Liverpool Football Club’s reliance on analytics(nytimes.com)
nytimes.com
Liverpool Football Club’s reliance on analytics
https://www.nytimes.com/2019/05/22/magazine/soccer-data-liverpool.html
58 comments
This season makes all these decisions look better in hindsight; but as mentioned in the article it didn't work at first. Liverpool tried to Moneyball their way into Top 4 and failed often. Classic examples being when they needed strength in positions like DM, CB or FB. They ended up going for players which made sense stat/pound. It led to LFC signing up players like Sakho or Moreno who had good stats for their value at that time but never worked out. Like in case of Sakho, he doesn't have the presence and leadership needed in the back and LFC needed it badly at that time.
As a Liverpool supporter of almost 11 years (beginning from one of the clubs lowest points), I find this is absolutely fascinating. Curious to know how the research team was involved in the signing of Andy Robertson – perhaps the best value signing in recent memory.
Anyone who's been watch European football for a decade knows how much disparity exists between teams that can afford data science and teams that can't.
Sports are filled with numbers and any teams/players who can afford it has such a leg up, its not even a joke
Sports are filled with numbers and any teams/players who can afford it has such a leg up, its not even a joke
> Sports are filled with numbers and any teams/players who can afford it has such a leg up, its not even a joke
I remain unconvinced. There are a lot of numbers and anecdotes thrown out. But how much is confirmation bias? And does it withstand robust hypothesis testing?
The A's have been in Oakland since 1968. They've been to the World Series 6 times and won 4. None of those World Series appearances occurred during the tenure of Billy Beane as GM, 1997 to the present day.
I remain unconvinced. There are a lot of numbers and anecdotes thrown out. But how much is confirmation bias? And does it withstand robust hypothesis testing?
The A's have been in Oakland since 1968. They've been to the World Series 6 times and won 4. None of those World Series appearances occurred during the tenure of Billy Beane as GM, 1997 to the present day.
There's still a lot to be said about locker room cohesion. Just look at any of the teams managed by Jose Mourinho. His teams quit playing because of his management style. (I still think the players are wrong from this.) Once he's gone, that same crap performing squad suddenly start winning. Personalities are a huge thing in sports. The English national team is an example of having an enormous amount of talent on paper, but just didn't play well together as a team. (Until this most recent squad at the World Cup).
I think part of that is "Moneyball" has tended be more successful over the course of a 162 game season rather than the 5 or 7 game series of the playoffs where there can be a lot more variation in outcomes.
I also think the analytics quickly caught on across the league so the A's advantage was short lived. Now pretty much every team is very heavily invested in it.
I also think the analytics quickly caught on across the league so the A's advantage was short lived. Now pretty much every team is very heavily invested in it.
I haven’t seen any reference to this. It seems to me the teams who dominate simply have the money to splurge on players (transfer fees and wages) facilities and managers. That’s it. Sometimes there’s an anomalous team who succeed without vast wealth (Leicester) or fail despite it (Tottenham) ... but a look at the teams taking part in the latter stages of the Champions League in the last decade doesn’t suggest a group of teams shrewdly applying their analytics teams ... but just those fat with cash who can pay their way to success.
But that's exactly my point. Teams with money buy the best players and managers.
But what is the definition of best? It used to be scouting and feelings. Today, it's almost down to great data science.
That's the reason why Jurgen Klopp gets $75M to buy Virgil Van Dijk and $60M for Allison. The data homework was done and they are almost guaranteed to succeed in Klopps setup.
Pogba probably doesn't fit in Mourinho style teams and that's bad data science setting a team back for a few seasons
But what is the definition of best? It used to be scouting and feelings. Today, it's almost down to great data science.
That's the reason why Jurgen Klopp gets $75M to buy Virgil Van Dijk and $60M for Allison. The data homework was done and they are almost guaranteed to succeed in Klopps setup.
Pogba probably doesn't fit in Mourinho style teams and that's bad data science setting a team back for a few seasons
I don't think Pogba's problem is style. During and even after Mourinho's reign most of the time Pogba looks like someone who couldn't be arsed to play for MUFC. He has been extremely lazy in tracking back, goes missing in games.
