www 101

All you need to know about the internet

Have a Question?

If you have any question you can ask below or enter what you are looking for!

What’s in a Name? Data Analysis of 5,820 Steam Games

What’s in a name? More than you’d think. Whether you’ve been working on your game for a long time or you’re just entering the planning stages, at some point you’re going to need a name. The title of a video game pushes the boundaries of our imaginations and causes us to recall memories. 

With more than 6,000 games being released on Steam each year, how many stir up player interest? Let’s take a look at some research we’ve done on titles released this year.

Why Are Titles Important?

Before we get into looking at the numbers, let’s quickly talk about why the name of a video game is so important. While it may seem like a minor part in the overall experience, the title is key—especially when it comes to marketing your game. In the fast-paced world of digital marketing, your title, a tagline, and an image or two may very well be the only thing a potential player sees. 

When searching through the Steam store, your title and a small image are all you have to promote your game. In the earlier ages of digital gaming, there was also a more limited amount of games being released overall. Pairing the glut of titles being released with this reduced marketing space makes a name more important than ever.

Beyond that, there are many other points to consider. Word-of-mouth marketing, search engine rankings, potential confusion with other existing titles, and more are all reasons to put extra effort into making sure your name is foolproof.

Now let’s take a look at the data.

Entire Set Statistics

We analyzed 5,820 games released this year (no free-to-play games included) to gather the following statistics. We focused on standalone titles released as their own product, so DLC and similar expansions were excluded from the data set. Brief explanations of the data and methods follow each data point.

Title Size

Of the 5,820 titles analyzed, the average title length in characters was 17.15, with a standard deviation of 10.26. So in general, games released this year ranged between approximately 7 and 27 characters, excluding outliers.

In terms of words, the average title length was 2.8 words with a deviation of 1.7, giving a general range between 1 and 5 words.

This shows a general trend towards more concise titles, which leaves two potential choices: following the trend, or deliberately avoiding it. Both offer distinct advantages in the current state of the market, and should be considered carefully when coming up with your own game title.

In addition to these ranges of character and word lengths, many of the games analyzed showed similar naming structures. From the early stages of analysis, there were less than two dozen title structures that accounted for more than 90% of the games. Of note is the current prevalence of titles using the structures “{noun} of {noun}” and “{noun}: {adjective} {noun}” (for example, West of Loathing and Divinity: Original Sin 2 respectively).

Title Size

Multi-Part Titles

15.14% of the total dataset was found to have a title with multiple parts. For the analysis, a title was considered to have multiple pieces when one of the symbols “:”, “-”, or “|” appeared and provided two distinct pieces of information. For example, “Yooka-Laylee” was not considered to have multiple parts, whereas “Steel Division: Normandy 44” was.

Unique Word Usage

One of the more interesting pieces of data collected was the usage of “unique words” within a game’s title. For our purposes, a word was considered unique if it did not exist as a dictionary definition within the ten most popular languages utilized by Steam players. Overall, 27.93% of titles were found to include a unique word within their title.

Unique Word Usage

References to Other Intellectual Property

Looking at our dataset for references to other Intellectual Properties (IPs) yielded a new subset that warrants further analysis later. 11.96% of the games in the dataset were found to reference a previous IP in some capacity—whether the IP was a previous game, movie, book, or other medium.

Using this subset to further research the relations between game sequels, movie-based games, and book-based games could make for an interesting project for those that want to do some number crunching of their own.

Measuring Up

Let’s look at how these titles stack up against top sellers. By looking at the top 10% of games and comparing them to the entire dataset, we can create a vivid picture of what’s important in video-game titles.

Title Size Comparison

The top 10% of games released this year, by sales, averaged a title length of 17.63 characters. Compared to the 17.15% of the general dataset, this shows that the top-selling games have a slightly longer title on average.

The average for title lengths by word count shows a similar result. The top games come in at 2.92 words, compared to the 2.8 words of the full dataset. Enough to show that the top games tend to be a little wordier.

Multi-Part Titles

19.17% of games in the top 10% of sales had multi-part titles, compared to 15.14% for the full dataset. The difference between these two numbers suggests that there may be a benefit to longer titles. This is likely due to the improved utilization of limited space developers have to attract potential players.

Usage of Unique Words

When it came to using unique words, the best-selling games were stellar! 48.19% of titles utilized at least one unique word, compared to just 27.93% of titles in general. This is a huge difference in data, and suggests that using unique words may make a game more memorable, or more likely to entice a would-be customer.

References to Other Intellectual Properties

In regards to other video games, movies, and books, top-selling games referenced them in 20.73% of cases—nearly double the 11.96% of the entire dataset. This shows a clear trend that extensions of existing franchises tend to perform better, or at least have a higher chance of being successful.

References to Other Intellectual Properties

Datasets, Sources, and Analysis Techniques

The dataset used for the research in this piece are provided entirely through SteamSpy and the Official Steam Web API. If you’d like to do your own research or try out other algorithms, those are the best places to start.

When reading through the data, keep in mind that limitations of the datasets and analysis techniques may have left gaps that could affect your results. For example, inherent issues exist, such as the effect of free weekends on sales data and games with sales equivalent to the margin of error.

While this data is still useful, these issues were difficult to control for.

Data Interpretation: What Does It All Mean?

We’ve gone over a lot of data, but what does it mean for you? Here are some of the clearest trends that we found:

Keep Titles Between 1 and 5 Words

Most games stay in this range, with the best-performing ones hitting the 3-4 word range. Try to choose your words carefully, and add as much description as possible within that limited space. If you’ve already found the perfect title, and it is a little short (1 word) or a little long (6-8 words), don’t worry—a little variation isn’t a bad thing.

Multi-Part Titles Yield More Interest

Players tend to prefer titles with multiple parts. The likely reason is that doing so allows for more information to be conveyed within a small space. Using the multi-part title to evoke additional emotions or description is a solid plan.  

Keep It Unique and Interesting

Titles including non-dictionary words are much more likely to perform better, so having something unique about your title can give it an edge among a list of average titles.

Sequels and Franchises Perform Better

Basing your game on a previous work gives it a better shot at being successful. This likely stems from the familiarity that players have with the original medium, but is also good news for indie devs who would like to turn their singular game into a series.

Further Research for Data Enthusiasts

Interested in how to take this to the next level? Here are two ways to expand the analysis done here for even more in-depth results.

Break Down the IP Subsets

We talked a little about the percentage of games that refer to already existent intellectual properties, but this subset can provide interesting data of its own. How many refer to previous games? Movies? Books? Another interesting line of research to follow would be to compare how multiple titles within a series perform. For example, are fourth games as good as third games?

What About the Worst Games?

We compared the top 10% of games to the entire dataset, but what about the bottom 10%? Segmenting this part of the data could lead to interesting new finds, especially when it comes to naming decisions to avoid.

Wrapping Up

The trends revealed can give you a leg up when it comes to naming your game. If you’re interested in doing a more in-depth version of this study, have questions about our findings, or would like to add to the discussion, leave a comment below!