They have been a long time in the CraqueStats workshop but now the revamp of our radars is complete and ready for launch for forwards.

There are lots of radars or polar bar charts out there, so why do another?

We built our first radars back in 2017. The main reason was that, while there were many out there, we needed our version to run the players we needed rather than relying on what was already posted on social media and the rest of the internet.

If you want to produce content, you need to be able to output visuals to support it. Without being able to do so, you will have to focus your work on what is available. If you are wanting to talk about Stephen Lichtsteiner joining Arsenal, for example, good luck searching for a radar you could use.

Or, when a young Bundesliga forward is snapped up by Bayern, you may want to compare him to others who have much more hype about them.

There was also a concern about how radars were being viewed in general on social media. It felt like a collection of what is considered the ten most valuable metrics in a position put together with the best player being whoever has the largest surface area.

The problem with this is the quantitative fallacy (or McNamara Fallacy). Some people assume that everything in the radar is of value or equal value. That everything not on the radar can be ignored. Or that which cannot be measured (e.g., movement to create space) isn’t important.

The first step is to measure whatever can be easily measured. This is OK as far as it goes. The second step is to disregard that which can’t be easily measured or to give it an arbitrary quantitative value. This is artificial and misleading. The third step is to presume that what can’t be measured easily really isn’t important. This is blindness. The fourth step is to say that what can’t be easily measured doesn’t exist. This is suicide.

Daniel Yankelovich, Corporate Priorities: A continuing study of the new demands on business.

How do CraqueStats Radars help identify different ‘types’ of players?

We aimed to break the radar into phases of play to make the radars to make them more tactical. With of our forwards’ radar, this should help highlight those who:-

  • press from the front
  • drop deep to help in the early building up of attacks
  • progresses the ball up the pitch either through passing or carrying the ball
  • offer penetration by committing opponents or getting the ball into dangerous areas
  • play between the lines as a creator
  • are more focal points in attack whose job is to stay at the highest point in the formation, stretch the opposition defence, and be an outlet the moment the ball is regained by their teammates.

As a result, you can identify which phases of play certain players are most involved in and create stylistically similar comparisons.

CraqueStats Radar comparing Gabriel Jesus of Manchester City and Roberto Firmino of Liverpool. Data as at 21 Match 2021.
Radar Comparison: Brazilian Internationals Gabriel Jesus and Roberto Firmino. Click to Expand.

Are you only doing them for forwards?

At the time of launch, yes. We are building this in Tableau and will simply duplicate the workbook and change the metrics and categories for each position. Before we do so, we want to make sure there are no major issues with it before duplicating it to avoid having to solve the problem multiple times later.

Therefore, now we are launching forwards to allow us to get content out into the world. If no major problems appear with thousands of extra sets of eyes on it, then we will get to work on the others.

Can I Make a Request?

Sure. Priority is given to our Patreons, of course. Feel free to tweet us @CraqueStats or me personally @BabuYagu and we will output them for you when possible. Please note, some may be held back if they relate to an article we are currently working on. 

Can I use them on social media/an article?

Sure. We only ask that you attribute it to us when you do so (@CraqueStats) and include a link to the article you have taken it from, if applicable.

Any known issues?

The dataset is run based on a minimum minute played threshold. This prevents small sample sizes from skewing the data. For now, we have done this based on 20% of the maximum minutes played by any player. I have seen others simply cap theirs at a flat 900 minutes played which works out at roughly 30% of a full Bundesliga season. A lot of fringe players and breakout young players may struggle to play a third of available minutes in a season.

Also, some players who may pass the ‘minutes test’ for 2 clubs (like Taki Minamino) may be excluded if their time at either club is below the cut-off mark. We are working on a solution to this. However, it affects a handful of players.

Lastly, we have some players whose data isn’t pulling through properly and we have excluded them to avoid problems. The only well-known player this affects seems to be Ryan Bertrand. We are working on a solution to this also.

How do I report any problems I find?

Like making a request, you can write to CraqueStats or me via Twitter.

Can you explain what the metrics mean?

Some of the metrics are already in our glossary, which is linked below. I also plan to write something in the coming week to help people read the radar, explain the power bars and the frequency distribution plot in the coming weeks. Before doing so, I wanted to get some public feedback to see if any major changes needed to be made before doing so.

I want to see more of there. Are there any articles or threads using them available at the moment?

At the time of launch, there is one accompanying article featuring Tammy Abraham and Timo Werner. Links to any further articles or threads will be added below in due course.

Tammy vs Timo: Why do both Lampard and Tuchel prefer the German International?
Adama Traoré vs Jérémy Doku: Attackers who can’t be judged on end product alone

A glossary of all the terms used in this article and throughout the site as a whole is available here. Also, click on any image in the article for a full-size high definition version.

All data used in our articles is sourced from Understat, FBRef, Sofascore, Transfermarkt and 538.

Leave a Reply