Introduction
I’ve been playing video games for over 20 years now and the video game industry has obviously grown and changed a lot. I have my classic games that I hold near and dear to my heart like Skyrim, Halo, and Fable but I do wonder if those old titles still hold up compared to the newer generation of games. With gaming being more widespread, there’s a lot more people chiming in and new markets developers need to fill. So this analysis will look at how the video game landscape has changed over time in regard to player engagement, game popularity, and genre popularity.
Primary Dataset: Backloggd Video Game Data
Video games represent one of the fastest growing entertainment industries in the world, spanning decades of technological change, creative innovation, and shifting player preferences. With thousands of titles released across platforms and genres, the gaming landscape generates a massive amount of user activity, ratings, reviews.
The primary dataset for this project comes from Backloggd and contains information on thousands of video games released between roughly 1980 and 2025. Each row represents a single game and includes information such as its title, release date, development team, user rating, genres, and several community engagement metrics (plays, backlogs, wishlist counts, and more). Backloggd is a place where people virtually track their game collections. Here they can list what games they’ve played, review them, and check their friends games and reviews, somewhat like Letterboxd.
Data Dictionary (Key Variables)
| id |
Unique identifier for each game |
| title |
Name of the game |
| release_date |
Date the game was released |
| team |
Name of developing studio |
| rating |
User rating (0–5) |
| times_listed |
# of times game was listed by users |
| number_of_reviews |
# of reviews the game received |
| genres |
All game genres |
| summary |
Summary provided about the game |
| reviews |
User-written reviews |
| plays |
Users who have played the game |
| playing |
Users currently playing the game |
| backlogs |
Users who have access but haven’t started |
| wishlist |
Users who wish to play the game |
Here’s a link to the data set:
Games Data Set
Interactive Table of the Data Set
Summary Statistics to get Familiar with the Data
Here are some quick summary statistics of the data set in order to get familiar with everything. N_games refers to the number of unique games in the data set, first_year is the earliest release year of a game, last_year is the opposite, and then average and median ratings speak for themselves.
High-level summary of the Backloggd video game dataset.
| 1512 |
1980 |
2025 |
3.719346 |
3.8 |
Distribution of User Ratings
This visualization shows us the overall distribution of ratings across games. We can see a pretty strongly left tailed skew in the data, where most people are giving ratings between 3 and 4.5. Meaning lower scores are pretty rare, but this is pretty common in user rating systems.
Number of Games Released per Year
This visualization shows us the obvious, which is that as time goes on, more games are being released each year. This makes sense, since the video game industry is constantly growing and this is one of the major reasons behind this analysis. Since the video game industry has grown so much since I started playing games, how has it changed exactly aside from sheer quantity of games?
Analyzing the Data
To start the analysis of the data, I wanted to investigate if there was a positive relationship between ratings and popularity. This lets me know whether or not user ratings and player engagement actually move together, since sometimes they don’t. I plotted each game’s average user rating against the number of times it appears on player lists.
Relationship Between Ratings and Popularity
The scatterplot shows a clear but imperfect relationship between popularity and quality on Backloggd The upward sloping trend line indicates that games with higher user ratings tend to appear on more player lists overall, meaning well liked games generally attract more attention. However, the points are widely scattered, which suggests that popularity and quality are not the same thing. Some highly rated games remain obscure, while some moderately rated games are popular. Popularity and rating move in the same direction, but there is a lot of variation in how individual games behave.
Average Popularity of Games by Release Year
The average number of user lists fluctuates up and down over time without forming a strong upward or downward trend. These changes appear fairly regular for almost 40 years, suggesting that they are normal variations rather than meaningful shifts in player engagement. The one major exception occurs around 2020, where the average popularity drops sharply. While this could be influenced by external factors like the COVID-19 pandemic affecting release schedules or gaming habits, it is more likely explained by the fact that newer games have simply had less time to accumulate listings compared to older titles. Because popularity on Backloggd and other user rating sites builds over years, recent releases naturally appear lower in this kind of analysis.
