Games are a huge part of life for most people, so throughout this project, our group wanted to find the relation between games from the data we acquired. SOme of the key findings throughout the data analysis of this video game project include finding th ebetter platform between PlayStation and Xbox, finding correlation between critic scores and user scores, predicting similar genres, and predicting what genre a game belongs to based on their description.
Our main motivation for this project was that each of us had a close connection to gaming. A lot of people have played games since their childhood, have a close attachment to games, and throughout the years have many questions on these games. Our main focus was to see if our personal experience in gaming would relate to our research. Would we be able to predict games similar to a game we previously played, and our model return games similar to it and perhaps also be games we played because of their similarity? Can we finally determine whether PlayStation or Xbox are better platforms based on their game scores? This led us to 5 questions we wanted to answer with our data: Is there a significant difference in the meta/user score between platforms and the top 20 genres, can we predict user scores based on meta scores, what does the battle between PlayStation and Xbox look like and which wins based on popularity and their meta/user score, could we build a recommendation system from keywords in the game description, and could we predict the genre of a particular game based on key words in the game description.
- Is there a significant difference in the mean meta score and user score between the platforms and the top 20 genres?
- Can we predict user scores from meta scores?
- What does the battle between PlayStation and Xbox look like? Who wins based on popularity and mean meta score and user score?
- Can we build a recommendation system from key words in the game description?
- Can we predict the genre of a particular game based on key words in the game description
You can also try our Recommender System application Release.