Football Hackers: The Science and Art of a Data Revolution

[Olé]

Took the opportunity of the international break to read the above book, published in Germany last year and a revised English language version came out this year.

It's an incredible read that brings together all the background and trends in football analytics including key prior books and with practical team/player examples.

Jens Hegeler's ideas are a key thread to the book with decent coverage of "packing" and it has the best history and explanation of xG ratings I've read (hat tip @ExiledAjax).

It also explores the work of the Brentford owner to make a business out of beating the bookies, among other things pointing out prior results aren't very relevant.

The book is full of things that'll make you change the way you look at the game - it dispels a lot of traditional metrics as noise and then introduces more relevant ones. 

A section on cognitive bias is as brilliant as it is familiar to all of us on a football forum, I will feel a lot more guilty for recalling key events as a basis for away report/ratings.

Anyway highly recommend (it is almost certainly @Davefevs dream book) and full of fascinating examples - incidentally see excerpt below which may sound familiar!

I did like this comparison - does the second highlighted example after Pep remind you of anyone not long ago? Perhaps the data helped save him when the results wouldn't.

[Maesknoll Red]

Thanks for the heads up Rob and the highlighted section in your post would suggest the author was a long time reader of this forum.....

[handsofclay]

I had already read the first 5% of this book on kindle that then persuaded me to purchase it at £1.99. In that 5% I found things of interest such as how percentages could be used to get a better picture of how a team is doing. The example to start with was the penalty. In the History of the Premier League 76% of penalties have been scored. So, if a team has missed three penalties in a game that is not a true reflection of their performance as instead of having a score of nil they should have, on percentages, scored two goals. Every spot on the field where goals have been scored from is then given a percentage based on how many goals have been scored in PL history from how many attempts. Thus, a shot from 35 yards might have a 5% chance of a goal, whereas a shot in open play from 14 yards in front of goal might have a 50% chance.

Thus, in a simple example, Derby could have 10 shots from 14 yards and score 1 while their opponents, QPR, have 3 shots from 35 yards and score 2. The Rangers fans are happy with their manager and how the team are playing although admitting that it took 2 wonder strikes to win but it was deserved because of their dogged defensive performance, while the County fans are really peed off with their manager and players. Yet, had the game, as it should've done, followed the percentages then Derby would've won 5-0. Had they done so their fans would've been ecstatic and the Rangers fans calling for the manager's head.

That's what I gleaned from just the first 5% of the book. Once I've read the other 95% I can see LJ installing me as his guru! (Ive got the latest Shardlake eBook and a Michael Palin history of a ship called Erebus to read first.)

 

[Davefevs]

One for the Xmas list from the kids!

[spudski]

It's a great book...and gives a greater understanding to the modern game.

Regarding @Olé quote regarding Pep...TBF to Pep he has consistently stated that it's not so much his tactics, but the quality of the players he has at his disposal that win games.

However, it's true about managers trying to employ tactics that can't be accomplished well by the players they have. It's a fine line between seeing what will improve them and what's a step too far.

I think many a top player who becomes a coach at a lower league team, has become guilty of that. Trying to get lesser quality players, playing the same way as they did.

Glen Hoddle was famous for it, to his dismay.

[elhombrecito]

Currently £1.99 for the Kindle version. Can't go wrong... 

https://www.amazon.co.uk/Football-Hackers-Science-Data-Revolution-ebook/dp/B07NQM3YGK/ref=sr_1_1?crid=3EHA31WT97QJ9&keywords=football+hackers&qid=1571223416&sprefix=football+hackers%2Caps%2C144&sr=8-1

 

[swanker]

33 minutes ago, Davefevs said:

One for the Xmas list from the kids!

And my mother in-law, she just gone into a nursing home suffering from dementia. I’ll tell her Hegelelar who used to play for city had a mention in it, she’ll like that!

[Port Said Red]

1 hour ago, spudski said:

It's a great book...and gives a greater understanding to the modern game.

Regarding @Olé quote regarding Pep...TBF to Pep he has consistently stated that it's not so much his tactics, but the quality of the players he has at his disposal that win games.

However, it's true about managers trying to employ tactics that can't be accomplished well by the players they have. It's a fine line between seeing what will improve them and what's a step too far.

I think many a top player who becomes a coach at a lower league team, has become guilty of that. Trying to get lesser quality players, playing the same way as they did.

Glen Hoddle was famous for it, to his dismay.

