#7 | Raheem Sterling - 2021/22 Performances

[meeesh]

You were doing so well and i liked it until this bit

And I don’t think any City fan doesn’t rate Sterling.

I could name you several , i wont but i could

I'll help you with one

 

[Kippax77]

You'll have to tell that to every top club in the world who are investing in data experts and analytics applications.

It is however comforting that there appears to be a correlation between not rating Sterling and not appreciating advanced analytics.

What does the data translate for Declan Rice running around in circles in the centre of the pitch for 90 mins non stop say?
That he should be rated?
Stats are very overrated in my opinion

 

[England Fan]

The difference there is they have several experts collecting ALL the data and use all the data in their analysis. The average random twitter account or media personality that isolates one random stat and presents that as proof of something, without tge surrounding context that the analytics departments at clubs will clearly have is utterly pointless. Stats have a use to the club, but the majority of them have next to no use to the fans and the media.

It depends what you are trying to evidence - I was using the analytics to show that people who call Raheem a poor finisher are not basing this on facts and using non-pen xG vs goals to do that is stronger than the "eye test" - though I doubt people who hate Raheem will want to hear that or even engage with the evidence. All that data points like xG do are take simple data points (e.g. shots) from multiple games so we can get a sense of how things are compared to an average (which you cannot do from the "eye test" as your time and memory is limited). Even clubs do not have perfect data sets - there is only about 5 years of xG data and that is even pretty sketchy outside the big 5 leagues so things like league conversion rates will be completely subjective - but it is better than just using the "eye test" as there is only so much football you can watch / remember.

The point further down about Bruno Fernandes being a player that people who use advanced analytics would rate very highly couldn't be further from the truth (unless when you mean stats you just mean raw goals and assists which I would class as descriptive stats rather than predictive). Most analytics fans have been calling Bruno overrated for a number of months given his reliance on penalty goals.

There are also stats to quantify the value of things like the pass before the pass, etc which aren't as readily available as xG data (xG chain, On Ball Value), you probably just won't like or agree with what they say

 

[England Fan]

What does the data translate for Declan Rice running around in circles in the centre of the pitch for 90 mins non stop say?
That he should be rated?
Stats are very overrated in my opinion

Stats aren't perfect (in any sport as there will always be randomness / variance given seasons are a small sample size) but they are better than the "eye test".

You can also do a sense check on most stats, e.g. when comparing non-pen xG to goals scored Messi shows up to be one of the best finishers in the world and the rest of the list broadly makes sense from what I observe with watching football which adds credence to that value of the data.

 

[Kippax77]

Stats aren't perfect (in any sport as there will always be randomness / variance given seasons are a small sample size) but they are better than the "eye test".

You can also do a sense check on most stats, e.g. when comparing non-pen xG to goals scored Messi shows up to be one of the best finishers in the world and the rest of the list broadly makes sense from what I observe with watching football which adds credence to that value of the data.

Welcome to the forum Max Headroom.
Enjoy your stay.

PS West Ham are our favourite "cockney team" on here

 

[England Fan]

Welcome to the forum Max Headroom.
Enjoy your stay.

PS West Ham are our favourite "cockney team" on here

Haha - I had to look that up to understand the reference but much appreciated.

There is a "special relationship" between City and West Ham - probably because we are used to dealing with awful fans from the Man U / Liverpool and Spurs / Chelsea respectively. I have a (potentially irrational) support of Raheem as a player so you can thankfully avoid me if you don't stray into this fred!

 

[Pablo ZZZ Peroni]

It’s why Gundogan and Bernardo play in City’s team rather than us selling them and going for Bruno Fernandes who has better stats than both of them. de Bruyne has always said that stats aren’t important. Pep has said himself that he’d often rather see his players, in the final third, take a chance and it not come off than just keep the ball for the sake of it and to boost stats figures.

When Wenger started going down the “moneyball” stats route, he took Arsenal from challenging for titles to barely finishing in the Top4.

Stats are often insignificant because they only tell a tiny fraction of what football is all about. Some of the most important aspects to this sport don’t have stats. Football is a sport that has situations that might only happen once a game and never again in any game again that season, but are some of the most important things that happen in that game. There could be half a dozen of these individual things that occur that are all different to each other that just occur out of the blue that certain players always seem to be involved in because they are great players and understand the game, but will never attain a stat for them.

You don’t get a stats table for the assist to the assist to the assist to the assist even though it was the most important pass in that goal, or the movement that draws opposing players out of position, or one individual quick one touch pass that sets the tempo for a whole move, or knowing how to make runs to lose a marker, understanding where to run to get on the end of a cross, or where to position yourself when the opposition move the ball into certain areas of the pitch… there are no stats to tell us which players have the best concentration levels, or players who communicate all game and tie everything together with motivation and organisation etc.

But they could all be more important to football than anything you can show me in a stat.

