Courses/CS 461/Winter 2006/Chris Lemcke/Week 5
From CSWiki
Contents |
[edit] Week 5 Homework
[edit] Introduction
I am working on a phenomena called the "Meta-game" phenomena that occurs in a popular game I play called Magic:The Gathering. More information on the game can be had from the MTG website, and a basic tutorial from the wikipedia page on the rules.
However, I will sum up the neccissary knowledge for the "Meta-game" Phenomena. Each player has a deck of 60 cards that look something like the following:
[edit] Card Back:
[edit] Some Example Cards:
[edit] Deck Contraints
It is unnessicary to describe the rules of the game, but there are a few basic rules that govern what cards are allowed in each deck. The most popular set of rules, or format, is called Type 2. In this format, only the most recently published cards are allowed, and only a certain amount of each card is allowed. What happens is that players build similar decks using the more powerful cards, but these decks tend to fall into certain catagories. For example, an "aggro" deck attempts to play as many spells as fast as possible to overwhelm it's opponent, while a "control" deck tries to maintain control of the game in it's favor. Since there are a limited number of cards, and some are more powerful than others, players end up playing very similar decks. Although the each game is randomized (by shuffling the decks before playing), overall match-ups are very predictable. Thus, what arises is often a loose "Rock-paper-scissors" arrangement, where deck A usually beats deck B, which beats deck C, which beats deck A.
[edit] The "Meta-game" Phenomena
The "meta-game" describes how certain decks become more popular, and even how "anti-decks" arise. What happens is that players will play more successful decks, thus limiting what decks are played, and also (hopefully) creating emergent patterns of deck popularity. An anti-deck arises if one deck becomes so popular that some players begin to play decks that are designed just to defeat that deck, but aren't very good versus others. However, these decks still often do well!
[edit] Modeling this
In my model, each patch is assigned a deck value. The three decks I'm using are "aggro", "control", and "tricky".
