Prisoner's Dilemma Card

[Glix]

I was tinkering around with the idea of the Prisoner's Dilemma today, and thought it would make an interesting card in practice.  I came up with the following after a little bit of tinkering:

Weaver of Pacts BB
Creature- Human Adviser
Shadow
At the beginning of each player’s upkeep, target player chooses “To Cooperate” or “To Defect” and conceals his or her choice.  Both players simultaneously reveal their choices.  If a player chose “To Cooperate”, each he or she discards a card.  If a player “To Defect”, he or she sacrifices a permanent.  If one player chose “To Cooperate” and the other chose “To Defect”, each player loses 2 life.
2/1

I'm still not sold on the benifits and risk of each proposal, and it might be overpowered as is, but I really like the overall concept.

[Godder]

I like the concept, but it's missing the big point of a Prisoner's Dilemma - someone can't "win" with this card. I'd also only trigger it during "your" upkeep. Finally, Shadow is unnecessary baggage. Maybe this should be an enchantment as well, to avoid comparisons with Dark Confidant.

[Slack]

This has the risk of being a really wordy card.  If you want to stay true to the spirit of the Prisoner's Dilemma, it should be something like this:

Interrogation [cost]

Enchantment

At the beginning of your upkeep, you and target player each secretly choose whether to "cooperate" or "defect" and then reveal your choices.  If both players chose to cooperate then they each lose 1 life and put a 1/1 prisoner token into play.  If both players chose to defect then they both discard one card.  Otherwise, the player that chose to defect draws a card and the player that chose to cooperate discards two cards.

The best scenario for a single player is for them to choose defect and have their opponent choose cooperate.  The point being that if both players are greedy then they will both suffer.  If only one is greedy then he will make out like a bandit.  If they both cooperate then they both get hurt, but less so.

This correlates to the original problem where if they both cooperate they serve 6 months, if they both defect they serve 2 years and if they choose differently one goes free and the other serves 10 years.

This might not be what you're going for, but this is just my thoughts when considering a prisoner's dilemma style card.

Edit: Grammar on the card

[Godder]

"Target opponent", not "target player", but looks good otherwise.

[Anusien]

Is the point of this to make it an iterated game?

By the way, can we bump up/tweak the payments significantly?  The +3 CA one way or another is significant, but other than that I can't see a penalty to defecting every turn; it turns into Necrogen Mists with the potential to go broken.  To be fair, the Nash Equilibrium in a repeated Prisoner's Dilemma is to defect first.

I also lik ethe idea of mixed modes of payment so it becomes more difficult to compare.  What are viable costs?  Life loss, sacrificing permanants, discarding cards?  What about

Player 1 is down, Player 2 is across

	Cooperate	Defect
Cooperate	-2 life/-2 life	+1 card/-2 cards
Defect	-2 cards/+1 card	-2 permanants/-2 permanants

[jro]

I also like the idea of mixed modes of payment so it becomes more difficult to compare.  What are viable costs?  Life loss, sacrificing permanants, discarding cards? 

Even with the same type of payment, the costs aren't really equal.  2 life to a player with only 2 life left is the game; discarding 2 cards to a player with no cards in hand is nothing.  Given the inherent complexity of any card based on the IPD, I'd think keeping the payoff matrix as simple as possible would be desirable.

[Glix]

The point was to avoid the Nash Equilibrium.

I realize my design doesn't conform strictly to the original prisoner's dilemma, however it was difficult to create the IPD I sought, simply because by function, the ability is very black, and as such would only see play in a suicide black style deck.  Therefore, even if both the cooperate/coopearte and defect/defect results are universally negitive, if the more gain/more loss realut, if construed either way, would result in the player ALWAYS chosing the same choice (in the original problem, or the one proposed by Anusien, "To Defect" would always be chosen).  I realized that by placing the dilemma in constraints of Magic, more, less measurable, variables are introduced, at least strategically.

To resolve this, I wanted to force players to attempt to either disagree or agree intentionally, therefore imitating the Prisoner's Dilemma in the overall concept, but better representing the "dilemma" within the constraints of the game.  As such, I reasigned vaules to be, essentially:

                 Cooperate                       Defect
Dooperate   Loss/Loss            Greater Loss/Greater Loss

Defect  Greater Loss/Greater Loss      Loss/Loss

Because of this, a suicide player would usually seek the "disagreement" mode, whereas the opposition would likely chose "agreement".  I compounded this by making the loss for each differet, so the players need to each attempt to anticipate the opponant's most likely choice given board possition.  However, each one will try to think "one step ahead".  Therefore, this card will likely result in a more varied result, and by doing so the Nash equilibrium cannot be reached.

When desigining this card, one must anticipate the deck it will best fit in, which is suicide black (unless I am mistaken, but such an effect rewards aggressive playing, I believe).

That being said, I believe that the card should trigger only durring "your" upkeep.

