Tackling The Biggest Challenges With Football Data

At Sportlogiq we’ve built the biggest and best team of AI and Machine Learning researchers in the world of sport.

You might think that sounds impressive.  But you also might not know exactly why. After all, we’re far from the only company to talk about AI when it comes to sport, sports data and sports analytics.

So let’s dig into some of the ways in which we believe the team we’re building, and the technologies they’re working on, create genuine value for those at the cutting edge of sport.

And while this piece focuses specifically on football (soccer), we’ve proven this tech with hockey and football too, and we’ll do a deeper dive into those at another time.

Tackling The Biggest Challenges with Football (Soccer) Data

In basic terms, football (soccer) data can be split into three distinct areas - these are event data (what happens on the ball such as passes, tackles and shots), tracking data (the positional coordinates of all players on and off the ball) and physical data (speeds, distances covered).

Event data has become somewhat ubiquitous.  Teams across the world have access to event data from a variety of sources and over the last few years this has become commonplace in professional clubs around the world. 

Tracking and physical data are a bit trickier. These require camera installs, wearables or a mixture of the two.

So there are two problems inherent in the data world at the moment - you have disparate data sets that, when combined, give you exponentially more value.  And you also have a struggle to get two of the main data sets because the method of collection is so inherently unscalable.

So What’s Our Solution?

At Sportlogiq we believe that having the ability to collect data from Broadcast video solves both of these issues, and adds a whole heap of extra benefits at the same time.  It’s this technology that our AI and Machine Learning geniuses have been working on for years.  

But what is it, and why do we believe it is the golden bullet here?

The simplest explanation is that we’ve taught machines how to understand sport, enabling them to analyse any broadcast feed and collect data - contextual data (based off event data), tracking data AND physical data as they do so.

Our ability to collect data from Broadcast video, adding extra layers of context and metrics to existing event data, changes the game for a number of reasons.  Let me explain:

More Data = More Context = More Insight

Traditional event data only tells you so much.  Take the example of a pass - it’ll tell you who passed the ball, where from and to whom. Combining this with our broadcast data adds an entirely new layer of context that dramatically changes the possibilities for analysis. From a simple pass event will now also tell you the amount of pressure both the passer and recipient were under, the other options available to the passer and whether the pass broke a defensive line or bypassed opposition players. 

The possibilities when adding this layer of context to every event are almost endless.

Globally Scalable

Using our technology to collect data from broadcast feeds means that it is now possible to scale this level of data collection to over and above anything previously seen. Not just existing data sets, but historical ones too. This opens up a world of possibilities for global scouting and recruitment.

Teams are used to worldwide scouting using video but the ability to combine this with data-driven, objective analysis will be a real game changer for a sport in which geographical boundaries are increasingly being broken down.

Genuine Like for Like Player Comparison

By utilising our technology to collect data from broadcast video, a club will be able to compare players in all of the different categories - technically, tactically and physically and relate this directly back to the squad that they work with every day.  For too long teams have analysed their own players in one way, high profile players in another and little known potential stars in another.

A consistent data set, combining all of the ways in which performance is analysed, removes this issue.  For scouting this is a massive win, as you can be more certain that a potential recruit matches up more accurately to the player he is potentially replacing.

Recruitment is a high-risk, high reward activity, and one that can be the difference between success and failure within a club. Utilising a globally consistent data set is a key way to remove some of that risk.

So, What Does This Mean for Clubs

The key thing for clubs is to maintain the workflows that they’ve worked hard to streamline over the years. Match, team and player analysis isn’t a new thing in the professional game and ever since its inception professionals have been iterating and refining a way to maximise the efficiencies within the process.

Sportlogiq understands this, and we’ve made it as easy as possible to augment your current data supply with our contextual and physical metrics. This enhanced data we provide integrates directly into the platforms you use every day, allowing you to easily compare video sources against these all-new data categories.

This not only makes the data you already have access to more valuable, it does so without disrupting the way you work.

So for clubs this opens up a whole new world of recruitment analysis - new data, providing new insights, accessible in a way that you can use instinctively. 

Sportlogiq’s data will help clubs to See Football (Soccer) Differently.

For more information contact soccer@sportlogiq.com