Meaningful Decisions: Donald X. Vaccarino on Design Choices in Dominion

In our Meaningful Decisions series, we ask designers about the design choices they made while creating their games, and what lessons other designers can take away from those decisions.

In this edition, we talk with Donald X. Vaccarino, the designer of Dominion, about deck-building, expansions, politics in games, variable setups and more.

Dominion established deck-building as a game mechanism, and many other games have sought to play around with the formula. How did you settle on the core elements around drawing, purchasing, discarding and reshuffling?

The premise was that you were building a deck while playing, and that everything was in the deck.

Since everything was in the deck, resources were in the deck. I didn't want to have it be too easy to get a small deck; so resources stay in the deck, you don't lose them. They're income.

Initially I considered having a variety of resources. It's simpler to have one, and means you don't draw a bad mix of them.

Initially I considered having to play a card to buy a card. But you'd always need those cards; maybe you'd lose yours or something and have nothing you could do. And some turns you wouldn't draw one. So, you can buy a card every turn, it's built in. And Copper costs $0, so you can always gain a foothold.

If you just drew one card a turn, it would take forever to see the deck you're building. So, you draw 5 cards a turn. You recycle everything so you can keep drawing it.

Initially I thought the choice of cards to buy would be like a line of cards, and when you buy one it's replaced. I worried though that too much of the game would be having a good card turned over when you got first shot at it. When it came time to try the game, I hadn't solved the problem. So I just put everything on the table at once. I figured, it would be easy to find the broken cards, and if the game worked I could refine it later. Of course it turned out we liked being able to buy everything. Then when I added more cards, I limited it to 10 that vary from game to game, because it's hard to keep track of even 10 different things, and 10 changing gives you plenty of variety.

You play a single Action card on your turn. This was a very simple approach I'd already used in other games. It meant I could have cards that let you break that rule, which was a significant attraction. Similarly only getting to buy one card is simple and creates the opportunity to make cards that give +1 Buy.

It's hard to shuffle a very small deck. So, you at least start with 10 cards. I didn't want to make any decisions for you for how your deck was built; so the 10 cards are bad.

Since everything was in the deck, victory cards were in the deck. I made three sizes thinking you'd be able to go for a faster or slower strategy. In the end the game aims heavily towards Provinces in order to keep the game long enough to have fun building your deck.

I needed an answer to the question, what if a pile runs out? At the same time I needed an end condition. So the initial end condition was, the game ends when any pile runs out. Later I changed it to fix a problem (if two players went for Duchies, everyone had to).

A more in-depth history.

You've designed numerous expansions for Dominion. Did you have expansions in mind when you made the base set, and did you have particular goals when you set out to make each expansion?

Initially there were just some cards. There were enough cards that I split it up into a 25-card main set and two 15-card expansions. I moved on to other projects, but my friends just wanted to play Dominion, so I expanded the expansions and made more of them. When I showed the game to Rio Grande Games, I had a 25-card main set and five 20-card expansions.

The expansions have several goals.

  • They try to add something new and exciting to the game, while still being just more cards.
  • They try to play well by themselves. My feeling was that some people buying an expansion would initially play the expansion by itself.
  • They try to have certain game elements at certain frequencies. For example, about 1 in 8 cards gives +2 Actions or otherwise plays extra cards ("villages"). This is essential in order to have it be that, whatever mix of expansions you have, you'll get a decent game experience.

What strategies can designers use to find new design space for an expansion for their game?

An overriding concern is that the expansion be good for the fans of the game. You don't want to veer off too much. People who didn't like the base game aren't trying the expansions, so it's not like, say, a bidding-themed expansion could open up Dominion to bidding fans. The one thing you know about the audience for your expansion is that they liked the base game. So the expansion should shake things up some, but stick pretty close to the main set's premise, especially in terms of what the fun part is.

