Pandas
Scipy
Plotly Express
Matplotlib
Plotly
Cleaninig Data
EDA
Filtering
Misisng Values
Feature Engineering
Pivot Tables
Hypothesis Testing
T-Test
We are looking at data from Ice, an online retail store for video games. The data set includes categories based on genre, platform, rating, sales, release year, critic, and user scores.
The purpose of this project is to collect historical data, and apply insights to determine whether a game will succeed or not in the future. These insights will be used to optimize capital allocation in regards to advertising potentially big winners. We will be determining total sales, and the distribution of sales across the different platforms. We will illustrate which platforms lead, and lag in sales. Data analysis will determine the most popular platforms and genres. Investigations will determine if the average user ratings of certain platforms are the same, or if the average ratings of certain genres are the same.
Overall, we can look at our data to make predictions for 2017. First, inferences would need to be made as to the region we are most interested in marketing to. Since the North American region generally leads in sales, and since the European region follows a similar trend, we can maximize our results by focusing on North America. Since the most popular platform is the PS4, we will be looking for a game on that platform. Since the most popular genre in North America is action, we will be looking for the next big action game. Furthermore, we would want a game that could be played on XONE or PC as well, to maximize sales. Since ratings and user scores do not influence sales, we will not care too much about those factors. Yet, a game in the action genre will likely be M for mature. It is highly likely that the next iteration of Call of Duty or GTA will be widely successful. Alternatively, if we want to market to the Japanese region, we will look for a role playing game on the 3DS. However, profits are likely to be higher working with the previous strategy.