I am not sure the card is playable, however, without something in addition to a 2/1 body and the above effect.  It just seems unplayable due to its vulnerablitiy.  I suppose it does add excess bagage, but shadow is a one word abiilty that fits very well with the card conceptuatlly, and I felt pushed it to be a bit better.

Being turned into an enchament would make a lot of sense, too.  I'm not really sure which one it should be, as both means seem viable.  Enchantments are a lot less vulnerable, but seem a bit less viable.

Any more input would be appreciated.

[Norm4eva]

I won't pretend I understand the intricacies of what's being discussed here.  However, I would like to submit that there's sort of this unwritten rule about incredibly 'text' cards (there may have been an essay about it on Wizards.com but I don't recall for certain).  If you can't actually fit the concept onto a Magic card, it's probably just a bad idea.  Now this card is clearly very well thought out so it's clearly not a bad IDEA per se, but using the text from Spoils of the Vault and Umezawa's Jitte your card produces unfavorable results:

Spoils of the Vault: (regular sized text)

Name a card. Reveal cards from the top of your library until you reveal the named card, then put that card into your hand. Remove all other cards revealed this way from the game, and you lose 1 life for each of the removed cards.

Umezawa's Jitte: (shrunken text)

Whenever equipped creature deals combat damage, put two charge counters on Umezawa's Jitte.
Remove a charge counter from Umezawa's Jitte: Choose one - Equipped creature gets +2/+2 until end of turn; or target creature gets -1/-1 until end of turn; or you gain 2 life.
Equip 

Prisoner's Dilemma:

At the beginning of each player’s upkeep, target player chooses “To Cooperate” or “To Defect” and conceals his or her choice.  Both players simultaneously reveal their choices.  If a player chose “To Cooperate”, each he or she discards a card.  If a player “To Defect”, he or she sacrifices a permanent.  If one player chose “To Cooperate” and the other chose “To Defect”, each player loses 2 life.

I like to call this the Takklemaggot Dilemma.  Takklemaggot is actually an interesting card; it's supposed to be a parasitic cretin which (very) slowly kills all creatures in play and then attaches itself to the planeswalker with the misfortune of controlling the last guy in play.  However, even the Oracle wording is just an afront to nature, readability and in the end, fun.  High concept cards are excellent and necessary, but sadly they need to have the capacity to be simplified down to a level which the physical constraints of a Magic card can accomodate.  Even though our cards will probably never see actual print, these constraints are still part of the exercise of the CCF.

[parallax]

I don't think it is possible to implement a true Prisoner's Dilemma in a Magic card. The game of Magic is inherently zero-sum. Also both players losing one life or one card seems like a punishment to both players, in reality it is beneficial to one player (the "aggro" or "beatdown" player, generally) by exactly as much as it is detrimental to the other player.

[Nydaeli]

I like the idea a lot, although it's pretty hard to find the right implementation of it.  How does this interact with multiplayer?  It just doesn't make sense with more than two players.

[jro]

I don't think it is possible to implement a true Prisoner's Dilemma in a Magic card. The game of Magic is inherently zero-sum. Also both players losing one life or one card seems like a punishment to both players, in reality it is beneficial to one player (the "aggro" or "beatdown" player, generally) by exactly as much as it is detrimental to the other player.

I wouldn't say "exactly as much", but I largely agree.  I do think it's possible to incorporate non-zero-sum subgames into Magic.  For example, consider the card:

Rainbow Vale 2
Land
Tap: Add one mana of any color to your mana pool.
When ~this~ becomes tapped, target opponent gains control of it.  If they do, untap ~this~.

In theory, all players can gain arbitrarily large amounts of mana, making it a non-zero-sum game.  The game consists of whether or not your opponent decides to tap it for mana and send it back.  I don't think this amounts to a very good game (the best strategy seems like it would always be to just sit on the land until your turn, and possibly forever).  But still, I think this is an area of card design ripe for exploration, even if a card based directly on the IPD isn't feasible.

Here's a possible mechanism for the IPD card:
During each opponent's upkeep, you and that player choose a card in hand and simultaneously reveal them.  A player who revealed a land card draws a card unless another player revealed a non-land card.  A player who revealed a non-land card discards it unless another player reveals a land card, in which case that player draws two cards.

This gives a payoff matrix structured somewhat like the PD (solo defectors are rewarded), works in multiplayer, and is a little shorter.  It's probably still too long, and it depends on players having land and non-land cards in hands to make choices available, so it's still lousy.

Maybe the Prisoner's Dilemma card will just have to go in an Un-set.

[Matt]

Maybe just make it:

Prisoner

Creature -- Human
2/2
, : You and target opponent each reveal a card in his or her hand. If both players reveal a land card, both players draw a card. If both players reveal a nonland card, both players discard a card. Play this ability only any time you could play a sorcery. (The active player chooses first.)

You don't need to specify the cases of mismatches if you just want them to do nothing. Also, recognize that the mere act of using the ability represents a loss for the person controlling the prisoner, so the risk/rewards are not equal for both players.

It's not a prisoner's dilemma but it still has the flavor of cooperating/defecting and can be good times in Multiplayer.