There's much more at play than just analytics. Formations, club culture and coaching styles factor heavily into player performance.
Manchester United for example have spent VERY heavily on new players since 2013, and the ROI has been awful almost across the board. They acquired Alexis Sanchez in 2017, who was joint top scorer in the 2016/17 season with 24 goals. At United, he has scored only 5 goals in 18 months.
Chelsea FC are scoring one duck after another by signing Morata, Giroud and Higuain, all of whom have flopped as strikers despite doing well in their previous teams.
Arsenal are an example of a team that uses analytics to suss out bargains in the transfer market. And yet, two of their most expensive analytics-driven signings, defender Shokrodan Mustafi and Granit Xhaka, have fared poorly while cheaper buys like Sokratis Papasthopulous have thrived.
Manchester United for example have spent VERY heavily on new players since 2013, and the ROI has been awful almost across the board. They acquired Alexis Sanchez in 2017, who was joint top scorer in the 2016/17 season with 24 goals. At United, he has scored only 5 goals in 18 months.
Chelsea FC are scoring one duck after another by signing Morata, Giroud and Higuain, all of whom have flopped as strikers despite doing well in their previous teams.
Arsenal are an example of a team that uses analytics to suss out bargains in the transfer market. And yet, two of their most expensive analytics-driven signings, defender Shokrodan Mustafi and Granit Xhaka, have fared poorly while cheaper buys like Sokratis Papasthopulous have thrived.
I don't think you get the scale of analytics. It's not just goals, touches, formations anymore.
It's about entire models of game play style and running simulations with a prospective player to see how they fit in. The data science there is not querieable science. It's more about simulations with numbers.
It's about entire models of game play style and running simulations with a prospective player to see how they fit in. The data science there is not querieable science. It's more about simulations with numbers.
I'm wildly skeptical of this. Do you have anything I could read to better understand it? From where I'm sitting, this isn't Sabermetrics, chiefly because football isn't a series of set pieces with very clearly defined plays and variance of individuals. You sound pretty confident in what you're saying though so I'd be genuinely thrilled to be shown how idiotic my supposition is!
How do you know that Mustafi and Xhaka were data-driven signings?
From a 2016 article,. Quotes from a Premier League agent
"This summer had a familiar feel in many respects as Arsenal faltered in the transfer window (...) yet it concluded with Granit Xhaka, Rob Holding, Skhodran Mustafi and Lucas Perez arriving for a combined fee approaching £90million.
“Four years ago, they bought a company called StatDNA in Chicago,(..)They will take a player and collate data based on every pass, sprint, shot, assist, goal – you name it – he has ever made. These stats are fed into a computer that assesses strengths and weaknesses against every professional in the game before producing an overall value of that player.
Arsenal tend to use statistics to raise questions or confirm observations on possible targets and sometimes unearth players for scouting. "
https://www.standard.co.uk/sport/football/the-deal-arsene-we...
"This summer had a familiar feel in many respects as Arsenal faltered in the transfer window (...) yet it concluded with Granit Xhaka, Rob Holding, Skhodran Mustafi and Lucas Perez arriving for a combined fee approaching £90million.
“Four years ago, they bought a company called StatDNA in Chicago,(..)They will take a player and collate data based on every pass, sprint, shot, assist, goal – you name it – he has ever made. These stats are fed into a computer that assesses strengths and weaknesses against every professional in the game before producing an overall value of that player.
Arsenal tend to use statistics to raise questions or confirm observations on possible targets and sometimes unearth players for scouting. "
https://www.standard.co.uk/sport/football/the-deal-arsene-we...
> The Liverpool Football Club’s reliance on analytics
British sports team names are usually not written with a definite article like that. Maybe someone who knows more about grammar than me knows a proper term to describe why. Is it something to do with British English tending to treat companies, teams, etc as plurals rather than singulars ('Liverpool are' more often than 'Liverpool is')? Anyway, that 'the' shouldn't be there, and it isn't in the article.
