Soccer's top teams tap into data

Scott Lazarus our consultant managing the role
Posting date: 5/3/2016 1:39 AM

Half-time soccer team talks used to be based around eating oranges. Now they can analyze real-time data and react to presentations.

Soccer has become a numbers game.

There is still a huge debate in the game about the use of goal-line technology, but for everything behind the scenes it seems like that battle has already been fought and technology won.

It is no longer just 2-0, 1-3 or - if you are a Barcelona supporter - 4-0 but all about a near-infinite number of figures.

Possession, territory, passes completed, the list goes on and on and on.

It's only a very dedicated soccer fan who would know that Bayern Munich midfielder Javi Martinez won an average of 2.6 balls in the air during each Uefa Champions League game so far this season. But processing this kind of data gives teams a great advantage, both in training and matches.

For goalkeepers, the guess of where a penalty might go is now a mathematical algorithm rather than an instinct.
The goalkeeper will know where the taker has placed previous penalties. If a player has hit 95% of his penalties to one side of the goal, which way do you think the goalkeeper is going to dive?

What this means is that the millions of pieces of data created must be analyzed by teams. Manchester City, for example, have 10 analysts working full-time on analyzing data, four just for the first team, so that no stat that counts - or at least worth counting - is left un-analyzed.

Pick geeks first

City defender Vincent Kompany has initiated a meeting with fellow defenders and the analysts once a week to discuss the data. The club have even experimented with making all this information public and, because of how popular it has been, are expected to resume this next season.

There was a time, not too long ago, when teams in school playgrounds were picked and the ones left towards the end of selection were often those more at home in front of a computer screen than in front of goal. Now Premier League clubs are picking them first.

"We've got a record for every shot across the top few leagues for the last five years," says Sam Green, advanced data analyst at sports data company Opta. "Beyond penalties, the tendency for people to shoot under certain circumstances and where that shot was from." "A lot of things can be done with the data. Some of the things we use it for is to look at how teams react to corners or other set pieces. How likely are they to concede a shot on a counter-attack afterwards?"

But there are limits to how far data can take you, particularly with the often distracted or influenced mind of a soccer player.

Intent

"I don't think any club is at a stage where they can use data comprehensively across every aspect of the game," says Mr Green. "There are things that we're not confident in predicting - which way they would turn, for example. "There's still an issue of intent. The shot went in the top left-hand corner but we don't know necessarily that it's where the striker was trying to put the ball."

What is often forgotten about the current data "revolution" - as some people call it - is that recording the information of soccer matches goes back a long way.

RAF pioneer

"Soccer is the original hotbed of analytics," Prof Chris Anderson, of Cornell University, and author of The Numbers Game, said. "There was an accountant and RAF wing commander by the name of Charles Reep who starting notating soccer games in 1950. "He develops a coding system, sits down at matches and does it for 60 years."

The games must still be recorded manually as automation is difficult in such a complex game. "His work has informed how [data] has reverberated in clubs through the years. "He notated with his own system all the events on the pitch and collected reams of data that he put into practice working with clubs in the 1960s, 70s and 80s. He was the original soccer analyst."

Mr Reep, through his findings, has been cited as one of the main architects of the long-ball game.
Another man with a reputation as a long-ball manager, Sam Allardyce, is seen as one of the main advocates of modern data use.

Mystery

Beginning in 2000 with a change in approach at Bolton, a company called Prozone was enlisted to provide detailed statistics that Allardyce has quoted and used time and time again. He believes his use of modern techniques improves performance in a way that can be measured. But even his data use is a mystery to some managers.

One, seen as a modern, progressive coach, is critical of the reliance of statistics at all. In a similar way to soccer broadcasts, video is now a key part of feedback from coaches. "I have never used Prozone. I don't use it because I don't believe [in it]," Tottenham Hotspur manager Andre Villas-Boas said in a press conference. "You always have to be very, very careful with statistics. It doesn't mean that we negate them completely - we just don't use them to the extent that people might think. "We have a scientific department that deals with that but we don't prepare our training or players based on the physical data we get from matches.

'Different approaches'

"The mind and how the player feels is much more important for us, rather than statistical data.

"For me it's useless but it varies from coach to coach. We all have different approaches." For a manager seen as one of the world's new breed of technically astute managers, it seems like the numbers do not add up.

If that's the case, he probably does not even know how many balls Javi Martinez has won in the air.