I have often thought that Pep's greatest quality must be his man management skills, there must be some huge egos in the kind of dressing rooms he has managed, yet you rarely read of players being unhappy with too little or too much game time etc.

[ExiledAjax]

2 hours ago, Olé said:

Took the opportunity of the international break to read the above book, published in Germany last year and a revised English language version came out this year.

It's an incredible read that brings together all the background and trends in football analytics including key prior books and with practical team/player examples.

Jens Hegeler's ideas are a key thread to the book with decent coverage of "packing" and it has the best history and explanation of xG ratings I've read (hat tip @ExiledAjax).

It also explores the work of the Brentford owner to make a business out of beating the bookies, among other things pointing out prior results aren't very relevant.

The book is full of things that'll make you change the way you look at the game - it dispels a lot of traditional metrics as noise and then introduces more relevant ones. 

A section on cognitive bias is as brilliant as it is familiar to all of us on a football forum, I will feel a lot more guilty for recalling key events as a basis for away report/ratings.

Anyway highly recommend (it is almost certainly @Davefevs dream book) and full of fascinating examples - incidentally see excerpt below which may sound familiar!

I did like this comparison - does the second highlighted example after Pep remind you of anyone not long ago? Perhaps the data helped save him when the results wouldn't.

This is the next one on my reading list! Just finishing Michael Cox's The Mixer - glad to hear it's good!

[petehinton]

Anyone who loves the kind of stuff, would really REALLY recommend the below podcasts too, if you don’t already listen:

Tifo football

Football analysis podcast 

Statsbomb

Training ground guru

All amazing and give really in depth analysis of club structures, certain games etc. 

The twitter accounts of the above, as well as Coaches Voice, Footballfactman all great too.  

[Cowshed]

7 minutes ago, Port Said Red said:

I have often thought that Pep's greatest quality must be his man management skills, there must be some huge egos in the kind of dressing rooms he has managed, yet you rarely read of players being unhappy with too little or too much game time etc.

I have read the book, what it tends to skim over is that Mr Guardiola improves teams. Guardiolas Barcelona may be the best in history. They improved under his leadership. Bayern? Maybe not. Man City? Certainly. 

Results do follow money and that can be quantified (80% +) but what about what the Coach can affect. It is hard to measure the Coaches ability to affect players psychologically. Pep Guardiola is quick to get rid of players he cannot get to buy into his football. Tactics are psychological. If you think what the coach wants is poor it will affect performance. If you think the Coaches ideas are exciting, a challenge to be met and be part of = It will almost certainly result in positive outcomes.

[Port Said Red]

3 minutes ago, Cowshed said:

I have read the book, what it tends to skim over is that Mr Guardiola improves teams. Guardiolas Barcelona may be the best in history. They improved under his leadership. Bayern? Maybe not. Man City? Certainly. 

Results do follow money and that can be quantified (80% +) but what about what the Coach can affect. It is hard to measure the Coaches ability to affect players psychologically. Pep Guardiola is quick to get rid of players he cannot get to buy into his football. Tactics are psychological. If you think what the coach wants is poor it will affect performance. If you think the Coaches ideas are exciting, a challenge to be met and be part of = It will almost certainly result in positive outcomes.

So many aspects of all sports are pyschological, I was thinking the other day when Kipchoge broke the 2 hour barrier, I wonder how many will follow suit now? I am not quite old enough to remember it but I remember reading that once Roger Bannister broke the seemingly impossible 4 minute mile in May 1964, that six people had also achieved it by the end of the year, John Landy ran it in 3.58 only a month after Bannister despite trying for 5 years previously. I wish fans would consider the psychological effect on players they can have.

[spudski]

40 minutes ago, Port Said Red said:

I have often thought that Pep's greatest quality must be his man management skills, there must be some huge egos in the kind of dressing rooms he has managed, yet you rarely read of players being unhappy with too little or too much game time etc.

 

 

If you get a chance to watch the Man City Documentary 'All or Nothing', I can highly recommend. It shows you how he works over a season.

[mozo]

1 hour ago, Cowshed said:

I have read the book, what it tends to skim over is that Mr Guardiola improves teams. Guardiolas Barcelona may be the best in history. They improved under his leadership. Bayern? Maybe not. Man City? Certainly. 

Results do follow money and that can be quantified (80% +) but what about what the Coach can affect. It is hard to measure the Coaches ability to affect players psychologically. Pep Guardiola is quick to get rid of players he cannot get to buy into his football. Tactics are psychological. If you think what the coach wants is poor it will affect performance. If you think the Coaches ideas are exciting, a challenge to be met and be part of = It will almost certainly result in positive outcomes.

