Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Titandrake]

Disclaimer: I in no way claim I am a professional designer, or even an average one.

Seems like there are multiple people who are considering making custom sets. This is a list of various pointers to help with the designing process.

1. The set needs a theme.

So far, the official Dominion expansions have had the following mechanical themes.

Intrigue: Choices
Seaside: Durations
Alchemy: Potion costs
Prosperity: VP tokens
Cornucopia: Rewarding variety
Hinterlands: Effect on gaining/buying

I'm ignoring Base because it was the first set. You could argue its theme is simplicity.

A mechanical theme keeps the set together, and makes sure each set is distinguishable from the next. This also has the effect of allowing the players to choose the gameplay they want, by selecting sets with effects that they want to play with.

Sets should have a flavor theme as well. Dominion is not a purely mechanical game; a card like Witch or Thief would initially feel different if they were named something else. However, flavor tends to be easier to add, as the name of the card can always be adjusted, and to be honest not all cards have names that make sense.

2. Do not forget bread and butter effects.
If you have not read the stickied Fan Card Creation Guide, you should. It does a good job of describing what a card should and should not do.

What it does not address is how cards interact with each other. For instance, say you have a batch of 20 well-designed, balanced cards. You decide to try playing with these cards. And the experience is awful. Why? Because although all your cards are good, you have no Village effects. Or you have no good card draw. Or you are missing some other general effect that is part of what makes Dominion play the way it is.

As far as I remember (and you can correct me if this is wrong), all Dominion sets have the following cards:

Cards with +2 Actions or more.
Actions that are terminal Silvers with a bonus.
A card that gives out curses.
A card that gives terminal card draw.
A card that synergizes with itself (i.e. a one card engine.)
A card that trims the deck.
A card that gives +Buy.

Some of these effects need to show up more often than others (you don't 10 trashers in one set, and you need more than 1 +buy card in a set). But all of these effects should show up at least once.

3. Playtest the set. With other people. A lot.
It seems that the one constant about card design is that without playtesting, we are terrible at judging power levels. Don't get me wrong, it doesn't take a genius to know that a $2 cost Action that gives +5 Cards, +3 Actions is horribly broken. But it's difficult to see how an idea plays out without actually playing it.

I had other stuff to say, but I seem to have forgotten what it was. I'll edit it in if I remember it, I think I typed enough of a text wall already.

[Elyv]

Minor nitpick: Prosperity's main theme was definitely not victory tokens; there were a total of 3 cards that use them. I feel like its main themes are treasures and cards that are like other cards, but bigger(colony<=>province, grand market<=>market, KC<=>Throne Room, etc)

Definitely a good read overall, though.

[Diving Pikachu]

Donald has written in his Secret History of the Alchemy Cards that "'has a potion in its cost' is not actually a huge connection functionally, so [he] supplemented the main theme with a 'cares about actions' sub-theme." And Seaside has a much more pervasive "next turn" theme, even though not all its cards strictly use the duration mechanic. I also read that from Donald's secret history of the set.

In fact, I'd feel that any discussion on set design would be incomplete without the invaluable design notes that are Donald's Secret Histories, which Theory has conveniently posted in the subforum called The Bible of Donald X.

[ChaosRed]

Nice post.

I wanted the underline your advice about play testing. Play test, a lot, as much as you can. It reveals a lot about your cards and how they work with one another, how valuable they are based on what's on the board and where you are at with the game. Pay special attention to cards that seem to sit dead on the board, in almost every game, similarly, take a hard look at cards that are depleted almost every single game they are used.

Nothing, and I mean nothing, beats play testing for helping you refine and balance your expansion.

[Titandrake]

Thanks for the comments.

Yes, you should read the secret histories too. I'll admit I haven't read them recently, so I'm not exactly up to date.

Prosperity is an odd set. The idea of "expensive but better" is much closer to a flavor theme than a mechanical one, but when compared to VP tokens, I can't argue that it is not the dominant one. It shows up a bit mechanically when you look at Bishop and Peddler as well. If I recall the Secret History talks a bit about Peddler and variable costs.

A lot of sets have these interesting subthemes, sometimes coming from themes that don't fill an entire set.

Also, one last thing. This normally comes with themes, but a set should have synergy with itself. Examples: Workshop/Gardens, Ironworks/Great Hall, Minion/Conspirator, Vault/Grand Market, Embassy/Tunnel, University/Vineyard. This should come about naturally, either from your theme or just because it happens, but it's worth keeping in mind...