Everyone knows the Agüeroooooo! goal, yet do you remember the most important thing that any player did in that goal? It wasn’t particularly Agüero scoring, it wasn’t Balotelli’s assist, it wasn’t de Jong breaking forward and passing it to Agüero before the 1-2… it was the run that Kompany made to move the defenders and create the space that Agüero eventually ended up in with a free shot at goal… what stat does Kompany attain for that? None at all, yet we don’t score that goal without that run.

Stats are often misleading and should never be used in isolation. A player can find himself doing three good things in three incidents that add up to 15 seconds in a game of 90 minutes and have an absolute stinker for the other 89 minutes and 45 seconds, but leave the field with a goal and two assists. The rest of the time, tripping over the ball, constantly giving it away, running down blind alleys and doing sod all defensively. Also, you can end a game - or a season - having the best passing stats in the game/league, but a player with the most passes and most successful passes might be dropped by Pep in the next game if those passes have been too safe, too many backwards passes, and too many that lacked incision and positivity.

Football isn’t like American sports where stats are king. US sports are played in bursts and are formulaic and robotic, where it’s all about stats. Football flows too much and has too many variables for stats to matter to the general influence on the game.

Stats are usually just the end result to a load of randomness. The person who attains the stats might have been provided with an easy chance to make an assist or score the goal with barely any influence in the randomness beforehand that set that situation up. The randomness comes from football being a free flowing game, stats are just the end product of all that. But what’s more important in football: the end stat, or the randomness to set the stats up? It’s the randomness, because without it there’d be no goal and assist.

Football is mainly about exploiting space. If you can influence the game to create space in the opposition’s defensive set-up, and exploit that space, you will win. There are no stats to tell you how much you exploited space in a game of football.


And I don’t think any City fan doesn’t rate Sterling. He’s been part of the best league side in English football history and is appreciated for it by us all. He’s provided us with some brilliant moments! But he can be a very frustrating watch, between the things you can attain stats for he can be anything from brilliant to abysmal and you never know which one you’re going to get. There have been periods of seasons where he’s been consistently poor and he’s useless to us, there have been times where he’s been consistently poor but got a load of goals and assists, there have been times where he’s been on fire and got no goals or assists or he’s been on fire and bagged a few of each.

That’s why I’ve changed my outlook on him; rather than expect too much, just take the rough with the smooth, and I try to not get frustrated with him. But I have in the past and it’s all worth talking about. But it has to be said, he’s nowhere near as good a-player as many others we’ve had.

The world has moved on since Arsene Wenger

In pro football there has been an explosion in the use of data. Some of it is mind blowing and much of what you refer to in the post above can be captured and is used ie your Kompany example is old hat as far as a data is concerned. As for exploiting space well, yep, they are cracking that with the evolution of tracking data and AI. The 2 best proponents are with City and Liverpool who are probably the most advanced in world football. They are scientists with big big brains. Here is the City guy:

Training Ground Guru | Man City land big signing in quest to be the best in data science

A PhD in computational astrophysics, developing trading systems for a £30bn hedge fund and policy advisor for the British government: not your typical CV for a member of backroom staff. However, these are just some of the credentials of Laurie Shaw, who has been appointed Lead AI Scientist by...

trainingground.guru

City women have a vacancy at the moment for a 1st team performance analyst. The remit follows and note in this role the repeated reference to performance data and video analysis:

ole specific accountabilities:
1	Work closely with the MCWFC Head of Performance Services to evolve data driven ‘What It Takes to Win’ models for the Women’s 1st Team plus a ‘What It Takes to Develop Players’ model for the Development Squad and Regional Talent Centre.
2	Create detailed opposition reports and tactical analyses utilising video and performance data. Work in partnership with other MCWFC performance analysts to provide actionable insights to support optimal preparation for upcoming matches.
3	Provide in-game performance analysis functions including filming, live coding and feedback as required to support in-game decision making.
4	Contribute to timely post-match analysis reports utilising video and performance data to review match performance. Communicate via appropriate methods to support player, coach and practitioner understanding and use of insights
5	Utilise appropriate performance analysis means (clips, data) to support individual player development plans and performance preparation as required.
6	Work in partnership with the MCWFC 1st Team Performance Analysts, coaches, and practitioners to provide training analysis as required.
7	Continually develop an evolving library of analysis (video & data) relating to opposition teams to allow longitudinal analysis of MCWFC and the opposition at a team and individual level.
8	Support the response to ad hoc queries and analysis requests from coaches, players and practitioners as delegated by your line manager
9	Collaborate with the Data Insights & Decision Technology Team (DIDT), coaches and practitioners to evolve data driven methodologies for Talent ID and recruitment.

As for skills and expertise...