Raheem Sterling is one example of the influence Guardiola and his team can have on players. 

[PHILINFRANCE]

1 hour ago, mozo said:

Raheem Sterling is one example of the influence Guardiola and his team can have on players. 

I was going to mention Raheem Sterling on the England vs Bulgaria thread.

It seems only months ago that many on here, and I was certainly one of them, were deriding him for his poor finishing ability, epitomised by the 2018 World Cup.

What has happened since? A revelation, no less.

Is that really due to Pep?

[Davefevs]

Drifting away from the OP, interesting tweet tonight:

 

[Kodjias Wrist]

Sweet ill be buying this for myself and a couple of others.

[swampy]

Results do follow money and that can be quantified (80% +) but what about what the Coach can affect. It is hard to measure the Coaches ability to affect players psychologically. Pep Guardiola is quick to get rid of players he cannot get to buy into his football. Tactics are psychological. If you think what the coach wants is poor it will affect performance

This is so true as Ray Savino once quantified .. 

[Olé]

From the debate it stirred I'm regretting including that Guardiola analogy - the book is not of course so stupid to imply that he isn't responsible for the success that he produces, it is just illustrating a rather neat point that there is a lot of other data required to be able to truly understand and appraise a football team, in just the same way that in the second half of the analogy, actual results aren't always the best guide to an expansive coach at a non-elite team (the bit I thought reminded me of early tenure LJ).

It's just the style of the book to give you reasons to question the traditional measures, and not to rely on results alone - plenty of other data would back up how influential Guardiola himself is - but using metrics like results the book shows is then exacerbated further by the cognitive "confirmation" bias that follows, ie fans and media build a narrative to justify their interpretation (great example given of reaction to Pardew's Newcastle finishing 5th vs relegation battle - with the exact same xG).

It's when the book explains how confirmation bias of this kind also affects our interpretation of a single match we've watched in a way full data does not back up - ie we base our ratings/views disproportionately on a memorable miss or one brilliant tackle, that is probably required reading for any football fan! It goes onto unpick cliches on luck - pointing out that football unlike most other sports, is very susceptible to luck and coincidence due to the low number of scoring events - really revealing stuff.

[Cowshed]

5 hours ago, Olé said:

From the debate it stirred I'm regretting including that Guardiola analogy - the book is not of course so stupid to imply that he isn't responsible for the success that he produces, it is just illustrating a rather neat point that there is a lot of other data required to be able to truly understand and appraise a football team, in just the same way that in the second half of the analogy, actual results aren't always the best guide to an expansive coach at a non-elite team (the bit I thought reminded me of early tenure LJ).

It's just the style of the book to give you reasons to question the traditional measures, and not to rely on results alone - plenty of other data would back up how influential Guardiola himself is - but using metrics like results the book shows is then exacerbated further by the cognitive "confirmation" bias that follows, ie fans and media build a narrative to justify their interpretation (great example given of reaction to Pardew's Newcastle finishing 5th vs relegation battle - with the exact same xG).

It's when the book explains how confirmation bias of this kind also affects our interpretation of a single match we've watched in a way full data does not back up - ie we base our ratings/views disproportionately on a memorable miss or one brilliant tackle, that is probably required reading for any football fan! It goes onto unpick cliches on luck - pointing out that football unlike most other sports, is very susceptible to luck and coincidence due to the low number of scoring events - really revealing stuff.

An interesting point about luck is that there is differing opinion of what luck is between coaches and the writers. Many coaches feel there is very little luck in football. Good teams do not get luckier(cliché), they are better prepared and have players that can problem solve more easily due to their skill sets - Virtually everything is controllable (it could and can be affected) v uncontrollable. The bald bloke up in Manchester feels by focussing on his five elements of football and creating governing principles for its variables luck only then exists in rare extremes .. 

The effects of confirmation bias is now being introduced to coaching. My coaching mentor uses a phrase do you know what you know. Most of us are biased. Our experiences and environment forms our bias. Britain's physical football affects, or used to our mindset and bias. Great value was placed upon crossing and tackling because they are traditional values of hard work. Crossing is widely ineffectual, tackling frequently is a symptom of a team lacking ability … But fans will roar and Mangers will frequently want more of that hard work = Its good when the reality can be different.. 

As humans we are also hard wired to remember negatives and the positives. It is natural. Norms are not so easily easily remembered. It will cloud judgement.

Data helps to reframe both the above.