[rinkworks]

As far as I remember (and you can correct me if this is wrong), all Dominion sets have the following cards:

Cards with +2 Actions or more.
Actions that are terminal Silvers with a bonus.
A card that gives out curses.
A card that gives terminal card draw.
A card that synergizes with itself (i.e. a one card engine.)
A card that trims the deck.
A card that gives +Buy.

This got me curious to examine the frequency with which these broad card categories appear in the official sets.

Terminal Silvers

These are terminal actions that offer a flat +$2 bonus plus something extra.  I've excluded attacks, since those fulfill a different purpose.

Base: 2
Intrigue: 1 (counting Steward but not Conspirator)
Seaside: 3
Alchemy: 0
Prosperity: 1
Cornucopia: 0
Hinterlands: 4
Average: 1.83 per full set

Attacks

Base: 5
Intrigue: 4
Seaside: 5
Alchemy: 2
Prosperity: 3
Cornucopia: 3 (not counting Followers)
Hinterlands: 3 (not counting Ill-Gotten Gains)
Average: 4.17 per full set

Cursers

These include non-attacks (Ill-Gotten Gains) and cards that only give out curses sometimes (Jester, Torturer).

Base: 1
Intrigue: 2
Seaside: 2
Alchemy: 1
Prosperity: 1
Cornucopia: 2 (not counting Followers)
Hinterlands: 1
Average: 1.67 per full set

Terminal Drawers

Included are terminals that provide at least +2 Cards or up to 6.  I wasn't going to include attacks, but sometimes (unlike with terminal Silvers) attacking drawers are used as much for the drawing as the attacking.  So two separate numbers are provided here, one for non-attacks and one that includes attacks.

Base: 4 / 5
Intrigue: 2 / 3 (not counting Tribute)
Seaside: 1 / 2
Alchemy: 0
Prosperity: 2 / 3
Cornucopia: 0 / 0 (counting neither Young Witch nor Followers)
Hinterlands: 1 / 3
Average: 1.67 / 2.67 per full set

Self-Synergizing Cards

The definition for this is very inexact, but I decided to count any card that matches one of the following:

- Non-terminal or treasure that reliably grows hand-size (e.g., Laboratory but not Wishing Well).
- Non-terminal or treasure that draws at least one card and yields at least +$1 (e.g., Market) or lowers card costs (e.g., Highway) with each play.
- Non-terminal or treasure gainer that also enables the gained cards (e.g., University but not Horn of Plenty; also not Baron, since it's a terminal).
- Non-terminal or treasure that produces extra effect in multiples (e.g., Treasure Map, Fool's Gold).
- Card that offers choices (with at least one non-terminal choice) that, combined together, effectively accomplish one of the above (e.g., Nobles but not Pawn).

Base: 2
Intrigue: 3
Seaside: 4
Alchemy: 6 (includes Golem, as it effectively increases hand-size)
Prosperity: 3
Cornucopia: 3 (not counting Trusty Steed)
Hinterlands: 5
Average: 4.33 per full set

Trashers

These do not include attacking trashers (e.g., Saboteur), only cards that trash from your own deck.  The first figure excludes trash-for-benefit cards (e.g., Remodel, Bishop) and the second figure includes trash-for-benefit cards.  Self-trashers (e.g., Mining Village) are not included.  I decided to include Island in this category, since it's still a deck-slimmer, but I left Native Village out, since cards on the Native Village mat are usually subsequently returned to the deck.

Base: 1 / 4
Intrigue: 2 / 4
Seaside: 2 / 3
Alchemy: 0 / 2
Prosperity: 2 / 6
Cornucopia: 0 / 1
Hinterlands: 1 / 3
Average: 1.33 / 3.83 per full set

+Buy

Included cards don't necessary provide +Buy every time they are played.  Gainers are not included.

Base: 4
Intrigue: 3
Seaside: 2
Alchemy: 1
Prosperity: 6
Cornucopia: 1 (not counting Princess)
Hinterlands: 3
Average: 3.33 per full set

Gainers

These are cards that allow a card to be gained outright.  A well-designed set probably has this category balanced with the +Buy category, although they do fulfill a slightly different niche.  The Remodel family is not included (I've counted them among the trash-for-benefit category, above), and nor are one-shots (e.g., Feast, Treasure Map), since those fulfill a different purpose.  I've also excluded attacks (e.g., Thief, Jester), which may not have been the right call.

The first number excludes gainers that can only gain a specific card (e.g., Bureaucrat, Baron, and I've included Explorer among these too); the second number includes these.