Role specific knowledge, skills and experience:

Essential: Undergraduate qualification in Sports Science, Performance Analysis or a related area. Relevant experience in a performance environment as a performance analyst. Experienced using performance analysis related software (SportsCode, Hudl etc) Using performance analysis methods to analyse and understand football tactics Capability to facilitate conversations with video and data visualisation and engage people from a range of backgrounds with novel / actionable insights. Track record of building effective working relationships with coaches, players, and practitioners from a diverse range of backgrounds. Experience in a performance analysis role identifying and communicating actionable insights. Experience of working with sports performance related data either in previous employment or as a hobby. Flexibility and openness to work irregular and unsociable hours outside of ‘normal’ office hours following a senior 1st team schedule (evenings, weekends, and bank holidays)

Desirable: Master’s Qualification in Performance Analysis or a related area Previous experience in professional football Competent using a visual analytics tool such as Tableau, QlikSense or Power BI. ‘UEFA B’ Licence football qualification

 

[Pablo ZZZ Peroni]

Peak Sterling.
Image from an old thread in October 19.
Funny reading the posts around that time... usual praise and criticism on a game to game basis :)

 

[mccity]

Peak Sterling.
Image from an old thread in October 19.
Funny reading the posts around that time... usual praise and criticism on a game to game basis :)

View attachment 35730

Significantly enough he was on the left wing for much of that year.

That's not me saying he's better on the left, more me saying that the wing he operates on isn't totally fundamental to his output on a larger basis.

 

[LongsightM13]

It’s why Gundogan and Bernardo play in City’s team rather than us selling them and going for Bruno Fernandes who has better stats than both of them. de Bruyne has always said that stats aren’t important. Pep has said himself that he’d often rather see his players, in the final third, take a chance and it not come off than just keep the ball for the sake of it and to boost stats figures.

When Wenger started going down the “moneyball” stats route, he took Arsenal from challenging for titles to barely finishing in the Top4.

Stats are often insignificant because they only tell a tiny fraction of what football is all about. Some of the most important aspects to this sport don’t have stats. Football is a sport that has situations that might only happen once a game and never again in any game again that season, but are some of the most important things that happen in that game. There could be half a dozen of these individual things that occur that are all different to each other that just occur out of the blue that certain players always seem to be involved in because they are great players and understand the game, but will never attain a stat for them.

You don’t get a stats table for the assist to the assist to the assist to the assist even though it was the most important pass in that goal, or the movement that draws opposing players out of position, or one individual quick one touch pass that sets the tempo for a whole move, or knowing how to make runs to lose a marker, understanding where to run to get on the end of a cross, or where to position yourself when the opposition move the ball into certain areas of the pitch… there are no stats to tell us which players have the best concentration levels, or players who communicate all game and tie everything together with motivation and organisation etc.

But they could all be more important to football than anything you can show me in a stat.

Everyone knows the Agüeroooooo! goal, yet do you remember the most important thing that any player did in that goal? It wasn’t particularly Agüero scoring, it wasn’t Balotelli’s assist, it wasn’t de Jong breaking forward and passing it to Agüero before the 1-2… it was the run that Kompany made to move the defenders and create the space that Agüero eventually ended up in with a free shot at goal… what stat does Kompany attain for that? None at all, yet we don’t score that goal without that run.

Stats are often misleading and should never be used in isolation. A player can find himself doing three good things in three incidents that add up to 15 seconds in a game of 90 minutes and have an absolute stinker for the other 89 minutes and 45 seconds, but leave the field with a goal and two assists. The rest of the time, tripping over the ball, constantly giving it away, running down blind alleys and doing sod all defensively. Also, you can end a game - or a season - having the best passing stats in the game/league, but a player with the most passes and most successful passes might be dropped by Pep in the next game if those passes have been too safe, too many backwards passes, and too many that lacked incision and positivity.

Football isn’t like American sports where stats are king. US sports are played in bursts and are formulaic and robotic, where it’s all about stats. Football flows too much and has too many variables for stats to matter to the general influence on the game.

Stats are usually just the end result to a load of randomness. The person who attains the stats might have been provided with an easy chance to make an assist or score the goal with barely any influence in the randomness beforehand that set that situation up. The randomness comes from football being a free flowing game, stats are just the end product of all that. But what’s more important in football: the end stat, or the randomness to set the stats up? It’s the randomness, because without it there’d be no goal and assist.

Football is mainly about exploiting space. If you can influence the game to create space in the opposition’s defensive set-up, and exploit that space, you will win. There are no stats to tell you how much you exploited space in a game of football.


And I don’t think any City fan doesn’t rate Sterling. He’s been part of the best league side in English football history and is appreciated for it by us all. He’s provided us with some brilliant moments! But he can be a very frustrating watch, between the things you can attain stats for he can be anything from brilliant to abysmal and you never know which one you’re going to get. There have been periods of seasons where he’s been consistently poor and he’s useless to us, there have been times where he’s been consistently poor but got a load of goals and assists, there have been times where he’s been on fire and got no goals or assists or he’s been on fire and bagged a few of each.

That’s why I’ve changed my outlook on him; rather than expect too much, just take the rough with the smooth, and I try to not get frustrated with him. But I have in the past and it’s all worth talking about. But it has to be said, he’s nowhere near as good a-player as many others we’ve had.

Exactly, my brother.
American sports and even the likes of cricket are mainly maths and science. The stats bores.
Football is art and creative writing, a new story and a blank page or canvas, every minute of every game.
The greatest artists transcend the scientific/mathematical attempts to constrain and quantify their ethereal creativity