iRacing GT3 Series Player Deep Dive

A new GT3 series in iRacing

iRacing introduced its first fixed-setup GT3 series for Season 2 in 2021, the Fanatec GT3 Challenge. One of the main barriers for progress in the sim racing world —particularly as you climb up the skill ladder— is access to good car setups. Good may mean different things to different racers. Some drivers lean towards having stable setups above anything else, others prefer setups that prioritize raw speed, and many others go for something in between. However, many sim racers are not particularly interested in spending precious racing time every week setting up a car, or in paying/looking for setups online.

Excessive spare time is not a luxury that many people have in a world with an ever-increasing array of competing entertainment options. The newly created series gives this audience the option of logging in, picking a car, and going racing in a quick 20-minute race with a level playing field.

There may have been some concern that the new series would cannibalize drivers from the popular GT3 open-setup VRS Sprint Series given that it’s also a B license series, but the team at iRacing scheduled it in such a way that it wouldn’t overlap with the already established one. I —for one— was worried that the new option would dilute the GT3 driver base and spread it thin between the two series. I was curious to dive into the data and analyse the composition of the driver pool, how I fared in comparison to the rest, and how players raced in these two series this season. In the process I uncovered some interesting insights across different regions and divisions in the GT3 iRacing world. This post is aimed at sharing some of those findings.

Motivation and Goals

The initial goal was just to take a quick look in Excel at the distribution of drivers across divisions in these series to compare myself. Since I came back to iRacing two seasons ago, I’ve been a Division 3 driver with an iRating that hovers just above 2000. The competitive part of me had always been curious about what that meant with respect to the rest of the population and that’s where the idea for this post came from. Once I had the data in my hands and did a quick histogram, I immediately had many more questions:

• What is the skill level in each division and how far am I from getting up to division 2 or dropping down to division 4?
• How do regions compare to each other and how do I compare to the rest of the drivers in my region?
• Are players racing in both series or are they favouring one over the other?

There are almost endless possible ways of dissecting the data. In the end, I chose to try to analyse the data focusing on division and region —which iRacing calls Club— as my dimensions of choice.

Preparing the Analysis

Data

The data used for this analysis is publicly available in the series stats iRacing member site. Once season 2 was over, I downloaded the corresponding data for the two series in csv format. In total there were over 36k records altogether, one per player per series, with a total of 24,404 unique players across both series.

Tools

The analysis was done in R, using the tidyverse suite of packages, the gt package to create the custom summary table, and the patchwork package to print multiple ggplot objects to a single image. I also used tvthemes to customize my charts.

Process

The workflow for this analysis was nothing out of the ordinary for a data analytics project:

• Dataset preparation using dplyr
• Exploratory data analysis and summarising
• Data visualization using ggplot, gt, and patchwork
• Analysis and presentation of results

Analysis

VRS and Fanatec GT3 Driver and iRating Distributions

To start the analysis I wanted to understand the overlap between the two series and summarise the breakdown by division. The results are shown in the table below.

There were 24,404 players who raced at least one week in either of the two series. The first interesting bit is that 50% of them raced in both series, whereas the other 50% was more inclined to race in the fixed series only (29%) compared to the open setup series only (22%). During Season 1 there were 19,407 players who raced in the VRS series, which means that for season 2 there was an increase of 25% in the total number of players racing in the GT3 series, likely because of the lower entry barrier of the fixed series. However, there was a decrease of 10% in the number of participants in the VRS series from season 1 to season 2, which went from 19,407 to 17,409 drivers.

Another interesting finding —although not surprising— is that participation was higher in the fixed series, with 9.8% more participants compared to VRS. Once again, this is probably related to the lower entry barrier of not having to worry about car setups. This theory is reinforced by the fact that the lower the division, the higher the proportion of players who raced in the fixed series, either exclusively or along with VRS (up to 90% in division 10 compared to 65% in division 1).

Overall, these numbers indicate that there was no clear dilution of the player base in either of the two series, and on the contrary, overall interest and participation in GT3 racing increase significantly during the first season of the Fanatec GT3 Fixed series.