Base: 1 / 2
Intrigue: 1 / 2
Seaside: 1 / 2
Alchemy: 1 / 1
Prosperity: 2 / 3
Cornucopia: 1 / 1 (not counting Followers)
Hinterlands: 2 / 3 (not counting Cache)
Average: 1.5 / 2.33 per full set

Villages

Cards that offer at least +2 Actions, though not necessarily every time they're played.

Base: 2
Intrigue: 3 (not counting Tribute, due to lack of control)
Seaside: 3
Alchemy: 1
Prosperity: 2
Cornucopia: 2 (not counting Trusty Steed)
Hinterlands: 3
Average: 2.67 per full set

Sifters

These are cards that improve either your hand or the top of your deck as their primary effect.  They can do other things, so long as this is their main purpose.  Thus, I'm including Spy but not Young Witch, though obviously there is lots of gray area here.  Drawers like Laboratory are not included unless they also sift (e.g., Oracle, Farming Village).  I've excluded returns that put themselves on top of your deck (e.g., Walled Village, Treasury) but included cards that put other cards on top (e.g., Royal Seal, Scheme).

Base: 3
Intrigue: 2
Seaside: 4
Alchemy: 2
Prosperity: 2
Cornucopia: 1
Hinterlands: 8
Average: 3.67 per full set

Cantrips

For strategy discussion, a broader definition of "cantrip," one that would include Villages and Laboratories, seems to have the most practical application.  But we've already accounted for both of these in categories above.  So here, the definition of "cantrip" I'm using is:  A card that gives exactly +1 Card and exactly +1 Action at least some of the time.  That means Conspirator and Wishing Well are included, but Laboratory and Village are not.  Somewhat arbitrary, I'm allowing mandatory discards, but only after the draw; thus, Oasis counts but Cellar doesn't.

Base: 2
Intrigue: 5
Seaside: 4
Alchemy: 2
Prosperity: 2
Cornucopia: 2 (not counting Hunting Party)
Hinterlands: 4
Average: 3.5 per full set

Non-Terminals

One of the more fundamental balances is non-terminals vs. terminals.  Cards that are always non-terminals count.  Cards that are sometimes non-terminal (e.g., Conspirator, Pawn) count half.  All types of non-terminals are included: cantrips, Villages, Laboratories, Attacks, etc.  I'm counting Throne Room and King's Court as half, since they can be doubled-up to allow extra actions.

Base: 6.5
Intrigue: 9.5
Seaside: 10
Alchemy: 6.5 (Potion costs are more suited to non-terminals; thus, this is a bit of an outlier)
Prosperity: 4.5 (unusual number of Treasures make this number a bit smaller than it would be)
Cornucopia: 5 (not counting prizes)
Hinterlands: 7.5
Average: 8.25 per full set (8.56 if Alchemy and Prosperity are excluded)

Card Costs

It's also important to balance card costs.  These are the number of cards priced at $2 / $3 / $4 / $5 / $6.

Base: 3 / 4 / 10 / 7 / 1
Intrigue: 3 / 6 / 7 / 8 / 2
Seaside: 5 / 5 / 8 / 8 / 0
Alchemy: (Potion costs disqualify this one as an outlier)
Prosperity: (Colony game design disqualifies this one as an outlier)
Cornucopia: 1 / 2 / 5 / 4 / 1
Hinterlands: 3 / 5 / 6 / 10 / 2

Average $2 Cards: 3.33 per full set
Average $3 Cards: 4.88 per full set
Average $4 Cards: 8.0 per full set
Average $5 Cards: 8.22 per full set
Average $6 Cards: 1.33 per full set

If you're making a Potion-based or Colony-based set, we only have one data point apiece, but here they are:

Colony Set: $2 Cards: 0
Colony Set: $3 Cards: 3
Colony Set: $4 Cards: 5
Colony Set: $5 Cards: 9
Colony Set: $6 Cards: 3 (one with a buy restriction)
Colony Set: $7 Cards: 4
Colony Set: $8* Cards: 1

Potion Set: $2 Cards: 1
Potion Set: $5 Cards: 1
Potion Set: $P Cards: 2
Potion Set: $2P Cards: 3
Potion Set: $3P Cards: 3
Potion Set: $4P Cards: 1
Potion Set: $6P Cards: 1

[ChaosRed]

More notes from someone now 80 games into my own expansion's testing:

Pay attention to your cost distribution, just as rink says.

One of things I did "right" was I mirrored the cost distribution of base set exactly and it turned out to be a smart move. My testing has generally produced a nice array of costs/features for each random board and when combined with base set, the ratios stay the same, so the set still works in this regard.