Genre Popularity by Decade
The genre trends show that while popularity rises and falls over time, the overall order of which genres players prefer hasn’t changed much. RPGs, platformers, shooters, and adventure games stay strong across every decade, with only small bumps that likely reflect short-term trends rather than big shifts in taste. Indie games move around a bit more, which makes sense since that category includes a wide mix of styles and release sizes. Overall, players consistently come back to the same core genres, even as the industry grows and changes.
General Co-Occurrence Heatmap
After looking at how genre popularity shifts over time, I also wanted to see how often different genres appear together within the same game. The heat map shows that Adventure is the most flexible genre, appearing with every other major category and often with the highest counts. Shooter, Puzzle, Platform, and Strategy games all pair with Adventure more often than with each other, suggesting that many titles borrow core adventure elements, which makes sense. Meanwhile, genres like Turn Based Strategy and Simulator appear much more isolated.
Emotional Sentiment Distribution in Games
After looking at genres, I also wanted to see what kinds of emotions game summaries lean into, since that can play a big role in drawing someone into a game. Using the NRC sentiment lexicon, I counted the emotional words in each description and found that positive and negative terms appear most often, which fits the wide range of tones games can take. Fear, anticipation, trust, and anger also show up frequently, supporting that mix of tension and excitement. Softer emotions like joy, surprise, and sadness appear less often, but they still contribute to the overall emotional tone.
Emotional Tone of Game Summaries Across Decades
After looking at the overall emotional distribution in game summaries, I wanted to see whether those patterns changed over time. By grouping the NRC sentiment categories by release decade, the heat map shows that the emotional tone of game descriptions has stayed consistent. Positive sentiment is always the most common, with negative, fear, anticipation, and trust following close behind. Even as the industry has grown and changed, the way games are described has remained almost the same, suggesting that players have been drawn to a similar emotional experience across the last 40 years.
Secondary Data Source (RAWG API)
In order to get a secondary data-based opinion to compare and help flesh out the analysis, we are going to pull data from RAWG using their free API. RAWG is a huge database, so we are only taking a sample of 40 games, specifically the top 40 highest-rated modern titles, to create a manageable comparison to the Backloggd data.
A quick look at the RAWG sample data set
Most Rated Modern Games on RAWG (2015-2023)
After pulling the highest-rated modern games from RAWG (2015–2023), I first looked at which titles have attracted the most user ratings on that platform. This gives a sense of which games are not just highly rated, but also widely played and discussed in the broader gaming community. There definitely seems to be a lean here, likely due to the small sample size, since the Witcher 3 appears 4 times. However, that is a widely critically acclaimed video game, so it isn’t completely unreasonable.
Backloggd Ratings vs RAWG Ratings
To compare the two platforms, I matched the games that appeared in both my Backloggd dataset and the RAWG sample. Each point in the scatterplot represents a game, with its Backloggd rating on the x-axis and its RAWG rating on the y-axis. The trend line slopes upward, showing that the two platforms generally agree on which games are rated higher or lower. A few games fall above or below the line, but there’s no clear pattern of one site rating titles more generously. Overall, this suggests that players across both communities evaluate major modern games in a very similar way.
Conclusion
This analysis suggests that even though the video game industry has grown a lot in size since I started playing games, what players actually like has stayed pretty stable. Higher rated games are usually more popular, but there is still room for cult favorites and big mainstream games that are only decent, a tale that will seemingly stand the test of time . Across decades, the same core genres keep showing up at the top, and genres often blend in familiar ways, with adventure elements glued onto almost everything.
The emotional side tells a similar story. Game summaries lean on a blend of positive, negative, fear, anticipation, and trust, in their summaries and that hasn’t really changed over time. When I compared Backloggd with RAWG, the two platforms mostly agreed about which games count as “good” which makes it look like players across different communities are reacting in similar ways. The data points to a gaming culture that keeps evolving on the surface while player preferences stay surprisingly consistent over more than 40 years.