By Division

This histogram of drivers by division shows the number of drivers in each division, their average iRating, and the percentage that they represent out of the total number of drivers in the two series. The main purpose of this plot is to illustrate the size of each division.

Division 3 drivers make up the largest group with 23.5% of the total GT3 pool and an average iRating of 2,501, followed by Division 4 and Division 2 drivers who have an average iRating of 1,942 and 3,252 respectively. As it can be seen, being in Division 1 or Division 2 means being part of the top 20% drivers in iRacing, with Division 1 being a considerably more exclusive club. The top 3 divisions make up 43.2% of the total driver pool. Even though the iRating data point in the histogram above is interesting to know, it is clear that the average alone is not a good summary point to describe a group as it can be highly impacted by outliers.

The next step was to look at a more detailed breakdown of the iRating distribution by division for the 24,404 GT3 drivers. This box-plot comparison shows a much more informative picture of how iRatings are distributed in each division, and how to divisions compare between them.

As it can be seen, there is a significant spread within each division, and some strange outliers. The horizontal line in each box is the median iRating (50% of drivers are on either side of the line), with the bottom and top borders being the first (25th percentile) and third (75th percentile) quartiles of the distribution. This chart gave me the first answer to my question of where I was relative to the rest of the driver population. If I want to start thinking about division 2, I better up my game and aim for an iRating of round 2,500 at least. Furthermore, if I ever want to get closer to the aliens, I need to somehow find the pace to get closer to an iRating of 3,900 to 4,000. In any case, this chart can give you an idea of where you are within your division and how far you are from going up or down. Needless to say, I don’t know the algorithm used by iRacing to assign you to a division, and almost certainly it is also dependent on your region.

By Club

Speaking about region, I also wanted to know where in the world are the fastest iRacing drivers. In order to find out, I did the same breakdown as above but this time comparing regions or clubs instead of divisions. I wanted to know which clubs have the largest driver pools and how do they rank in terms of iRating.

Let’s start with the driver distribution by region. The largest Club is Germany-Austria-Switzerland (DE-AT-CH), with 13% of drivers coming from that region, followed by Iberia in second (10.5%) and UK and Ireland in 3rd with just over 10% of the drivers. These three regions alone make up for just over a third of the total. If we were to sum all 20 clubs in the United States, they would represent just over 22% of the drivers, making it the largest country in the service. However, for this analysis each club is considered individually. I found it a bit surprising to see Australia and New Zealand in fourth given that their total population is just over 30 million people.

If you look a bit closer in each bar, you can also see the average iRating for each region. If this was our only indicator, DE-AT-CH would also be the fastest region. However, once again the average is not enough to draw this kind of conclusion.

As with the breakdown by division, I created the iRating box-plot comparison by club to get a better understanding of the pace comparison between clubs. The horizontal line in each box is once again the median iRating for each club.

Now that we have a more detailed breakdown of each club, Finland rises to the top with a median iRating of almost 2,000, followed by Hispanoamérica in second and Italy in third. Germany-Austria-Switzerland drop to a very respectable fifth place considering the total size of the club (3,278 players), which is almost 8 times the size of the Finland club, over 4 times larger than Hispanoamérica, and almost 6 times larger than Italy. As for myself (at just over 2000 iR), I’m close to the 80th percentile within my Canada club, which stands in the 20th spot.

More questions

The analysis presented here barely scratches the surface of the insights we can get from this data. Some interesting questions that come to mind are:

• How do these metrics compare between the two series?
• Does one series attract faster drivers than the other?
• Are there differences between regions?
• Does one of the series have a higher engagement than the other?

I may re-visit this data in the future or look at different series to keep finding more about the iRacing community.

Conclusion and acknowledgements

This project was born from a simple question about my pace in iRacing and evolved into a personal project of building a blog website and dusting off my R coding skills to work on a topic that I’m passionate about as is sim racing. Big thanks to iRacing.com for making this data available to the community, to Alison Hill for her Blogdown posts, and the members of the R community for sharing their knowledge.