
Quite often, the people who have the most startling impact on games are one-hit wonders -- Gygax and Arneson, for example. Pace both men, but neither produced a game worth the powder to blow it to hell after D&D.
Garfield, by contrast, has certainly never designed another game with the commercial impact of Magic, but his other games, of which this one, are all worth playing.
What's particularly interesting about Dalmuti is the way that it encourages a particular style of socialization above and about the gameplay. That is, if you look at the game as a simple, formal system, it's a mildly interesting Go Fish variant. What makes it interesting beyond that are the roles it assigns to players, the modest game impact those roles have, and the game's encouragement to players to assert their status. Paradoxically, it is a card game that's more fun to play with roleplayers.
It's worth noting in passing that Garfield is the first person to articulate the concept of the metagame (in the sense that game designers now use it), and the appeal of Magic is, to a high degree, the metagame of collection and competitive deck construction it encourages. Here, Garfield is experimenting with the metagame in a different way.
At game start, the four game roles are, in essence, chosen at random: one player is the Great Dalmuti, one the Lesser Dalmuti, and two others the Greater or Lesser Peons. (If there are others, they are called Merchants.) The cards are all dealt out, at random, and the Great Dalmuti leads. Before he does so, taxation occurs; the Great Dalmuti gives the Great Peon any two cards he wishes, and the Great Peon must give the Great Dalmuti his two best (lowest numbered) cards. Similarly, the Lesser Dalmuti taxes the Lesser Peon for one card.
The deck consists of cards ranked from 1 to 12, with the rank also corresponding to the number in the deck -- that is, there are twelve 12 cards, one 1 card, and so on. There are also two "Jesters," which serve as wild cards. The leader plays any number of cards of the same rank. Play then goes around the table; each person may play cards of a better (that is lower) rank, but must play exactly the same number of them as played by the leader. If he cannot, or chooses not to play, he passes. Once all players pass, the last player to play cards leads the next round.
The objective is to get rid of all your cards first. The player who does so is the Great Dalmuti for the next round of play, the second player to do so is the Lesser Dalmuti, and so on.
The Great Dalmuti is thus a game with a strong first-mover advantage, normally considered a flaw in a game design. Here, however, the "flaw" is transformed into a central part of the game's appeal; combined with taxation, it gives the Great Dalmuti a commanding advantage in play which, particularly if he roleplays the part of an arrogant noble indifferent to the suffering of otehrs, can irritate the other players in a fruitful way. When it happens that cardplay shifts the order of hierarchy, this becomes a highly enjoyable moment.
Dalmuti is not a game of great strategic depth; you rarely hesitate over what to play, though there is some modest advantage to card-counting. But if you view it as a beer-and-pretzels game with a bit of roleplaying appeal, it's not bad at all.

















The Peon Shuffle
My friends play with several additional house rules that have no effect on the actual play rules. My favorite of these is that whoever's stuck with the Leser Peon role after a round has to sweep up all the cards, shuffle, and deal them. (Unless the Dalmuti enjoys dealing, in which case the peon must pass the shuffled deck to them, bowing and scraping appropriately.)
How novel is Garfield's variation on this traditional game?
Isn't the Great Dalmuti merely a "cleaned-up" variation of a popular card-game that pre-dated it? According to Wikipedia's article, it's a variant of the traditional game President.
What did Garfield bring to the table other than the variation in card distribution? Wasn't the meta-gaming construction always in the game? Or did he formalize it with his taxation mechanics? (My understanding of President is that it at least incorporates the meta-gaming role-based bits of forcing the person at the bottom of the totem pole to perform menial chores such as card raking and shuffling, etc).
related trivia
There used to be one bit of lore passed around the Wizards of the Coast center in the University District (in Seattle) about the game--allegedly, the word "dalmuti" actually means "asshole" in some uncommon language. The idea, at least, amused workers there greatly, as the attached restaurant was called "Dalmuti's".
There's not much that's
There's not much that's original in The Great Dalmuti, apart from the custom deck. As others have pointed out, there is a large pre-existing family of card games with the same mechanisms. (My friends and I started playing it under the name "Tahimi" back around 1992 or so.)
There's also a Dilbert-themed version of The Great Dalmuti called "Corporate Shuffle" that's kind of entertaining due to the corporate role names (Custmer Support, Marketing, Engineering, etc.) and Dilbert comics attached to the different card ranks.
Emperor and Scumbag
My friends used to play this as "Emperor and Scumbags" and we used to get into the metagame a lot. The high ranking players got the comfortable chairs while the scum sat on the floor and had to do favors for the emperors.
We added a nice house rule for 'Assassination' in which the supreme Emperor had to make sure he always finished first or else he fell all the way to scumbag. This made holding onto the the top position riskier and competition between the emperors was a nervous thing. Some players preferred not to go for the top position, but rather sit more comfortably in second or third place.
To the other extreme...
...I once went to an SF convention where we kept a game of Dalmuti going non-stop for the whole weekend (I think we lasted 56 hours in the end.) One of the great things about the game is that it can cope with people coming and going very easily and will still work even when you are reduced to a rump of three (in the dark reaches of the night...) or when there are a ludicrous number (13 is just too many I think.)
As a result I am very, very good at the game - but never want to play it again :)) It has more strategic depth than is immediately apparent, although I grant you it's still not *that* deep.
And yes, it's an adaptation of an existing game, but the card distribution is the thing that transforms it and elevates it to greatness.