PITCH IN WITH THOSE FACTS
While some data is great for training staff and coaches, other stats are just interesting to drop into conversation

169 - David James has the most clean sheets in Premier League history, 169 in 572 games
126 - Morgan Schneiderlin has made the most interceptions in the Premier League this season
123 - Arsenal have seen the most opponents caught offside this season
88 - Stephane Sessegnon has been fouled more often than any other Premier League player this season
44 - Ashley Williams has blocked the most opposition shots in this season's Premier League
7 - No Premier League team has conceded more penalties than Aston Villa this season. Manchester United are the only side yet to concede one.

Stats courtesy of Opta. Correct as of 27 April 2013

Click here to watch the video and see the related article on the web.

Related blog & news

With over 10 years experience working solely in the Data & Analytics sector our consultants are able to offer detailed insights into the industry.

Visit our Blogs & News portal or check out the related posts below.

3 Ways Machine Learning Is Benefiting Your Healthcare

With Data-led roles leading the list in the World Economic Forum’s ‘Jobs of the Future’ report, it is no surprise that Data Science continues to be the main driving force behind a number of technological advancements. From the Natural Language Processing (NLP) that powers your Google Assistant, to Computer Vision identifying scanning pictures for specific objects and the Deep Learning techniques exploring the capability of computers to become “human”, innovation is everywhere.  It’s unsurprising, then, that the world of healthcare is fascinated by the possibilities Data Science can offer,  possibilities which could not only make your and my life better, but also save several thousands of lives around the world. To just scrape the surface, here are three examples of how Machine Learning (ML) techniques are being used to benefit our healthcare.  COMPUTER VISION FOR IMAGING DIAGNOSTICS  Have you ever had a broken leg or arm and saw a x-ray scan of your fracture? Can you remember how the doctor described the kind of fracture to you and explained where exactly you can see it in the picture? The same thing that your doctor did a few years ago, can now be done by an algorithm that will identify the type of fracture, and provide insights into how you should treat it. And it’s not just fractures; Google's AI DeepMind can spot breast cancer as well as your radiologist. By feeding a Machine Learning model the mammograms of 76,000 British women, Google’s engineers taught the system to spot breast cancer in a screen scan. The result? A system as accurate as any radiologist.  We‘ve already reached the point where Machine Learning and AI can no longer just outsmart us at a board game, but can benefit our everyday lives, including in as sensitive use-cases as the healthcare industry. NLP AS YOUR PERSONAL HEALTH ASSISTANT  When we go to our GP, we go to see someone with a medical education and clinical understanding who can evaluate our health problems. We go there because we trust in the education of this person and their ability to give us the best information possible. However, thanks to the rise of the internet, we’ve turned to search engines and WebMD to self-diagnose online, often reading blogs and forums that will convince us we have cancer instead of a common cold.  Fortunately, technology has advanced to the point where it can assist with an on-the-spot (much more accurate) evaluation of your medical condition. By conversing with an AI, like the one from Babylon Health, we can gain insights into possible health problem, define the next steps we need to take and know whether or not we need to see a doctor in person.  There’s no need to wait for opening times or to sit bored in a waiting room. Easy access from your phone democratises the process and advice can be received by anyone, at any time.    DEEP LEARNING DRAWS CONCLUSIONS BETWEEN MEDICAL STUDIES Despite their extensive qualifications, even medical researchers can feel overwhelmed by the sheer amount of Insights and Data that are gathered around the world in hospitals, labs, and across various studies. No wonder it’s not uncommon for important Insights and Data to get forgotten in the mix. Once again, Machine Learning can help us out. Instead of getting lost in a sea of medical data, ML algorithms can dig deep and find the information medical researchers really need. By efficiently sifting a through vast amounts of medical data, combining certain datasets and providing insights, ML sources ways for treatments to be improved, medicines to be altered, and, as a result, can save lives. And this is only the beginning. As Machine Learning continues to improve we can expect huge advances in the following years, from robotic surgery to automated hospitals and beyond. If you’re an expert in Machine Learning, we may have a job for you. Take a look at our latest opportunities of get in touch with one of our expert consultants to find out more. 

