• You may have to login or register before you can post and view our exclusive members only forums.
    To start viewing messages, select the forum that you want to visit from the selection below.

Liverpool are using incredible data science during matches, and effects are extraordinary

Status
Not open for further replies.

King Binny

Part of the Furniture
Honorary Member


0_export-2020-01-14T172208832.png

Liverpool form a compact block in the centre of the pitch late in the match with Spurs

87 minutes of football had been played in North London last Saturday, as Jose Mourinho's Tottenham Hotspur clamoured in a desperate search for an equalising goal. Liverpool managed to gain a lead earlier in the contest through Roberto Firmino but had been unable to add to their tally, resulting in the final stages of the match descending into chaos, with ball moving rapidly up and down the pitch, and Spurs missing a pair of excellent chances through Heung-min Son and Giovani Lo Celso.

But in one moment, Liverpool managed to kill Spurs' momentum dead. As Dele Alli took control of possession of the ball near the halfway line, Jürgen Klopp's team morphed into a solid, narrow block and gave their opponents a simple ultimatum - 'go around us'.



0_export-2020-01-14T172208832.png

Liverpool form a compact block in the centre of the pitch late in the match with Spurs

10 players wearing red had positioned themselves in the centre of the field, separated from back-to-front and from side-to-side by a total of no more than 20 yards. Unsurprisingly, the ploy worked and the contest ended with Liverpool securing their sixth clean sheet in a row.

The Reds have now conceded at least seven fewer goals than any other Premier League outfit this season, and that record stems from the team's ability to control the pitch. The large majority of football clubs have data analysis departments, but very few, if any at all, incorporate such knowledge into top-level decision-making and processes to the extent that Liverpool do.

Michael Edwards, the club's sporting director, is a former analyst himself having previously spent time at Portsmouth and Spurs. He's since been promoted through the ranks at Anfield, with his current role centred around catering for the medium-to-long term interests of the club as well as ensuring that every department is seamlessly aligned.

Fenway Sports Group, the owners of Liverpool FC, have a history of embracing data. John Henry ventured into baseball before exploring the world of football, and he's renowned for offering Billy Beane, the man behind the concept of Moneyball, a $12.5m deal to become general manager of the Boston Red Sox in 2002. The leading figures on Merseyside have been assembled by FSG over time and are now making a difference to the team's success behind the scenes.

Ian Graham is the director of Liverpool's research division. The Welshman has a doctorate in theoretical physics and believes in evaluating players and wider trends in the sport by crunching sophisticated numbers beneath the surface. Breakthroughs can be made once analysing detail to such a granular level, with the Reds making specific strides in regard to gaining an understanding of what is labelled as pitch control.

Tim Waskett, an astrophysicist, and Will Spearman, who has a doctorate in philosophy, are two members of Liverpool's data science team. Both have spoken publicly about the concept of pitch control, with an example of the type of graphic used to capture the notion pictured below.

0_pitch.png

A pitch control visualisation which captures the regions of space controlled by certain players

The player who is circled in yellow has possession of the ball and, essentially, his team has access to the areas of the pitch that are blue, while the red areas are mostly controlled by the opposition. The player is best advised to pass into a teammate positioned within a blue zone.

Waskett has since provided greater context having appeared at the recently held Royal Institution Christmas Lectures. He stated that by combining event data and tracking data, Liverpool can comprehend how each action on the pitch impacts the probability of a goal being scored. One of the graphics that was used to demonstrate that theory can be seen below.

0_epv.png

Liverpool's pitch control concept considers which areas of the pitch are best to use at any given time

Waskett stated: "The red team is Liverpool and the areas in red are the places they can get to sooner than the players in blue. Everything is turned into a goal probability and this value, 1.3%, is the probability that a goal will be scored with the ball in this position within the next 15 seconds." Such information is highly complex but once the specifics are filtered and used within the club, the findings can provide a platform for Liverpool to gain an edge over competitors that are yet to develop an awareness.

