Videogame Sales based on Genre and Platform

Introduction

Since the dawn of mankind, one question has always been asked, “What can I do to ease my boredom?” People have tried various methods such as drawing, painting, litterature, sports, even games. One of today’s common day answers has been to play a game, in particular a video game. The earliest known video game was Pong, a game created during the 1950s which was used to mirror tennis. As technology improved, new types of video games were created and devices made to handle them. Today, the video game industry is a multi-billion dollar industry with a wide variety of games. Video game genres that can appeal to different types of people such as strategy, action, adventure, racing, role-playing, sport, shooter, puzzle, simulation, platform, fighting, etc. Strategy video games are video games that emphasize skillful thinking and planning to achieve victory. Racing video games are video games where the player participates in a racing competition. Role-playing games are video games where players assume the roles of characters in a fictional setting. Sports games are video games that simulate the practice of sports. Shooter games are video games where the focus is almost entirely on the defeat of the character's enemies using the weapons given to the player,usually firearms. Puzzle games are video games that emphasize puzzle-solving. Simulation games are games that are typically designed to closely simulate real world activities. Platform games are video games where the objective is to move the player character between points in an environment. Fighting games are video games that involve combat between pairs of fighters.

People are now able to play on devices such handheld consoles, consoles, PC, smartphones, etc. Some iconic video games include Super Mario Bros., Pac-Man, Pokemon, Tetris, Pong, etc. Everyday, video games are becoming more commonplace than before. According to Statista, there are roughly 3.24 billion gamers across the globe. And with a multi-billion dollar industry, one might say that video games are quickly becoming the preferred pastime of people. There is a lot of money spent in developing games and by analyzing factors such as sales, genre, and region. We will be able to pick up on trends to help maximize potential profit.

The goal of this assignment is to analyze sales of video games across the years, different genres, and regions to answer questions such as “Which is the most popular genre?”, “What genre sells best by region?”, and many more. For people unfamiliar with video games, we hope to provide an adequate understanding.

Getting Started with the Data

We used Python 3 with a few imported libraries: pandas, numpy, matplotlib, and seaborn. The Video Game Sales dataset, "Video Game.csv", we are using for this project we got from Kaggle

Reading the Data

In the below Video Games dataset, there are 15 columns. We have

Tidying and Modifying the Data

We decided to drop the columns "Publisher", "Critic_Score", "Critic_Count", "User_Score", "User_Count", "Developer", and "Rating" because these columns will not be relevant in our analysis, and many of the entries in the dataset were missing data in these columns. Due to those missing entries, it would be impossible for us to accurately use information based on those columns.

We also decided to drop games released before 2000 to make the data more recent along with games after 2016 since there are only 4 of them, games that have DC(Dreamcast) or WS(WonderSwan) as their platform since these platforms had little data and we had never heard of them until looking at this dataset, and games with a missing value for their Platform. We also convert the years from float to int to help with visualization and other future operations

Exploratory Data Analysis

One of the first questions we want to ask is “Which genre sells best by platform?”

First, we divided the data based on platform where

Just from these tables we can see which Game was the most sold for each platform and how many games were sold on each platform. Based on this data, from 2000 to 2016 3DS sold 512 units of games, DS sold 2120 units of games, GB sold 27 units of games, GBA sold 786 units of games, GC sold 542 units of games, N64 sold 70 units of games, PC sold 912 units of games, PS sold 274 units of games, PS2 sold 2127, PS3 sold 1306, PS4 sold 392, PSP sold 1193, PSV sold 427, Wii sold 1286, WiiU sold 147, X360 sold 1232, XB sold 803, and XOne sold 247. We will also be using these tables to further analyze the data later.

Pokémon, Mario, are exclusive to Nintendo platforms so it makes sense that those games would have high sales on those devices. It is the same for Halo which is exclusive to Xbox, and Final Fantasy is typically exclusive to Playstation. For Xbox, Japanese sales tend to be very low due to Japanese preferring Sony's Playstation and Nintendo over Microsoft's Xbox.