Along with the Mixer I would recommend people read Football hackers. Read both.

[spudski]

5 hours ago, Olé said:

From the debate it stirred I'm regretting including that Guardiola analogy - the book is not of course so stupid to imply that he isn't responsible for the success that he produces, it is just illustrating a rather neat point that there is a lot of other data required to be able to truly understand and appraise a football team, in just the same way that in the second half of the analogy, actual results aren't always the best guide to an expansive coach at a non-elite team (the bit I thought reminded me of early tenure LJ).

It's just the style of the book to give you reasons to question the traditional measures, and not to rely on results alone - plenty of other data would back up how influential Guardiola himself is - but using metrics like results the book shows is then exacerbated further by the cognitive "confirmation" bias that follows, ie fans and media build a narrative to justify their interpretation (great example given of reaction to Pardew's Newcastle finishing 5th vs relegation battle - with the exact same xG).

It's when the book explains how confirmation bias of this kind also affects our interpretation of a single match we've watched in a way full data does not back up - ie we base our ratings/views disproportionately on a memorable miss or one brilliant tackle, that is probably required reading for any football fan! It goes onto unpick cliches on luck - pointing out that football unlike most other sports, is very susceptible to luck and coincidence due to the low number of scoring events - really revealing stuff.

Love the description in your last paragraph mate. It's spot on.

For example...so many fans will remember how great a pass or cross the likes of Palmer or Eliasson make, but would fail to notice anything, say Tommy Rowe does.

The stats would show the passes, tracking, pressing, interceptions, movement, marking etc, etc that largely go unnoticed, but are essential in doing well as a team.

The fan well remember a great pass, or shot or cross from a player, and that may cloud their judgement as to how well that player actually performed during the rest of the game. Did he do the basics well, that largely go unnoticed.

I'll use our Club as an example...a large amount of fans love Palmer and Eliasson for having an eye for a pass or cross...but some fail to notice that other parts of their game are largely inferior at this level. 

That's why LJ rarely starts Eliasson as an example. 

Football has got to a standard where you just can't carry a flair player that can't do the graft work as well.

[Davefevs]

13 minutes ago, spudski said:

Love the description in your last paragraph mate. It's spot on.

For example...so many fans will remember how great a pass or cross the likes of Palmer or Eliasson make, but would fail to notice anything, say Tommy Rowe does.

The stats would show the passes, tracking, pressing, interceptions, movement, marking etc, etc that largely go unnoticed, but are essential in doing well as a team.

The fan well remember a great pass, or shot or cross from a player, and that may cloud their judgement as to how well that player actually performed during the rest of the game. Did he do the basics well, that largely go unnoticed.

I'll use our Club as an example...a large amount of fans love Palmer and Eliasson for having an eye for a pass or cross...but some fail to notice that other parts of their game are largely inferior at this level. 

That's why LJ rarely starts Eliasson as an example. 

Football has got to a standard where you just can't carry a flair player that can't do the graft work as well.

And that’s why it’s always useful to use both eyes and data. 

[spudski]

1 minute ago, Davefevs said:

And that’s why it’s always useful to use both eyes and data. 

Exactly mate.

[ExiledAjax]

6 hours ago, Cowshed said:

Along with the Mixer I would recommend people read Football hackers. Read both.

I'm just finishing the Mixer. Great book that brought back lots of memories. For me the most interesting parts were the early discussions around the changes that happened when the back pass rule was introduced int the early 90's - I was far too young to appreciate those changes at the time so to read about them was enlightening.  In particular it was interesting to read about that change just as we witness a second change in goalkeeping rules with the introduction of the short goal kicks this season. I suspect that may not have quite as seismic an effect in the long term but we've already seen a few teams struggling and getting caught out when trying to play out from the very very back.

Then the chapter on Leicester was amazing, honestly some of the best non-fiction I've ever read. To relive that season, but with a focus on actually how they won- was brilliant. To compliment what you say about luck; that title win wasn't luck. Fortune played a role in that other teams had bad seasons and underestimated Leicester at certain times - but that is fortune not luck - fortune favours the bold and Leicester were bold and thus got what they deserved.

Anyone read Zonal Marking yet? Waiting for it to hit paperback but would be interested to hear how it stacks up against The Mixer.

[ExiledAjax]

6 hours ago, spudski said:

For example...so many fans will remember how great a pass or cross the likes of Palmer or Eliasson make, but would fail to notice anything, say Tommy Rowe does.