You don't need to mirror an expansion exactly for cost, but the guidelines of previous expansions do provide a nice glimpse at how to keep your set balanced.

I also urge you to make a few 2$ cards. A lot of variants I see are 5$ or higher, but designing a cool, nifty 2$ card is very entertaining and satisfying.

I also want to warn you about attacks. They are very hard to balance correctly. Although they are fun to design, they require a lot of testing. Also attacks are best tested against live opponents. Testing attacks against a BM+Smithy will generally skew your results. Your opponents need to be able to counter-attack, or perhaps come up with a defense or find ways to mitigate the penalties of your attacks. If you really want to get a true handle on the validity of your attack card, play them against a sentient opponent, who can react or counter to the attack.

I have a lot more notes on this, but I'll hold back, as I'm still learning myself and I've made a lot of mistakes along the way.

[WanderingWinder]

I want to add that I don't think the specific 3 vs 4 split is SO important, but the total combination of 3s and 4s is. i.e., if there were 7 3s and 6 4s, man I don't think that's huge problem, though there is SOME reason you get more 4s (mostly 'cause 4s can be more interesting), but if you have twice as many 5s as 3s+4s or if you have as much as many 2s as 3s+4s, than that's going to be a big problem.

On the other hand you COULD break these patterns for the sake of the theme of your set (say you want an anti-prosperity, where small is the way to go), but you want to be VERY careful about doing this, 'cause otherwise you can end up really warping the game (i.e. in your 'cheap stuff rules' set, in order to make it powerful enough, you want to give some pretty good incentives to go small, which probably means making it fast, but on the other hand, if it's too fast, then cards from other sets aren't going to get played over them in game with them, and you've just made a bunch of game-warping super-fast cards). So I think these are good guidelines, and you can break them, but you need to be really sure/have a really good reason. And then, as is always the rule of the day, PLAYTEST.

[ChaosRed]

Yeah agreed that small variance isn't a big deal, but as you say, you should probably avoid tilting too far in one direction.

I really feel, the base set is an elegant design, on a lot of levels, and is a superb place to start to appreciate design. You don't have to mirror it exactly of course, but take time to appreciate how and why it is so elegant. We've all played the base set so much now, we take it for granted I think. But there was a lot of thought and care put into the base set. In its simplicity are some of the seeds to producing a well-crafted expansion.

[Titandrake]

The base set is nice because of how it has every element it needs in its simplest form. Donald X already has a post on complexity creep, but it's worth mentioning. You want to keep the complexity as low as possible. If you must have a complex effect, then it should ideally be tied into the card's flavor, because it makes the effect easier to keep track of.

And yes, cost distribution is important. You know, not really sure what gives me the right to talk about all this stuff when I have barely playtested my set at all

[rinkworks]

I added two more categories to my post above:  cantrips and non-terminals.

[chwhite]

Cursers

These include non-attacks (Ill-Gotten Gains) and cards that only give out curses sometimes (Jester, Torturer).

Base: 1
Intrigue: 1
Seaside: 2
Alchemy: 1
Prosperity: 1
Cornucopia: 2 (not counting Followers)
Hinterlands: 1
Average: 1.5 per full set

Intrigue has two Cursers (Swindler and Torturer); I think Seaside only has one? (Sea Hag)

Also, if you're excluding Prosperity from the discussion of card costs (because Colony=more expensive stuff), then I think you have to exclude Seaside as well.  Seaside is nearly as much an outlier, with lots of $2 cards and no $6 cards, and was specifically designed that way to balance out the high costs of Prosperity.

[Thisisnotasmile]

Intrigue has two Cursers (Swindler and Torturer); I think Seaside only has one? (Sea Hag)

Ambassador.

[AJD]

Intrigue has two Cursers (Swindler and Torturer); I think Seaside only has one? (Sea Hag)

Ambassador.

Embargo?

[chwhite]

Intrigue has two Cursers (Swindler and Torturer); I think Seaside only has one? (Sea Hag)

Ambassador.

Embargo?

Interesting- would Embargo count?  You can't force another person to take the curse.

If we're including Ambassador as a curse-giver, then Masquerade should probably count as well.

[AJD]

Intrigue has two Cursers (Swindler and Torturer); I think Seaside only has one? (Sea Hag)

Ambassador.

Embargo?

Interesting- would Embargo count?  You can't force another person to take the curse.

