The Naga Demon Dilemma

For about a week now, I have had tabs in my internet browser open for Wikipedia articles including "Prisoner's dilemma," "Unscrupulous diner's dilemma," "Peace war game," and "Nash equilibrium." Why? Because of NaGa DeMon - National Game Design Month. NaGa DeMon is where you create a game along with all the necessary pieces and play it at least once all within the month of November. This month is going to be the month where I design my very first original game.

Yes, for those participating in NaGa DeMon like I am, I know that research could (preferably) have been done before November started. During the latter half of October, however, I was focusing my efforts on building a Magic: The Gathering cube and designing Magic cards for the set pitches contest for the Goblin Artisans 2012 set project. But excuses, excuses; right? Anyway...

So, a week has gone by, and the percentage of my game I have completed by now amounts to about diddly-squat. This is what I've decided on so far: it's a card game, and the mechanics will have elements of the prisoner's dilemma. Why a card game? Because, like The Joker, I adore card games. And they're easier to make than a lot of other games. Why the prisoner's dilemma? Because I've been fascinated with it, especially since when The Legend of Zelda: Four Swords implemented it.

The prisoner's dilemma is a situation where two players are given the same set of choices that they will each make secretly from each other: cooperate or betray. There are four outcomes with two players with varying degrees of rewards. The best thing that can happen is one player betrays while the other cooperates. The worst thing is when you get betrayed while you cooperated. So why would you ever cooperate? Because the second-to-best thing is when you both cooperate. The second-to-worst thing is when you both betray each other. If you both keep betraying each other, you're both getting some bad rewards. So, then you both cooperate to get better rewards. But then somebody can get selfish and reap the best rewards and betray. Dilemma!

The Legend of Zelda: Four Swords implemented this by making collecting rupees a competition. The players must both cooperate to traverse the dungeon, defeat the monsters, and solve the puzzles. At the same time, there are treasure chests and rupees lying around alongside these important and dangerous obstacles that require teamwork. Players have the ability to mess with the other players, even being able to go so far as to throw other players into a bottomless pit.

So, let's say you just enter a room with this boss monster and some treasure chests lying around. The boss monster starts attacking the players, threatening the welfare of everyone. Three of the four players immediately start working together to fight the monster. The fourth player, however, decided that the other players will probably fight the monster, and three would be more than enough to keep the monster distracted. So, the fourth player sneaks over to the treasure chests and starts collecting rupees, helping them gain the lead in the rupee competition. The rest of the players are doing all the work while you're making out like a bandit!

In another scenario, EVERYONE gets this mindset of desiring to get to the treasure chests first before dealing with the monster. So all four players starts heading over there - but nobody is taking care of the monster. Then the monster attacks and hurts everybody. Sad times for all. But, hey, you might have been able to sneak one treasure chest or something. ...But if you ALL WORKED TOGETHER and beat the monster first, you could have still gotten one treasure chest or something but without all that damage you endured just now.

However, Four Swords doesn't just work like the prisoner's dilemma - it works like an iterated prisoner's dilemma. This is different in that there are multiple times where you must make the decision to cooperate or betray. And you have memory of all the outcomes and decisions made by the players from previous prisoner's dilemma situations. Knowing what has happened before can alter your decision in the next prisoner's dilemma. This is great. But then there's a problem.

The problem with the iterated prisoner's dilemma is that when there is a known finite number of times the prisoner's dilemma will be played out, this will usually influence the behavior of the players to betray during the last known round of the prisoner's dilemma - because there will be no consequence to their actions afterward. But since you know that betrayal will happen during the last round, you might as well gain the upper hand and betray during the second-to-last round. And then it goes on and so forth like this until you've determined that the best decision is to relentlessly betray right from the beginning.

So, what is needed is an unknown number of rounds of the prisoner's dilemma. How do you do that? Well, a deck of cards is great for randomizing things in an order that you do not know. However, simply using a deck of cards with events in it with a "game-ending" event inside of it won't be enough. Because you'll know how likely it is when the game will end the closer and closer you get to the end of the deck of cards containing the game-ender.

One solution is to have a deck of cards with randomized encounters along with that one card you'd want to POSSIBLY end the game. This one card would actually be random itself in whether you continue the game or end the game there. When you get a "continue card," you reshuffle this deck of encounters along with all previous encounters and include a new special event card that may, again, end the game or continue the game. Or you could use a new deck. Doesn't matter. Just as long as the randomizer of encounters does not reveal a good approximation of when the game will end.

Within the context of a "medieval high fantasy" theme (I don't want to use such a played-out theme, but I may end up ultimately using it), you may use a Dungeon Deck full of monsters and treasures and a single special card that determines whether you continue to the next area (shuffle up the deck again with a new special card) or you've found the dungeon exit. Of course, you can always associate boss monster events, etc. to these special event cards.

You can control the randomness of the special outcomes in the game by parsing the deck into multiple sections like what is done in Pandemic. In Pandemic, there are special Epidemic cards. In a regular game, there are four of them in the deck. However, during the set-up of the game, the deck is divided into four separate piles with a randomized Epidemic in each pile. Then the deck's four piles are stacked on top of one another. What this does is create four separate "stages," where, within each stage, you'll experience an Epidemic sometime. This way, you KNOW that the game will not suddenly throw three Epidemics at you three turns in a row and the game quickly ends from there. If you find an Epidemic card during your first turn, then you know you've got quite a few cards to go before you experience another one.

I think I need to determine just how long I want the game to last and then design the game in a way that the game will end within the general timeframe that I want it to end. This can be fine since knowing how much time you've got in the beginning before the game will end isn't as detrimental as knowing how much time you've got in the game left.

In terms of theme, this is what I've been banging my head against a wall over for the past few days. I just can't come up with one satisfying enough and whose flavor can accommodate the mechanics of the game. Having the theme determined from the beginning along with the mechanics of your game is a huge boon, if not vital, when you're designing the game. This is because design can influence the game's theme while the game's theme can influence the design. Design and theme play off of each other and reinforces one another. Perhaps that "finding love while riding a train" theme influences the design to contain some "crying baby" content or random large groups popping up to take all the seats.

I've got the general idea of how the game's pieces will be played out: you'll be revealing cards from the top of the Dungeon Deck (I'll be calling it this for the purposes of this article and perhaps even myself until I settle on a different them or a better name) that may contain monsters, treasure, or that special event card. These will be laid out face up for everyone to see to represent what's currently available to engage with. Then the players will, with their hand of "action cards," decide what they will engage with. This decision will be simultaneous (simultaneous turns) yet secret. What you can engage with will be one of the revealed cards (in a four-player game, I'm guessing having three cards revealed at a time would be best), another player, or yourself. When you engage with another player, you'll be putting them at a disadvantage. When you engage with yourself, you'll give yourself some kind of benefit. It still wouldn't be as good as engaging with the laid-out cards but it's at least a safe bet. Perhaps you'll hide or heal yourself (healing yourself would then force players to not always engage with themselves since the benefits of engaging with yourself every turn would have diminishing retunrs).

...so this is where I'm at. I've got some project planning to do, so I can make sure I keep on track, but I'll figure that out soon enough. Right now, as I write this, it's past midnight, I'm tired, and it's my nana's birthday later on today. Gotta snooze!