whitestreak
SC Supporter
- Dec 8, 2006
- 833
- 3,417
Interesting article about Poch's tactics and our inability to score goals.
http://cartilagefreecaptain.sbnatio...io-pochettino-defensive-manager-diego-simeone
(I've removed all the images)
When Spurs first hired Mauricio Pochettino I argued that this was basically Daniel Levy saying that Andre Villas-Boas’s ideas were right; AVB just wasn’t the one to actually execute those ideas at White Hart Lane. Pochettino was a manager who would again try to develop a distinctive tactical system that allowed Tottenham to punch above its financial weight and compete with richer domestic rivals. That had been the plan under AVB and it would continue to be the plan under Pochettino.
The funny thing in year three of the Poche era, it looks more and more like we did more than just find a manager to, like AVB, develop a distinct tactical system that allows us to compete with wealthier rivals.
We may have actually found a manager who is developing AVB’s tactical system that allows us to compete with our richer domestic rivals. Specifically, we have found a defense-first manager who sees football in essentially attritional terms and sets up his team to grind out 1-0 results.
This is still perhaps a bit counter-intuitive with Pochettino. His reputation in England, as best I can tell, has always been more Guardiola than Mourinho. Certainly his Southampton teams were praised for being entertaining, as was last year’s Tottenham team. Moreover, Guardiola himself has raved about Pochettino on multiple occasions.
But here’s the funny thing: When you take another look at the actual evidence, Poche doesn’t look like a Guardiola disciple, despite the Catalan’s high praise of him. Instead, he looks more like a Mourinho-style defense-first manager with some limited Guardiola influences.
Put another way, Poche looks a whole lot like AVB.
What was AVB’s style?
Here are some numbers on AVB (NOTE: It’s hard to get super great advanced stats going back this far):
Pochettino has always been a defense-first manager.
We’re now in the fifth season of Mauricio Pochettino managing in England. And here’s what we know: Poche is a lot more like AVB or Mourinho than he is Guardiola or Jurgen Klopp.
To demonstrate this, let’s begin by working backward to Pochettino’s Southampton days. Upon arrival in January of 2013, Pochettino inherited a Southampton team that knew how to attack, but... well, defending wasn’t a priority for them.
Here’s how they did in the first 22 games of their first season back in the Premier League played under Nigel Adkins:
Teams that defend like Adkins’ Southampton did don’t last long in England’s top flight.
In 16 games under Pochettino and with the exact same players, those numbers transformed dramatically:
Here’s the impressive thing: The only teams in 2013-14 with better defensive records in England were Jose Mourinho’s Chelsea and eventual champions Manchester City.
When you look back at his normal XI during that season, that defensive record makes more sense. Pochettino played a 4-2-3-1 system that typically looked like this:
Boruc
Shaw, Fonte, Lovren, Clyne
Wanyama, Schneiderlin
Lallana, Davis, Rodriguez
Lambert
A few things should stand out about that lineup:
Indeed, we can push the point further: His record at Southampton showed that Pochettino’s system requires a great deal of running and work, but that this running and work is closer to the work demanded by Diego Simeone than Klopp.
In an interview I did with him at Just Football, Into the Calderon editor Robbie Dunne said that Simeone believes in running and a high work rate because he sees football as being basically attritional in nature. That, of course, sounds an awful lot like both AVB and Pochettino.
How does Pochettino’s style explain results this season?
As at Southampton, Pochettino’s style at Spurs has been to generally play a more conservative midfield two with one player used in an almost exclusively defensive role. The first priority is to be defensively solid and to not concede bad goals. The attack must be built on this platform.
For this reason, Pochettino does not typically attack with numbers in the way that Klopp, Schmidt, Guardiola, or Thomas Tuchel do. Recall the numbers highlighted in this piece by Ricardo Tavares. Amongst the top teams in England last season, Spurs were the only ones whose main passing combinations were between defenders.
Rather than slow, patient possession, there are three main ways that Pochettino teams create chances:
Moreover, once teams get wise to this style and begin sitting back themselves, you typically end up being badly out numbered. Long ball attacks from Spurs often begin with four Tottenham attacking players going up against five or six opposition defenders but against teams like West Bromwich Albion earlier this season it was more like seven, eight, or even nine defenders.
Those numbers begin to change a bit once the fullbacks get forward, but as we saw in the Champions League, if you can keep Spurs’ fullbacks pinned deep, the attack sputters.
What happens when teams eliminate Tottenham’s primary attacking outlets?
This brings us back to AVB. When you play defense first and rely on direct attacks and counters to create scoring chances, it is relatively easy for opponents to say, “Fine. We’ll sit in a low block ourselves and then you can’t create chances.”
The result of this is that Tottenham games against weaker opposition begin to all look more-or-less the same: Spurs have a high amount of possession, but they have no ability to break down a packed defense because their players either don’t know how to open up such a defense or lack the ability to do so.
As a result, they end up taking lots of shots from distance and other lower quality chances. This is, in fact, a common problem with defense-first managers.