Then we create 2 separate tables displaying the total amount of sales for each platforms. We decided to separate each table based on the type of console since it would make more sense to compare a handheld console against other handheld consoles.

We also want to ask is “Which genre sells best for each region?”. Now we divided the data based on Genre

Just from these tables we can see which Game was the most sold for each genre and how many games were sold on each genre. We will also be using these tables to further analyze the data later.

Then we create another table displaying the total amount of sales for each genre, and the average amount of sales for each genre based on region.

Now we will start analyzing graphs

We chose a custom palette of colors because the default colors were very similar so it was hard to observe the graphs

In this scatter plot you are able to see how much each individual game sold for each year along with what genre they belonged to. Most games stayed under the 20 million mark but some games like Wii Sports in 2006 or Mario Kart Wii were able to hit higher number like 82.53 and 35.52 million. There does not seem to be a trend in the data just a few peaks for games that did really well.

This bar graph shows the Total Global Sales for all games in each Handheld Platform. DS is at the top with about 800 million total sales and GB is at the bottom with less than 50 million total sales.

This bar graph shows the Total Global Sales for all games in each At Home Platform. PS2 is at the top with over 1200 million total sales and N64 is at the bottom with less than 100 million total sales.

In this bar plot you are able to see the Total Global Sales for each Genre. Action is at the top with over 1500 millions sales and Strategy is at the bottom with less than 200 million sales.

In this bar plot you are able to see the Average Global Sales for each Genre. Shooter is at the top with about 0.8 million sales per game and Adventure is at the bottom with about 0.15 million sales per game.

In this bar plot you are able to see the Average North American Sales for each Genre. Shooter is at the top with about 0.45 million sales per game and Adventure is at the bottom with about 0.75 million sales per game. The main difference this plot has vs the global sales plot is that Racing games were much higher than Role-Playing games in North America.

In this bar plot you are able to see the Average European Sales for each Genre. Shooter is at the top with just over 0.25 million sales per game and Adventure is at the bottom with just under 0.05 million sales per game. Just like North America, Europe also sold more Racing games than Role-Playing games.

In this bar plot you are able to see the Average Japanese Sales for each Genre. Role-Playing is easily at the top with just over 0.175 million sales per game and Shooter is at the bottom with just under 0.025 million sales per game. This Graph is drastically different from the first 3. The first thing to note is Role-Playing is far at the top where in the other graphs it was just average. The second thing to note is that all the other genres are at about the same level. The third thing to note which is the most shocking is that Shooter games were at the bottom for Japan, but in the first 3 graphs it was at the top.

In this bar plot you are able to see the Average Other Regional Sales for each Genre. Shooter is at the top with over 0.08 million sales per game and Adventure is at the bottom with just under 0.01 million sales per game. Other Regions had a similar graph to North America and Europe which means Japan is almost like an outlier.

In this graph you can see how each genre performed on average for each year. Shooter is the one that sticks out the most and seems to have an overall postive trend. Platform also seems to have a positive trend until 2013.

Now we start analyzing genres based on platform

We decided to do the more recent platforms since the other ones don't really make games anymore

This bar graph shows the average global sales for each genre for the 3DS. At the top is Racing at 1.5 million sales per game and Adventure is at the bottom with about 0.1 million sales per game.

This graph shows the average global sales for each genre for the 3DS for each year. In 2011 both platform and racing were both at the top, in 2012 platform and simulation were at the top, in 2013 role-playing and simulation were at the top, in 2014 fighting was far at the top which is most likely because thats when Super Smash Bros was released, in 2015 it was simulation, and in 2016 it was role-playing.

This bar graph shows the average global sales for each genre for the PC. At the top is Role-Playing at 0.5 million sales per game, and both fighting and puzzle are at the bottom with about 0.5 million sales per game.