There are two kinds of expansions: ones that really change the game, and ones that are just more stuff. I have leaned heavily on just having more stuff. Dominion has cards, and you don't use them all each game, so an expansion can seamlessly add more cards. Kingdom Builder has boards and you don't use them all; the expansions add more boards. But other people also make expansions that aren't like that; Cities & Knights is a big change to Catan, not just more of the basic stuff.

I don't know what advice I can give on expansions that really change things. I personally haven't done them, and might lean more towards making a spin-off rather than an expansion there.

For expansions that add more of what you've got, there are several good basic ways to find stuff to do.

  • General functional theme. Anything you can do normally here and there, you can focus on doing a lot of in one expansion. Wizards of the Coast does this a bunch with Magic; any set may have artifacts, and then one set will have lots and lots of artifacts.
  • Program flow theme. This is where there's some basic what-you-do logic that you do a lot of. This is a major source of Dominion expansion themes. Intrigue has "choose one" as a theme; Seaside tries out "now and next turn"; Hinterlands has "when you gain this"; Dark Ages has "when you trash this."
  • Qualifier theme. Cornucopia has a variety theme; several cards care about cards being different from other cards. It's an aspect of things rather than what-you-do.
  • Flavor theme. You can start with flavor instead of functionality. This is a better fit for games that allow more complexity; when you have very simple cards, it's harder to hang meaningful flavor on them. Again, Magic does this one a bunch.
  • Break conventions. In Dominion the convention is that Victory cards don't do anything else, they're just worth points at the end. Intrigue breaks that convention by having Victory cards that do things. A good trick here is to establish conventions, planning to break them later.
  • Add data. Adding data to the game requires components to track the data, but it's otherwise a great, always-available way to find new things to do.
  • Add a value. Alchemy adds Potions as a resource. It gives you a new potential cost for cards.
  • Add rules. This makes your game more complex, and if you add both data and rules, it's like you added a game to your game. Still, sometimes adding a rule will get you somewhere good. Guilds adds both coin tokens and the rule that lets you trade them in.

Attack cards are the most direct form of player interaction in Dominion. Without them, player interactions are much more subtle. Did you seek to encourage certain types of player interaction in Dominion, or did you want to allow the players' preferences to determine this?

Dominion is a game where a lot of the interaction is "incidental," meaning, it comes from the card mix rather than the rules. To make sure there will tend to be a good amount of interaction, I have a certain proportion of interactive cards. It's gone up slightly over the years, though the percent that's attacks has gone down.

There are a few different kinds of interactive cards, not just attacks.

  • Attacks!
  • Cards that give something to other players. Council Room has the other players draw a card.
  • Cards that get data or a decision from other players. Contraband has the player to your left make a decision that limits you.
  • Cards that interact with something the players share. City cares about empty piles, which all players affect.
  • Cards that push competition for the cards. These are more mild but I count them for a fraction. For example, Gardens can cause us to compete over the pile.
  • Cards that feed off of attacks. I don't count these towards my quota since they don't increase the percentage of games with interaction, but they do interact in games with attacks. For example, Moat stops attacks.

I am strictly anti-politics. By "politics," I mean situations in a game where it's possible to talk other players into decisions that benefit both of you to the detriment of other players. You can't completely get rid of politics from interactive multiplayer games (that aren't reduced to 1-2 teams). You can dial it way down though, and that's what I do. So Dominion's attacks and interactive non-attacks all involve everyone else. Militia makes everyone else discard down to 3; you don't pick a player to hose. For me this is just key to enjoying the game. There are players who like to pick who to hose and, well, there are games out there that cater to them.

So: I try to have a good amount of interaction (though it varies with the cards used); I specifically avoid politics; and otherwise it's just down to what I can manage to do on cards.

What can designers do to encourage or discourage certain types of interaction, or to put the level of interaction in players' hands?

My games strive to have variety. Some of them get it in a player-selected way--the varying cards in Dominion, the varying boards in Kingdom Builder. That is all I do to give players control over how much interaction the game has. Anything else is just play style.

I am interested in most ways players can interact. What I do not like, again, is politics.

