Personal Motivation and Objectives
Modern board-games (also known as ‘Eurogames’) are of particular interest to artificial intelligence (AI) researchers because state variables of most modern board-games are discrete, and decision making is turn-based. The gameplay in modern board-games often incorporates randomness, hidden information, multiple players, and a variable initial setup that makes it impossible to use opening books (Szita, Chaslot & Spronck, 2010).
According to Pfeiffer (2004), Catan is interesting for AI researchers, because players can choose from a large action-set, they have to balance long and short-term decisions, always depending on the performance of their opponents.
There are papers which focus on analyzing optimal game strategies in Catan (e.g. Guhe & Lascarides, 2014; Szita, Chaslot & Spronck, 2010) but it is important that these analyses do not involve robbing strategies.
The mentioned AI researchers worked on the rational model of how to play Catan. Although when people are playing against one another, irrationality quite often occurs.
Today, various versions of Catan exist. Originally, we are talking about a board game, but since then it has been computerized, including offline and online versions. Playing Catan against strangers via the internet provides a chance to make decisions more on statistically and rational basis instead of emotional ones when playing against friends or family members.
Playing Catan online, it was experienced that revenge is an emphasized part of the game. It was noted, that revenge ruins the chances of winning and these experiences gave the motivation to look at this phenomenon more closely.
Thomas (2003) collected six types of decisions that are made during the game. One of them is ‘Evaluating the other players’ positions’. In Settlers of Catan, it is important to be able to accurately evaluate the other players’ positions to determine how close each one is to winning and what they need to do in order to win. This evaluation system (developed for AI) uses the unit of turns. Based on the evaluation process player chooses which opponent’s development to slow down (Thomas, 2003).
That method seems reasonable, but it would be a hard calculation method for a human player and due to this fact, there are differences in how player calculate or estimate their winning chances.
The aim of this paper is to analyze the optimal strategy regard to revenge, furthermore to try quantifying the impact of revenge on the winning chances.