This is a report given to the class "Experiment Design in Computer Science", taught by Professor Claus Aranha during Master's course in Computer Science in the University of Tsukuba in 2022. You can find more info about this course here.
This report aims to investigate the results of soccer games regarding the win-ratio of soccer teams in each place they play. That is, analyze the ratio of wins for the home team and compare different sets of positive results. A draw may or may not be a positive result, and for this reason experiments are done by considering all possible roles for it: draw as a negative result, positive result and not considering draw in the analysis.
It is expected by the end of this experiment to have ratios of positive results against negative results for the home team and determine with a certain level of confidence an interval of percentage of games that the home team obtains a positive result in different considerations for a draw.
This project started as a way to practice coding in pico8. I didn't finish all days of the challenge (yet?) but it was very fun coding and seeing stuff working in pico8.
This game was my first game and also my entry on lowrezjam-2023! It is a platformer where your goal is to take this bloob to the top!
The main goal of this project is to practice Evolutionary Algorithms using python. In order to do this, an evolutionary approach will be applied to a specific problem. In this case, the problem to be solved is to determine the combination of atributes of an NPC from a forum rpg so that, such combination of atributes achieves a specific probability of winning a fight agaisnt a given opponent.
In this RPG each character has atributes that are compared when they battle. It is difficult for a Game Master to build a balanced file for an enemy. The idea of this project is to take the burden of the GM, so they can focus on story alone. This code will receive the Player's data (atributes), and a probability that states the chances of this enemy winning a fight agaisnt the player. The GM might want to create an easy enemy, so they can give as input 30%. This means this enemy won the Player 30% of the times in X number of fights. The code will run an EA where each individual fights agaisnt the player and the % of wins acts as it fitness. The fitness function is how far from the desired win-rate the NPC is.
Updates:
04-26-2024 - Rough program is done. Now to make a simple interface.