This graph shows the average global sales for each genre for the PC for each year. At first misc is at the top in 2000, but this genre does not have data for every year on the PC. Simulation has the most obvious peaks in 2002, 2009, and 2014. Role-playing had the highest peak in 2004, and 3 much smaller peaks in 2007, 2012, and 2015.

This bar graph shows the average global sales for each genre for the PS4. Easily at the top is Shooter at over 2 million sales per game which is most likely due to Call of Duty being such a popular game series on the PS4, and puzzle is at the bottom with a number so small it might as well be 0 million sales per game.

This graph shows the average global sales for each genre for the PS4 for each year. Shooter is consistently at the top for all 4 years which makes sense when looking at its corresponding bar graph.

This bar graph shows the average global sales for each genre for the WiiU. Easily at the top is Racing at over 2.5 million sales per game and adventure is at the bottom with a number that seems to be less than 0.1 million sales per game.

This graph shows the average global sales for each genre for the WiiU for each year. Platform begins at the top in 2012 most likely due to a Mario game. In 2014 Racing and Fighting peak which is probably due to Mario Kart 8 and Super Smash Bros. Shooter peaked in 2015 due to Splatoons Release.

This bar graph shows the average global sales for each genre for the Xbox One. Easily at the top is Shooter at 1.6 million sales per game which is most likely due to Call of Duty being such a popular game series on the Xbox One, and Strategy is at the bottom with about 0.1 million sales per game. Also note that there is no puzzle games listed for the Xbox One.

This graph shows the average global sales for each genre for the Xbox One for each year. Shooter is consistently at the top for all 4 years which makes sense when looking at its corresponding bar graph.

Linear Regression

In this last section we will be predicting the trend for each genre using linear regression. By doing this we gain an estimate to how future sales are likely to do. A negative slope will suggest a decrease in sales globally. Where a postive slope will suggest an increase in sales globally. This could be useful for comapnies when they are trying to develop new games, and this could also be useful for investors trying to invest in companies.

For this graph there seems to be a negative trend for action games. There is also an outlier in 2013.

For this graph there seems to be a negative trend for adventure games

For this graph there seems to be a negative trend for fighting games

For this graph there seems to be a negative trend for misc games

For this graph there seems to be a slight positive trend for platformers. This is also the first graph so far to have a positive trend.

For this graph there seems to be a negative trend for puzzle games

For this graph there is a very slight negative trend for racing games

For this graph there is a obvious negative trend for role-playing games

This graph has a very obvious positive trend for shooter games since games like Call of Duty and Halo do very well.

There is a negative trend for Simulation games, but there is a bit of an outlier in 2005.

There is a positive trend for sports games since games like Fifa, Madden, and 2K usually do very well.

Last but not least we have strategy games which also have a negative trend.

Conclusion

With an ever growing industry, video games are becoming commonplace in society. Based upon what we have done in this project, we see the value that data analysts bring when analyzing vast amounts of data. Initially, we thought that the action genre would have been the best performing genre globally for sales but the shooter genre was the best performing genre globally. We also saw the regional difference between countries. For example, in Japan, the best selling genre is the role-playing genre whereas in North America, the shooting genre is the best selling genre.

If you are interested and you have more recent data involving more recent console platforms such as the Nintendo Switch, PS5, and Xbox Series X. You could analyze other factors such as micro-transactions, season-passes, and how COVID19 affected sales.

So in conclusion, we have modified the dataframe to remove columns and rows that were not needed for our analysis. Hopefully, this tutorial provided an insight into what game developers, and analysts have to take into consideration when trying to make a new game. We have graphs showing the global sales of video games by genre, region, and platform. This information can be useful if a developer wants to figure what type of games will potentially lead to the most profit. We have shown that we can answer questions such as “Would this genre sell well on this video game platform?”, “Which genre is best for maximizing sales globally?”,or “Is it worth it to develop a game in a particular genre?”