I mean, you can't force another person to take the curse with Torturer either. But obviously Embargo is a really oddball card; the "thing it does" in some sense is hand out curses, so in that respect it belongs in the same group as Witch and Ill-Gotten Gains; but it's clearly not very much like those, or even Torturer, and really belongs in a class by itself.

[dondon151]

I wouldn't personally consider Ambassador and Masquerade because 99.9% of the time, when they do hand out a Curse, there needs to be another Curse-giving card in the kingdom.

[Graystripe77]

I wouldn't personally consider Ambassador and Masquerade because 99.9% of the time, when they do hand out a Curse, there needs to be another Curse-giving card in the kingdom.

Not true, you could buy a curse to ambass it.

[rinkworks]

Intrigue has two Cursers (Swindler and Torturer); I think Seaside only has one? (Sea Hag)

I was counting Embargo.  Swindler was an oversight; I fixed the original post.

Also, if you're excluding Prosperity from the discussion of card costs (because Colony=more expensive stuff), then I think you have to exclude Seaside as well.  Seaside is nearly as much an outlier, with lots of $2 cards and no $6 cards, and was specifically designed that way to balance out the high costs of Prosperity.

Do you have a reference for this?  Because unless Donald said as much, I don't see how this can be true.  The expansions are designed to be balanced even if all you play with is one expansion plus either Base or Intrigue.  Many people will play Seaside and never play Prosperity, and, in fact, everyone did this for a full year after Seaside was released.  For that matter, Donald only ever mixes two sets at a time, so I understand; I'm sure he doesn't only ever play Seaside and Prosperity with each other.

Prosperity's costs are higher because of Platinums and Colonies and thus need not be balanced back the other way.  Platinums and Colonies *are* the balance.  In mixed games with enough Prosperity cards, you tend to play with Platinums and Colonies, and when there aren't many Prosperity cards, you don't need to counterbalance their costs.

Certainly Seaside's cost distribution is an outlier, but I see it as an outlier due to random noise within the common parameters that govern all the sets.  Prosperity's cost distribution, by contrast, is clearly by design.  Thus, if you're designing a Province-game custom set, you're probably okay if you model it after Seaside's cost distribution but not necessarily if you model it after Prosperity's.

[lympi]

Also, if you're excluding Prosperity from the discussion of card costs (because Colony=more expensive stuff), then I think you have to exclude Seaside as well.  Seaside is nearly as much an outlier, with lots of $2 cards and no $6 cards, and was specifically designed that way to balance out the high costs of Prosperity.

Do you have a reference for this?  Because unless Donald said as much, I don't see how this can be true.

From The Secret History of Prosperity (under the Origins 2007 section):

$2; Wishing Well. The Intrigue card, only for $2. Prosperity never had much in the $2 department. If you want to throw in expensive cards, something has to give. In the end I felt that it would fit the flavor to not have any $2's at all. I put extra $2's in Seaside to compensate, at least for people who play with everything.

So Seaside's lower cost average is sort of by design, sort of not. I mean, it is by design, just not by design of Seaside… by way of Prosperity.

[sherwinpr]

Attacks

Base: 5
Intrigue: 4
Seaside: 4
Alchemy: 2
Prosperity: 3
Cornucopia: 3 (not counting Followers)
Hinterlands: 3 (not counting Ill-Gotten Gains)
Average: 4.0 per full set

Seaside has 5 attacks: Ambassador, Cutpurse, Ghost Ship, Pirate Ship, and Ghost Ship Sea Hag.  This should bump the average to 4.17 per full set (if I did the math correctly).

Edit:  Thanks to Jimmmmm for catching my mistake.  It should be a Sea Hag rather than a second Ghost Ship, obviously.

[Jimmmmm]

Ghost Ship, Pirate Ship, and Ghost Ship.

Umm...

[AJD]

Ghost Ship, Pirate Ship, and Ghost Ship.

Umm...

...Also Sea Hag, and Ghost Ship.

[sherwinpr]

Ghost Ship, Pirate Ship, and Ghost Ship.

Umm...

...Also Sea Hag, and Ghost Ship.

Hahaha, okay so I messed up.   

But my point still stands.  Replace the second Ghost Ship with Sea Hag (I was even going in alphabetical order until the error).

[rinkworks]

Attacks

Seaside: 4

Seaside has 5 attacks: Ambassador, Cutpurse, Ghost Ship, Pirate Ship, and Ghost Ship Sea Hag.  This should bump the average to 4.17 per full set (if I did the math correctly).

Indeed you did.  I fixed the original post -- thanks!