British sports team names are usually not written with a definite article like that. Maybe someone who knows more about grammar than me knows a proper term to describe why. Is it something to do with British English tending to treat companies, teams, etc as plurals rather than singulars ('Liverpool are' more often than 'Liverpool is')? Anyway, that 'the' shouldn't be there, and it isn't in the article.
I loved the line from "The Good Shepard" when they ask why the article 'the' is never used when discussing CIA.
Richard Hayes: I remember a Senator once asked me 'when we talk about CIA why we never use the word the in front of it.' And I asked him 'do you put the word the in front of God? [0]
[0]https://en.wikiquote.org/wiki/The_Good_Shepherd_(film)
Richard Hayes: I remember a Senator once asked me 'when we talk about CIA why we never use the word the in front of it.' And I asked him 'do you put the word the in front of God? [0]
[0]https://en.wikiquote.org/wiki/The_Good_Shepherd_(film)
Is it something to do with British English tending to treat companies, teams, etc as plurals rather than singulars
Yes. Far more information than is perhaps healthy on this topic: http://itre.cis.upenn.edu/~myl/languagelog/archives/001874.h...
Yes. Far more information than is perhaps healthy on this topic: http://itre.cis.upenn.edu/~myl/languagelog/archives/001874.h...
Meta: it's a weird use of "The" in the title, it's about "Liverpool Football Club" and the definite-article is unnecessary. Indeed the "The Liverpool Football Club" makes it wrong as Liverpool is one of two [major league] Liverpool based football clubs [Liverpool, Everton], at least.
Being a pedant is annoying sometimes.
Being a pedant is annoying sometimes.
In the article itself, Liverpool are correctly referenced throughout, without "The".
I think this is a US/British English distinction, since saying "The New York Mets" or "The Dallas Cowboys" is the norm, but never "The Manchester United" or "The Rangers" (despite being a plural).
Come to think of it, does any NFL team not end in a plural?
Come to think of it, does any NFL team not end in a plural?
My favorite sports trivia question:
Name the 7 NCAA FBS (Div 1 college football) teams whose name is not a plural and also does not contain a color (i.e. Crimson Tide doesn't count).
Hint: Two of them are the same team name at different schools
Name the 7 NCAA FBS (Div 1 college football) teams whose name is not a plural and also does not contain a color (i.e. Crimson Tide doesn't count).
Hint: Two of them are the same team name at different schools
Stanford Cardinal, Norte Dame Fighting Irish, Illinois Illini, Syracuse Orange, the two Wolf Pack (NC State and Nevada), and I cannot think of a seventh school.
If I can’t use colors then I guess Syracuse doesn’t count.
If I can’t use colors then I guess Syracuse doesn’t count.
>Stanford Cardinal
Stanford Cardinal is also a color.
Stanford Cardinal is also a color.
William and Mary Tribe
You got 4
Answer: Fighting Irish, Fighting Illini, Wolfpack x2, Midshipmen, Minutemen, Thundering Herd.
Arsenal F.C. on the other hand are commonly referred to as the Arsenal.
My company's data analyst got his degree in soccer data analysis (and is a Liverpool fan!) It's been an interesting sector to learn about and ripe to learn from, too, as they seem to operate rather separately from what we might call "data science" here. Sport tech seems to be ripe for learning from and to cross over into more traditional tech scenarios.
Where did he study for that?
Soccer would be a much better sport if it didn't have such restrictive substitution rules. Freer substitution would open up a new dimension to the game and give coaches and data analysts more room to strategize both during the game and when hiring players.
It would then become American football where offense and defense get swapped out as soon as there's a possession change. No thank you. I like soccer just the way it is now.
Hockey manages to keep the action going while substituting half of the team at once. I think soccer could manage a couple guys at a time.
Rugby used to be the same with 3 subs, but these days you can pretty much sub half the team. It really destroys finding value in superior fitness, tactical planning of the starting side etc.
