How many Dominion cards do you think you have come up with, including different variations on the same card that were at least considered and maybe tested? Do you have a spreadsheet to keep track of them all, so you can record things like "reaction that hurts attacker | bad idea" and ideas for possible future promo cards, etc.?
There is an old ideas file, which has lists of ideas sorted by card type. I guess there's an even older file that I turned into this file. Ideas are sometimes marked with a rating, + for good, . for okay, - for bad; this doesn't reflect testing, just, how much do I like this idea. Some things have a comment in brackets after them, sometimes reflecting testing. Let's peek at the first five things on the discard-attacks list.
. att: each other player discards silver
. att: name treasure. each other player discards it [strong at 5]
. reveal top. if not silver, each other player ebbs. gain top silver. [multiplayer cumulativeness]
. if another pl. has < 5 cards, do x. otherwise, they discard.
each pl. looks at left's hand if they have 5+ cards, chooses a discard
As you can see I tested one of those, although I don't think I printed out anything for it. "Ebb" means "put from hand onto deck" - after the Magic card Time Ebb. "Do X" is of course a placeholder to just show off the actual idea; similarly most cards would also make +$2 or something; that isn't the idea part.
There are just 21 things on that particular list, plus a list of general approaches at the top. The file is 58K and also includes lists of general mechanics. These lists have been combed over; there are probably a few things that would be okay in there, but you know, the discard-attacks list, those are the 21 variants I passed over in favor of better ideas.
Then each set has a file, with ideas specifically for that set, and a list of the set as it stands at the top, with some notes on what cards fill what roles. For example for Dark Ages, the original list of ruins ideas is:
junk possibilities
- +1 action / +1 buy / +1 card / +$1 (ruined village / market / library / abandoned mine)
- look at top 2, may discard them (survivors)
- pass this left
- gain a copper / gain a card costing up to $2 / gain a card costing up to $1
- blank / trash this
- draw up to 4
- +1 card, -1 card
- action cards cost $1 less this turn
- worth 1 vp per 5 ruins in your deck
In that list the minuses are just for indenting, not passing judgment. Later I considered a few other things, including "play up to two ruins cards" for Ruined Village, but they aren't on this list. The initial five worked out so there wasn't a lot more work there.
The Dark Ages cards file is 59K (distinct from the file with the secret history and such). It is just an endless sprawl of card ideas, with some to-do list items like "fit in a 3rd spoils card."
Finally there are the image files. Dark Ages has 30 pages of card images to test (9 cards per page), plus full versions of the sets that sometimes include cards not on the other pages, especially the older ones when I wasn't saving everything yet. It is hard to meaningfully count those pages; it's 36 pages, not counting the original 3 page version or the brief 2 page version, but most of that stuff is redundant. And of course most of the images are things in the Dark Ages text file.
I am looking at the 4th page of Cornucopia images, which was the first page without a version of a card from Guilds. It has:
- Three versions of Horn of Plenty, two of them actions.
- Two version of Wandering Minstrel that may appear unrelated to it (it started out +$2, name a type, dig for one and leave it on top).
- Two random cards that didn't make it - "gain 5 silvers minus a silver per card in hand" and a thing that made other players ebb a card if they had any duplicates in hand.
- A precursor of Harvest that drew the non-duplicates in your top 5.
- Horse Traders but called Foreign Traders.
I flirted with posting the image, but man let's save that stuff for after Guilds is out, not have any slip-ups.
Not every version of every card makes it anywhere though. I say, "this game, this card will be different," and explain what I want to test and we test it. Maybe it works out and gets an image and more testing; maybe I'm immediately done with it. Maybe it seems promising and I change the cost and we try that, but the original cost is never in a file. You know.
I do not know how many cards ideas I've had, but there's some of the data I'd be looking at to guess that number. For a normal game I make maybe twice as many cards as end up in the game (distinct cards, rather than slight variations or wordings fixes or what have you). Like, for Nefarious, there were about twice as many twists tested as were used. I pared it down a couple times. There would also be a list of twist ideas I didn't try, however many, I'm not checking. For Dominion the numbers are higher; some cards just have one version, but some have 10 versions, and for any idea there are probably lots of ways you could do it that would be fine, and you can list them and consider them, but once you commit to one then the others aren't so interesting anymore.