Minecraft settlement generator

Abstract

Procedural Content Generation (PCG) refers to algorithms used to automate the creation of game material that an individual would typically produce. Being heavily motivated by the Generative Design in Minecraft (GDMC) competition, our proposed solution is a Minecraft settlement generator that intends to use PCG to emulate humans in constructing urban settlements. Our generator verifies that structures are properly situated and that no overlappings occur by continually analysing the changing environment. The generator also includes districts and regions with buildings adhering to a corresponding size and aesthetic according to their designated location whilst guaranteeing that every building is connected to the settlement through the A* pathfinding algorithm. We employ various evaluation strategies to analyse our settlement generation, including an expressive range analysis where we executed our generator 100 times to understand its capabilities better. We discuss the division of districts and demonstrate how the average number of houses in each district is approximately 25%. Moreover, we illustrate how the number of houses increases whilst expanding outwards due to the average house size decreasing, where we obtained a negative correlation for the inner and middle regions and no correlation for the outer region. Furthermore, findings indicate a negative correlation between the average path length in blocks and the number of houses, as when the number of houses increases, the average path length decreases. We also recognise that deciding whether the algorithm produces the expected result is subjective, so we gathered individuals to form part of a judging panel and help us replicate the judging process held for the GDMC competition. Three different settlements were used for this process, with two of them being the first and second-placed settlements from the competition’s 2018 iteration and the third being our generator. After participants assessed the three settlements, a focus group session was conducted to obtain comprehensive insights about our generation. Results indicate that our implementation performed well, gaining the highest score in the aesthetics category and obtaining a good overall score whilst receiving very positive feedback throughout the focus group session.