Freer substitutions would ruin the game. I believe that restricting substitutions to just 3 forces the managers to strategize their gameplan (especially in the event of a reduced team size due to player(s) getting Red Card) and in turn make the game more interesting.
> Soccer would be a much better sport
Very subjective. I personally think having a limited set of substitutions helps keep the game more interesting as every change can be hugely impactful to the team's tactics and the result of the game.
Very subjective. I personally think having a limited set of substitutions helps keep the game more interesting as every change can be hugely impactful to the team's tactics and the result of the game.
Obviously subjective but my fear is that it would become more mechanical and dominated by physical capacity and highly specialized players. Some people may see that as a good thing. Personally I like seeing how versatile players and smart coach deal with the constraints. On the other hand, I think this freedom works well for basketball and even futsal.
Are there any good soccer data collections (free / reasonably priced) for an individual programmer to build something like this?
Detailed per play/player stats involve quite a bit of effort to collect, and thus are not typically free. Sports stats in general is a fairly closely guarded and proprietary area. Here an interesting article about it from 2014:
https://fivethirtyeight.com/features/the-people-tracking-eve...
I have to assume computer vision will largely automate the process of collecting sports stats, even in real-time, if that hasn't already happened. For that I think you would need access (and licensing rights) to a quality video feed of the matches, different than what is broadcast for human consumption (i.e. shows the entire field of play the whole time).
I have to assume computer vision will largely automate the process of collecting sports stats, even in real-time, if that hasn't already happened. For that I think you would need access (and licensing rights) to a quality video feed of the matches, different than what is broadcast for human consumption (i.e. shows the entire field of play the whole time).
It does exist: https://www.hudl.com/
Lots of your local and state tax money for high school / university sports programs gets dumped into it
Lots of your local and state tax money for high school / university sports programs gets dumped into it
Why not aggregate the uploaded videos from hundreds of fans?
Sure it would very noisy, but synchronised feeds from many angles seems valuable.
Sure it would very noisy, but synchronised feeds from many angles seems valuable.
If I recall correctly, Opta is one (https://www.optasports.com/), but I'm not sure if it reasonably priced.
So I didn't read the article but 25 years ago I was part of a group of 100 15-17 year olds in the US recruited by Liverpool to see if we were good enough to go over seas to their academy to finish up school. So I've been following them ever since.
What I remember reading in their 2nd game against Barcelona is the data team told the coaches about how Barcelona was slow to setup for set pieces mostly corner kicks. The ball boys are all academy players and the coaches told the ball boys on corner kicks to get the ball to the Liverpool player as quickly as possible.
I believe it went as far as them having more ball boys then usually to just have 4 of them sit with a ball in each corner.
What I remember reading in their 2nd game against Barcelona is the data team told the coaches about how Barcelona was slow to setup for set pieces mostly corner kicks. The ball boys are all academy players and the coaches told the ball boys on corner kicks to get the ball to the Liverpool player as quickly as possible.
I believe it went as far as them having more ball boys then usually to just have 4 of them sit with a ball in each corner.
That's an amazing level of attention-to-detail. Liverpool seem very interested in perfecting every aspect of the game. Another example is that they have the world's only throw-in coach. https://www.bbc.co.uk/sport/football/45405476
This liverpool and Manchester City teams are setting the bar very high for what sports is going to be in the future. There is no way teams would go entire seasons with just one defeat (against each other) without paying attention to every tiny opportunities of success.
Case in point, in preparation of next weeks champions league final, Liverpool are at a camp simulating Madrid weather and pitch surface. Even if they don't win, the knowledge they gain and the precision data they obtain for next season is crazy!
Case in point, in preparation of next weeks champions league final, Liverpool are at a camp simulating Madrid weather and pitch surface. Even if they don't win, the knowledge they gain and the precision data they obtain for next season is crazy!
Given that's exactly how Liverpool just beat Barcelona to reach the Champion's League Final, I believe it.
https://youtu.be/4NhlXGMDNq4?t=250
https://youtu.be/4NhlXGMDNq4?t=250