The main way to reduce politics is to remove the common ground. When I attack you in Risk, every other player benefits. They can try to talk me into attacking you; it's good for them, it's good for me. In Dominion, Witch attacks everyone else; no other player shares in benefiting from me playing it.

Hiding information can also reduce politics. When you are choosing between playing Militia or Adventurer and I already have 3 cards in hand from someone else's Militia, I could try to talk you into the attack, which would hurt at least one other player but not me. But I don't know that you're considering those options until after you've decided.

Simultaneous decisions involve hidden information but further reduce politics. We're all busy during the time I'd spend trying to talk you into a mutually beneficial attack; I don't want to reveal what I'm doing at all via saying what I want you to do.

There are more extreme measures. You can reduce decision-making; reduce interaction period; have all players on one or two teams (a co-op or team game).

Encouraging interaction is not an issue. If the game provides ways to interact, well, if they're mandatory everyone will do them, and if they're optional then the people who enjoy them will do them. Providing ways to interact is important, but people need no encouragement there.

Dominion allows players to choose which card types to include in each session, giving the game a great amount of variability. There are also suggested setups. How did you determine what options to give players?

I tried to keep the set-up simple. The simplest thing is to have one kind of thing that varies, so I try that first. For some games you might be stuck having multiple kinds of things to vary, but this wasn't necessary in Dominion. For Kingdom Builder I vary the scoring separate from the boards.

I try to have games vary as much as possible but not too much. I don't want it to change so much that it's not that game you like, but otherwise I want to really shake things up.

My expectation is that, when there's a set of things to pick from randomly, many people will pick randomly! So that does affect what things are in the set; for example, if something is no good for new players, it might be stuck waiting for an expansion (or never happening). If a combo is too dominating, you might have that combo, so I have to fix it.

The suggested setups, after the first game one, exist because the thought was that some people might like them, as a way to not just jump into pure random. Originally I did not put much work into them; after all, the game is supposed to work with random cards. So I picked cards that went together in whatever ways and they were played once each and that's that. For much later sets, it was clear some people played all of them, and I started playtesting them and tweaking them a little. They are still not all that polished; again the game is supposed to work with random cards, and if it does then these lists will work too.

How much guidance should designers give players in variable setup? Do they need to be wary of an abundance of variety creating an unpredictable experience?

It's a concern in two ways.

First, you want the first game to be as good as possible. If you leave it up to chance, it will sometimes be the very worst case for a first game. So, just specify what portion of the variable setup to use for your first game. Dominion says, play with these 10 cards.

Second, there are all the other games, after the first game. If there's something you need in the game, you need to make sure it's there. For example if it was important to always have a +2 Actions card in Dominion, and I didn't want it to always be the same one, then there would need to be, say, a separate pile of those, so you took nine random regular cards plus one random +2 Actions card. It's nice to avoid having to complicate the set-up like that, but you may have to.

Players can play with whatever cards. If they pick randomly they'll see certain things as often as I like them; for example, a typical game will have a +2 Actions card, but some games there won't be one, for a different experience. But some players may wish to always have +2 Actions, or always have a defense if they have an attack, or whatever, and that's fine. They'll know how much they like it and can stop doing things they don't like. They can always go back to random.

The cards in the game almost all directly help you build up. So it's hard to have a random set of 10 that doesn't give you ways to move forward (without just buying treasures). A big trick here is that smaller effects will have basic resources attached, like the +$2 on Militia.

Cardboard Edison is supported by our patrons on Patreon.

SENIOR INVENTORS: Steven Cole, John du Bois, Richard Durham, Matthew O’Malley, Isaias Vallejo
JUNIOR INVENTORS: Stephen B Davies, Luis Lara, Behrooz Shahriari, Aidan Short, Jay Treat
ASSOCIATES: Robert Booth, Doug Levandowski, Aaron Lim, Nathan Miller, Marcel Perro
APPRENTICES: Kevin Brusky, Kiva Fecteau, Scott Gottreu, JR Honeycutt, Scott Martel Jr., Marcus Ross, Diane Sauer