Klopp is the man responsible for coaching the first-team. Any conclusions that are realised by the club's data science department have to be interpreted and applied in a footballing sense by the Liverpool boss.

The Reds have faced only 55 shots on target in the Premier League this season, which is 10 better than second-best Chelsea on 65, followed by Manchester City on 70. It is possible that Liverpool's defence and the team's dramatic central block late in the match with Spurs stems from the club's appreciation and understanding of pitch control.

On the offensive side of the game, Trent Alexander-Arnold and Andy Robertson are celebrated for passing to one another despite playing on opposite flanks. The actions are somewhat unconventional but again, there's enough reason to suggest that those actions stem from the concept of pitch control and which areas are valued as dangerous. Alexander-Arnold faced Spurs at Anfield earlier in the season and his passing network from the clash is pictured below.

0_EIM6Et9W4AIUlv9.png

Trent Alexander-Arnold's passing network against Spurs at Anfield (Image: Wyscout)

The sheer volume of diagonal passes that the young full-back made across the field suggests that his actions were deliberately instructed.

Liverpool's midfield has received criticism in recent years because of the general lack of creativity that it tends to employ, but what if pitch control has played a part? Risks in possession are generally taken by the team's full-backs, while Klopp's central midfielders remain disciplined and focus on retaining balance. Perhaps Liverpool's data science department have found the centre of the pitch to be the most crucial to defend, hence players such as Jordan Henderson and Gini Wijnaldum exhibit a preference to play safe.

On the other hand, the inside-right, the inside-left, and the spaces in behind opposing defenders may be considered to be the most effective to use when attacking, rendering the Reds' persistent probing of those areas an intelligent plan.

0_export-2019-12-05T124723756.png

Liverpool score against Everton by hitting a simple long pass over the defence

Nobody knows how much the Reds have actually discovered behind closed doors, but considering the team are world champions and have lost just one Premier League match since May 2018, it is conceivable that the subtle advantages provided by data science are drastically impacting results on the pitch.

The inclusion of data has changed the landscape of other sports but is yet to truly have a revolutionary effect on football. Until that time comes, Liverpool will continue to hold an upper hand over the chasing pack.
 
A podcast I listened to a couple of days ago mentioned that their stats provider (a Spurs fan), said he'd analysed our matches & since Fabinho has been out, that every single midfield 'pass map' for every match was identical in shape. Something they'd never seen with such consistency before.

That suggests that the 'drills' that we were aware we did, may actually be much more rigid than we (certainly myself) suspected, & may well be based on this kind of sports data science.
 
Fuck me this is mind blowing - we are the most technologically advanced football team on the planet.
 
We better plant computers chips into our players to signal data input and output? Hahaha more conspiracies!
 
Simply fascinating.

It’s also one of the big reasons I’m currently taking courses in data science, and hopefully change my career later this year or next.
Are you doing them distance or on campus? I've been thinking about doing a stats/data science course myself, with the same intention of changing my career.

I probably won't do it as I'm a lazy bastard...
 
A mixture of both. There are loads of free/cheap courses out there e.g. in Coursera, Kaggle (which is a very good place for real-life datasets), etc. I normally sign up using the trial accounts and download everything before it expires. But knowing myself, I'm also pretty lazy, else i won't be spending so much time on 6CM....lol...so i'm thinking of signing up for a professional 4-days full day course (classroom) that boasts of guaranteeing a job after completion, within 3 months.

Harvard Business School and Cornell offers an online e-learning course but many people advised me that you don't need those fancy degrees to venture into a data science career. What they're looking, like many other jobs, is for someone to get the job done, regardless of qualifications. I know of a vet (in his 50s) who got a job as a business analyst after taking a short data science course for a few months. So if someone who spends most of his life with animals and still get a job, i reckon anyone can do it too.
 
