An element (whether a resource, a card, an action, a faction, any part of a game) has inherent properties if it can be defined without depending on other elements. For example, the tracks in Endeavor have inherent properties: Industry determines which buildings you have access to, Culture determines how many workers you get, Economy determines how many actions you can empty, and Politics how many cards you can keep. The tracks in Terra Mystica don’t: their only differences are which round tiles are present in a given game.

This is where many games fall short in terms of making variability meaningful. Starting with a boost on the Industry track in Endeavor will feel very different than one on the Culture track. It will affect which building I can start with, but limit how many actions I can take. However, starting with a boost on the Blue track in Terra Mystica does not lead to a different game than one on the Red: it only matters in which round tiles will give you income, and on where other players will go. They are different, sure, but only because they will evolve differently.

Picture from the publisher

When it comes to variable setup, your goal is usually to push players into different experienced, but for those experiences to be meaningfully different, they have to have inherent differences. To make sure of it, describe the scenarios, and try to see how many links to other elements you have to make before you use a different verb (or verb phrase), rather than a different noun:

  • 0 link: “In Ethnos, there are 6 types of cards in each game, out of a pool of 12. Trolls allow you to break ties. Giants score points if you have the biggest stack. Wingfolks place their tokens anywhere.
  • 1 link: “In Spirit Island, players will choose a Spirit, which gives them different innate abilities and starting cards (noun). (LINK) This Spirit’s powers allow you to scare away the Invaders, while this one is specialized in defending the land. (verb)
  • 2 links: “In Gaia Project, there are extra scoring conditions if you can do certain actions during a specific round. This one will give you points if you build a Mine, and this one if you build a Research Center. (LINK) A Mine gives you Mineral, and a Research Center gives you Research. (LINK) Mineral is used to build more Mines, while Research is used to improve a technology.

The fewer links you need, the more inherent the difference, and therefore, the more meaningful the variability: each link takes away from that variability both mathematically (because those aren’t perfect relations) and psychologically (because each jump makes it feel less meaningful). Of course, it’s not a perfect relationship, but I use that thought exercise in my designs quite often.

Sometimes, no amount of links will ever come to a difference. There’s no difference in Coloretto between starting with a Red or a Brown card. In that case, that is not variable setup. In Coloretto‘s case, the goal is to have every player start with a different card, which helps make the first round interesting. It should not, however, be considered variable setup.


Another aspect to consider is the scope of that variability: how much of the experience does it actually change?

As an example of analyzing scope in starting situations, let’s take a look at everything that your choice of Spirit in Spirit Island will give you, thinking about how different they make a game from another, and for how long those differences last:

  • Starting presence: within two turns at the most, your board presence is virtually the same as everyone else’s;
  • Income abilities: Minor differences overall. Some provide more flexibility, some less, but they’re all quite similar;
  • Innate abilities: Very important difference. It’s abilities that only you can do, and they’re rather strong and stay useful throughout the game. They also push you to acquire certain cards, with the icons you need to trigger them. Those are a 1-link relationships, because cards with specific icons will also have specific types of effects, meaning a card with Fire as a trigger will tend towards aggressive cards throughout the game;
  • Starting hand: Fundamental difference early, but since you acquire new cards quickly, and get rid of those you started with, you will end up with most of your cards being from the central deck rather than your starting hand.
Picture from BGG user TheBratPrince

Add it all together, and those Spirits make Spirit Island a highly asymmetric game. The Innate abilities, however, do most of the heavy lifting. I’d even argue that the starting hands are probably there to support those innate abilities, make sure you have the icons you need and have cards that follow its theme, rather than to increase asymmetry.

When it comes to variability of scoring systems, let’s look at Isle of Skye, which offers you 4 different scoring conditions every game. A game of Isle of Skye finishes with average scores in the 60-90 range. On average, about 10 points will come from stuff unrelated to those conditions, whether leftover money or acquired Scrolls. That means that, on average, about 80% of your scores will come from those variable scoring conditions.

Compare that with Prowler’s Passage, a really neat 2-player game. In it, every game, you choose 3 objectives, and the first player to complete each will get points, between 3 and 6. Players score, on average, about 100 points each. Out of an average of 200 total points, 12-15 will come from the achievement cards, or 7%. How much do these objectives push you to do different things from a game to the next?


Focus is an aspect that came to me while I was writing the two previous ones. In my earlier examples, I’ve talked mostly about strategy games, mostly Euro ones, because that’s my jam, that’s what I design and play and know. However, if you design a narrative-centric game, or a very puzzly abstract, the scope you aim for and the kind of inherent properties you care about will change.

If I play a fantasy skirmish game, I don’t care whether I play Dwarves or Orcs: what are their abilities? How are they different mechanically? However, if I play an RPG, leading an army of Orcs is going to be quite different than an army of Dwarves, even if, mechanically, they’re the exact same, because different story challenges will come out of it.

Now, I’ve seen many board games where they offer variability, but without focus. Many games centered around “missions”, which are mechanically just recipe fulfillment, resource management things, with very different lore behind them, but only meaningless differences in the recipes you needed to fulfill. Is that variety? Only if the narrative aspect is what matters to your target audience!

I think of Munchkin, where the target audience will consider a Magic Sword and Laser Vision very different cards, even if both just give you +1 in combat. I think of Claustrophobia, which has a few different enemies in it, but many criticize its lack of variety because “they’re all demons”, even if they’re mechanically different. Can’t please everyone it seems…

And that is not just theme vs mechanisms: you can also use that to dig deeper. In a game were movement on a map is key, like Hawaii or Maquis, changing the connections on a map would make for a very different feel, much more so than it would in a game like Godfather: Corleone’s Empire or Spirit Island, where those spatial relationships are important, but secondary.

Being even more of a nitpick: the character in Marco Polo which allows you to choose the value of your dice is exciting in that game, because the game’s focus is one of optimization, and the dice you roll just add that limitation to your turn. That same power in Sagrada would take away half of the puzzliness of the placement, remove from the game’s focus: much less interesting. It’s not that the mechanism takes a different place, it’s that the experience you want the player to have is different.

I always do this: I start with a short article, and then keep on adding on to it until it reaches “geez no one’s ever going to read this”. If you made it to this part, AWESOME! You win a virtual cookie!

So yeah, those are the aspects that I think one has to keep in mind to avoid palette-swap, skin-deep variability: how inherently different are the two things you’re comparing; how noticeable is the difference; and how fundamental to the game is the element you’re changing.

Next week, I think I’ll get into some examples of games which have done variable setup particularly well!

3 thoughts on “Palette Swaps: when variety underwhelms

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s