Diversity In Data: An Overview Of Our Berlin Meet-Up

We started the year off right at Harnham Berlin, following the launch of our first ever European “State of Diversity Report” and working in collaboration with Smava to host an amazing event with three inspiring speakers on the topic. Our second event in Berlin, we wanted to continue with our mission to create a different type of tech meetup, moving away from purely technical discussions and focussing on important non-technical subjects within Data & Analytics and Recruitment. With Diversity & Inclusion more important than ever for both businesses and individuals, we wanted to do our bit to contribute to the discussion and talk about how the industry can move forward. As we were full on the day, and many of those who wanted to attend were unable to make it, I just wanted to put together a short piece on some of the highlights. Here are some of the top points covered on the day: Harnham’s State of Diversity Report David Webb – PrincipAL Consultant | Harnham As industry leaders, we feel it’s our responsibility to share our knowledge with businesses as individuals across the world of Data & Analytics. Alongside our annual Salary Guide, our Diversity report allows us to provide you with a comprehensive overview of the market and, in this presentation, we discussed the state of D&I in Europe and Germany specifically. Research has showed time and again that a diverse workforce drives profitability & increases staff satisfaction, so is it really surprising that having many people from different backgrounds can offer a company a broader range of solutions?Our report surveyed over 3,000 people and shows that not only can you increase profitability and improve staff satisfaction, a full TWO THIRDS of job seekers consider Diversity to be an important factor when analysing a job offer (which Bar Schwartz takes a closer look at during her talk).With a German workforce that’s only 25% female, there is still plenty of work to do in order to achieve greater equality. If you’d like a copy of the full report and want to talk through some of our findings in more detail, please just get in touch.  Everyone speaks about D&I, not everyone is ready for it Bar Schwartz – Head of Engineering | Signavio Diversity is not an outcome of hiring people of different gender or colour; it is an outcome of seeking and accommodating different personalities at work.Integrating diversity to your workplace or team requires education on what diversity is, what personality is, and how people differ. It requires challenging our biases on what the right ways to do things are and what is perceived as good or bad.It has to be a top-down, inside-out solution that covers everything from culture to leadership, every individual, and even your structures and roles. Change can start small. Integrating different people into the hiring process (even if they just observe), exposes people to profiles of diverse people and may challenge your unconscious biases. You can read more of Bar’s thoughts on creating a Diverse workforce here.  How the brain asks for Inclusion, not Diversity Kirsten Brueckner – CMO | mobile.de Our brain asks for inclusion, not diversity. Why? Our brain is incredibly smart in being as efficient as possible. This means that 95% of our decision making is unconscious and 70% of it is influenced by emotions (and we are great in post-rationalising). Most of the time we are on autopilot based on past experiences and knowledge and we don't even know this. We mix past experiences and knowledge with the input we get and form our own version of reality, which is a challenge in communication.What does that tell us about diversity? It’s difficult as we can't be on autopilot if we want to make progress. We need to discard past experiences and question our current knowledge. There are some simple tricks that transform recognising diversity into seeing inclusion; search for similarities (you will always find some), broaden your experience, be consciously conscious and enjoy the ride while learning. How to better advocate for Diversity & Inclusion Anna Mikulinska – CTO | Enterroom Why is it urgent to act? Without exposing the bias in Data, we use the inequality which will become a part of the design of the modern world and this is only amplified by the use of technology.How should we approach D&I? With empathy, and by addressing all the possible doubts Diversity & Inclusion raises. For Managers and Investors to spend money on supporting Diversity & Inclusion we need to make sure they truly understand the value of becoming advocates on their own. It’s not enough to convince someone for five minutes, they in turn need to be able to a buy in from their managers and partners as well. It’s time we stop avoiding difficult questions, let’s address them upfront.Everyone can act, but what can be done? Not everyone has to get on stage or into a board room. We can support progression with D&I by:Creating a D&I friendly work environmentBringing up the topic during the interview process as a potential candidateMentoring a young person willing to enter the Tech world and sharing your story with themNot being silenced by the argument “let’s not do politics” We will be running more events throughout 2020 which are already being planned and hope to see you all there! If you would like any more information, would like to get involved, or if you’re looking for support with your Data & Analytics hiring process, get in touch with our team of expert consultants and we will be able to advise you on the best way forward.  You can download our European Diversity Report here, and our Salary Guide here. 

RELATED Jobs

Salary

US$130000 - US$150000 per annum + Benefits

Location

Detroit, Michigan

Description

An IOT technology company are looking for a Senior Data Engineer to join them in the heart of Detroit! Click below to read more!

Salary

US$195080 - US$219465 per annum + Benefits

Location

Boston, Massachusetts

Description

An award-winning consultancy are looking for a Data Architect to join them in the heart of Boston! The salary is up to $180k, click below to read more!

Salary

US$130000 - US$160000 per annum + stock + bonus + equity

Location

New York

Description

Senior Data Engineer

recently viewed jobs