Last edited:
I've never met a Data Scientist who believed in Jesus. Not saying they don't exist or anything. Just that they're probably not very good at data science.
 
I've never met a Data Scientist who believed in Jesus. Not saying they don't exist or anything. Just that they're probably not very good at data science.
I've met a few very religious scientists. One didn't believe in evolution but was also an expert in animal nutrition...
 
A mixture of both. There are loads of free/cheap courses out there e.g. in Coursera, Kaggle (which is a very good place for real-life datasets), etc. I normally sign up using the trial accounts and download everything before it expires. But knowing myself, I'm also pretty lazy, else i won't be spending so much time on 6CM....lol...so i'm thinking of signing up for a professional 4-days full day course (classroom) that boasts of guaranteeing a job after completion, within 3 months.

Harvard Business School and Cornell offers an online e-learning course but many people advised me that you don't need those fancy degrees to venture into a data science career. What they're looking, like many other jobs, is for someone to get the job done, regardless of qualifications. I know of a vet (in his 50s) who got a job as a business analyst after taking a short data science course for a few months. So if someone who spends most of his life with animals and still get a job, i reckon anyone can do it too.

Yeah Coursera ones are very good, esp the ones including Python with data science.
 
@dantes can you give me a binny-fied version of this binny-fied work please? Is this nonsense or where it's at? Also are the data science courses a doddle as per the above?
 
I've met a few very religious scientists. One didn't believe in evolution....
As if “belief” has anything to do with it.

More accurately, He should say that he ignores the overwhelming evidence of evolution and instead believes whatever other baseless shite that he wants. It’s fundamentally anti-scientific, and surprising that he can call himself a scientist and keep a straight face.
 
There's a lot of shit data scientists coming out of some of these courses, a lot of them are too short and too easy to get through. So, I'd imagine some of them are a doddle. That said, there are lots of jobs in it, and good data scientists are essential for any decent size business.
 
@dantes can you give me a binny-fied version of this binny-fied work please? Is this nonsense or where it's at? Also are the data science courses a doddle as per the above?

Yes total nonsense. In this case there is some value, they aren't just doing correlation tests on a bunch of random variables and picking the best ones. Rather they are including some fundamental modelling of the actual game, as in the probability of a goal being scored given some prior play. That is not data science, but real science, and rescues these morons from their own stupidity.
 
What utter nonsense this is.

Clearly it’s blood doping, referees being biased and a VAR conspiracy.
 
There's a lot of shit data scientists coming out of some of these courses, a lot of them are too short and too easy to get through. So, I'd imagine some of them are a doddle. That said, there are lots of jobs in it, and good data scientists are essential for any decent size business.
I know what you mean - fucking short scientists!
 
Every top club is doing this and data science is gaining momentum in all work practices.

Government is usually a very late adopter to any new ways of working and even in that field I’m seeing most entities create data science teams.

Certainly a good career choice to future proof yourself.

I’d rather we didn’t share any of our ways of working though so pubically.

We seem to be doing it better than most atm
 
Impressive but not surprising given Henry’s background and those in the US saw the Red Sox turn to analytics in the early/mid 2000’s.
 
I've met a few very religious scientists. One didn't believe in evolution but was also an expert in animal nutrition...

He or she is very much an outlier in respect of evolution. I've met loads, have got to know some very well (two of them were my parents and another was my first serious girlfriend) and have never met one yet who shared that view of evolution.
 
He's very much an outlier in respect of evolution. I've met loads, have got to know some very well (two of them were my parents and another was my first serious girlfriend) and have never met one yet who shared that view of evolution.
It was one of Ian Paisley's lot, the Free Presbyterians...
 
For no particular reason that reminds me of a poster I once saw during my student days. The poster said "Free Nelson Mandela", underneath which someone had scrawled "...with every packet of Bold Automatic". 😀
 
Status
Not open for further replies.
Back
Top Bottom