We provide a third of EFL Championship clubs with physical and technical data, which is integrated with video footage and ready to analyse in Hudl Sportscode.

Now, with some recent additions to the data we capture, our product is even more powerful. We caught up with Hudl’s Kerry Morrow, formerly of Newcastle United, to find out more.

Kerry, talk us through Hudl’s EFL Championship data. How’s it gathered? What’s the process?

With our cameras now installed at championship grounds, we can track every movement of every player throughout an entire game. In total, it’s more than 3 million data points that are collated. From this, our PAS-3 technology then derives statistics including players’ physical loads, number of high intensity runs, sprints and more.

For events, we have a team of analysts go through each game to tag around 2,500 key actions. We strive to deliver this data to all clubs in a good timeframe to help analysts deal with the demands of the footballing calendar. 

How reliable is the data?

We have automatic detection processes to highlight errors in both the physical and technical data. For example, if the system is unable to track a player due to their close proximity with another player, then this will be flagged and manually corrected. Once a game is marked as complete, a quality control manager will then review it from start to finish.

We’ve con­duct­ed exten­sive test­ing and con­tin­u­ally invest in data valid­i­ty and reli­a­bil­i­ty for our tracking data. 

I worked as an analyst in the Premier League for several years so I’m very confident in what we’re providing here at Hudl.

Tell us about the changes Hudl has made in data collection recently.

We’ve listened to feedback from the market and added over 60 new technical statistics.

We’re aiming to provide analysts and coaches more detail than ever before. Set plays are hugely important in today’s game, therefore we are now breaking set play attempts into first and second phase. We have also added the likes of pull back crosses, quality of possession regains (time to regain and where on the pitch), and expected goals (xG).

On the physical side, we’ve added over 20 new statistics, which can also be viewed in XML format for the first time. High intensity and sprint distance are now broken down into possession states and are easily accessible to the sports science, medical and analysis departments.

You mentioned expected goals, can you tell us more about that?

Because of our tracking data, we can take all player positions and velocities into account when looking at expected goals. We can accurately evaluate the chances of each defender to disrupt the shot and look at many possible paths for the ball in three dimensions.  

Most other models only look at basics, like where the shot has been taken from and some subjective ratings. Our advanced xG model is powered by our tracking data and it can see the difference between a good goal chance and one where the defense has shut it down, even if they are taken from the same location.

Do clubs receive anything else from this data service?

Yes, we provide monthly benchmark reports so teams can see where they rank amongst other teams in the league. These have proven to be really popular with analysts, coaches and managers alike.

Feedback on our data has been very positive. Analysts across Europe are using our data more than ever before, and it’s becoming an influential tool across a multitude of departments.