 

Tommy Rowe is class. I saw him in one of the Tampa games and picked him out as the one guy that looked in control of what he was doing. Everything had purpose, he knew exactly what he wanted to do, what he wanted the opponent to do, and what he wanted his teammates to do. He communicated all of that to his teammates and was simply class. I cannot wait for when Dasilva is back fit and we can move Rowe centrally. He won't play every game but I think in tough away games against top opposition he will be invaluable.

[elhombrecito]

14 minutes ago, ExiledAjax said:

Tommy Rowe is class. I saw him in one of the Tampa games and picked him out as the one guy that looked in control of what he was doing. Everything had purpose, he knew exactly what he wanted to do, what he wanted the opponent to do, and what he wanted his teammates to do. He communicated all of that to his teammates and was simply class. I cannot wait for when Dasilva is back fit and we can move Rowe centrally. He won't play every game but I think in tough away games against top opposition he will be invaluable.

I'd forgotten that he is actually a central midfielder! 

[petehinton]

2 hours ago, ExiledAjax said:

I'm just finishing the Mixer. Great book that brought back lots of memories. For me the most interesting parts were the early discussions around the changes that happened when the back pass rule was introduced int the early 90's - I was far too young to appreciate those changes at the time so to read about them was enlightening.  In particular it was interesting to read about that change just as we witness a second change in goalkeeping rules with the introduction of the short goal kicks this season. I suspect that may not have quite as seismic an effect in the long term but we've already seen a few teams struggling and getting caught out when trying to play out from the very very back.

Then the chapter on Leicester was amazing, honestly some of the best non-fiction I've ever read. To relive that season, but with a focus on actually how they won- was brilliant. To compliment what you say about luck; that title win wasn't luck. Fortune played a role in that other teams had bad seasons and underestimated Leicester at certain times - but that is fortune not luck - fortune favours the bold and Leicester were bold and thus got what they deserved.

Anyone read Zonal Marking yet? Waiting for it to hit paperback but would be interested to hear how it stacks up against The Mixer.

Zonal Marking is ******* great.

Still think my personal favourite is either The Numbers Game, or Living on a Volcano

[spudski]

3 hours ago, ExiledAjax said:

I'm just finishing the Mixer. Great book that brought back lots of memories. For me the most interesting parts were the early discussions around the changes that happened when the back pass rule was introduced int the early 90's - I was far too young to appreciate those changes at the time so to read about them was enlightening.  In particular it was interesting to read about that change just as we witness a second change in goalkeeping rules with the introduction of the short goal kicks this season. I suspect that may not have quite as seismic an effect in the long term but we've already seen a few teams struggling and getting caught out when trying to play out from the very very back.

Then the chapter on Leicester was amazing, honestly some of the best non-fiction I've ever read. To relive that season, but with a focus on actually how they won- was brilliant. To compliment what you say about luck; that title win wasn't luck. Fortune played a role in that other teams had bad seasons and underestimated Leicester at certain times - but that is fortune not luck - fortune favours the bold and Leicester were bold and thus got what they deserved.

Anyone read Zonal Marking yet? Waiting for it to hit paperback but would be interested to hear how it stacks up against The Mixer.

I've noticed the new rule with goalkicks has really changed the game.

It's stretched the pitch even further, and teams are drawing players in, creating lots more space in midfield.

In all my years I've never witnessed football being played in such an expansive manner. 

Teams are stretched width wise, and length wise.

It wasn't long ago everything was congested in small areas of the pitch.

The game is constantly evolving.

Fitness and greater discipline and understanding, with better technique is allowing this evolution.

[Guest]

I'm not a stats man at all. I don't actually know how to use them as I know they don't always tell the true story.

I do think number of sprints in a game is useful for a manager to see how much a player is putting into a game.

And it can be handy for coaches I think to search for a player. We looked at Eliasson because of his number of crosses he statistically manages in a game. Then I guess you look at why when watching the player.

But if someone says to me Massengo had a bad game because of some stat when I've seen him play superbly well, I'm not really interested in it. Because there can be so many reasons for a stat to be lower than expected.

A ball down the channel to relieve pressure is statistically bad, but for the team it can be good. And an unsuccessful pass can be down to the person on the end of it more than the actual passer. It's all so complicated. 

There's no factual way of determining how well a player played. It's all opinion. 

As for that XG thing, don't get me started on that!

 

[ExiledAjax]

2 minutes ago, JonDolman said:

I'm not a stats man at all. I don't actually know how to use them as I know they don't always tell the true story.

 

Haha, I think admitting this is the first sign that you do know how to use stats!