The first stage of this research project involved accessing a software called MarI/O which is like a mario clone. The next part of this project is observing an AI mario attempt to play some levels and collected data from that. Through collecting and analyzing more data this project is trying to figure out ways to make the agent learn better.