Experiments with the Decktet

I was having a think and was wondering about a game that played a little like Coloretto, with rows of cards being built up and claimed before being put into rummy-style melds for scoring, and where, unlike Coloretto, the game keeps flowing as players claim cards, so the rows are constantly shifting and changing.

This developed into a thought about the Decktet which, in case you didn't know, is a beautifully illustrated deck of playing cards with six suits, but with the twist that most of the cards have more than one suit.  So, for example, you can have the "7 of Suns and Knots" or the "3 of Leaves and Wyrms".  There are heaps of games already designed using this deck; S and I often play "Bharg", which is a version of gin rummy, adapted for the Decktet.  Anyway, the idea of multiple suits combined with the previous thought got me to work...

Pretty Decktet cards, hard at work being experimented on.

So I figured that the main flow of the game would be that a player would add two cards from the deck to central rows of cards, with the rule that each row had to have one suit in common for all its cards, and that you had to add a card to a row instead of starting a new one if possible.  After placing two cards, you take all the cards from one row to add to your collection.  We would seed the table by dealing four starter cards.  Scoring at the end was uncertain.  I initially figured that you could arrange your cards in sets sharing a suit, and use Fibonacci numbers (1, 2, 3, 5, 8, 13, 21...) to score the sets.

S agreed to have a go at this with me.

Within a few minutes we agreed that this was not working.  We were running out of cards in the middle, and it felt like the only decision was to always take the largest amount of cards possible.  Of course, if we could come up with a more subtle scoring system (like Coloretto's capping and penalties for too many suits), this could make things more interesting.  But even so, the basic game play just wasn't working.

After abandoning that play, S came up with a couple of suggestions, which we discussed and then restarted.  This time, the plan was that instead of drawing two cards, you drew cards one at a time, adding them to rows if possible, but not adding more than one card to any given row in a single turn.  When a card was drawn that couldn't be added to a row, it started a new row and the player claimed a row.  In this way, every turn involved adding one row and removing another.

Within a couple of turns we could tell this was a far better game.  The flow of play was good and it felt like there were actual decisions to make at each turn of a new card.  The problem came at the end when we scored up: Fibonacci numbers work great for a while, but S had a 12-long suit and I had a 10, which left her overall score being twice mine.

So, overall, if I can come up with a good scoring system, this could be a workable game.  I'll spend some thought on this.  It also occurred to me that instead of having 6 suits with 12 cards in each, 12 suits with 6 cards in each might work better for this structure.  I'd have to generate some cards to try that out, but that would be something that nanDECK is very good at, so maybe I'll give that a try.