Here are a few sample xG maps from just this season for Poche and other defense-first managers:
Simeone’s Atletico earlier this season: 16 shots against Las Palmas, only one xG.
xG map for Atletico Madrid - Las Palmas.
8:34 PM - 17 Dec 2016
Here is Mourinho’s United against Watford earlier this season. 15 shots, 1.3 xG:
xG map for Watford - Manchester United. I've been expecting something like this to happen to United all season.
3:54 PM - 18 Sep 2016
That brings us to this season’s Tottenham. To be honest, the number of examples I could choose from is kind of terrifying. But here’s a small sample.
This is last weekend’s clash with Burnley. 30 shots, 1.7 xG:
xG map for Tottenham - Burnley. Spurs do it with volume. It's not pretty, and it's pretty risky.
Here is Bournemouth almost two months ago. Note that Erik Lamela played in this game so don’t think this is necessarily a “we-miss-Lamela” problem. 16 shots, .7 xG:
xG map for Bournemouth - Tottenham. Excellent defensive performance by Bournemouth, and they were unlucky this ended 11 v 11.
It’s a problem in Europe too. Here is the win against CSKA in Moscow. 23 shots, 1.3 xG
xG map for CSKA - Tottenham.
Another very Poche Spurs performance. Control midfield and take your time to unlock a set defense.
Aggregate xG Data for 2016-17
To make the point a bit more clearly, I went through and reviewed xG maps for all six top English teams so far this season. I took a very, very basic approach to the question: Let’s find out each team’s average xG/match, average shots/match, and average xG value/shot.
This is using only results from this season (and only results that I could find on Caley’s Twitter timeline). So the data is not perfect. Even so, the results are interesting:
Shot Quality Data
Team
xG / Match
Shots / Match
Avg. xG / Shot
Arsenal
1.7125
13.75
0.12
Chelsea
1.67
14.9
0.11
Liverpool
1.7
16.9
0.1
Man City
1.86
15.6
0.12
Man United
1.5
16.19
0.09
Tottenham
1.57
18.9
0.08
As you can see, our xG value per shot is the worst amongst the six sides with Champions League aspirations. Indeed, Tottenham’s xG value of any one shot is, on average, 50% lower than the xG value of Arsenal or City’s average shot.
One way of breaking down this data would be to identify two separate groups. I’m listing the members of each group in order from the most extreme example to the least extreme, though that is admittedly defined in somewhat subjective terms. (I’m looking for a way to do a chart like this one, but am having no luck—if someone can help me out in the comments, I can update the post with a chart.)
Higher Shot Quantity, Lower Shot Quality
At the other end of things, more attack-minded managers like Arsene Wenger and Pep Guardiola will generally produce higher quality chances because their teams are more comfortable attacking packed defenses and are better at opening them up.
Conte’s Chelsea, meanwhile, is sitting in the sweet spot between these two poles. On the one hand, they actually defend better than either Mourinho’s United or Pochettino’s Tottenham. But they do this without sacrificing shot quality: Their xG/shot is almost the same as Arsenal’s and City’s and is actually better than Liverpool’s, which has the most goals in the league so far this season. That is how you win a title.
It’s also worth noting that the team with the best chance at challenging Chelsea is probably Klopp’s Liverpool. They have also found a way to buck the trend by having a high shot quantity number and better shot quality than United or Spurs.
What can Spurs do to fix this?
This is the key question, of course. And, unfortunately, there aren’t easy answers. Conte’s Chelsea seems to have figured out how to retain a robust defensive structure without compromising shot quality, but then Conte’s Chelsea has N’Golo Kante in midfield, Eden Hazard in attack, and an on-fire Diego Costa leading the line.
So while Conte deserves major credit for his transformation of Chelsea, I am not sure how much Chelsea can be a template for other teams since most teams lack Chelsea’s financial muscle and consequent ability to recruit world-class players.
The more likely reality is that this sort of attack is probably more or less what we should expect from Pochettino teams and the best way to improve results is not necessarily by increasing shot quality, which may not be possible without compromising defensive soundness, but instead looking to eliminate mistakes that lead to high-quality chances for the opposition.
This past weekend’s fixture is probably as good a place as any to begin on that point: We weren’t bad, necessarily. In addition to our two goals, Dele Alli wasted a good chance in the opening minutes and came very close to scoring a curler from the edge of the box in the second half. Christian Eriksen also nearly scored from distance but for great work from Burnley keeper Tom Heaton and nearly got on the edge of a Harry Kane cross in the first half. None of these were great chances, but you also wouldn’t have been shocked to see any of those four chances find their way into the net.
What killed us is that we allowed Burnley two extremely good chances in the first half, both of which were totally unnecessary. In both cases, we had multiple chances to clear and we simply failed to do so. The second chance, which is the one that produced Burnley’s goal, also involved a couple lucky deflections for Sean Dyche’s team, but if Kyle Walker clears the ball when he has the chance to do so, those deflections never happen.
To put it another way, we probably shouldn’t expect Pochettino teams to consistently produce more than what this team is currently averaging, which is 1.57 xG/match. Pochettino teams are never going to create chances like an Arsene Wenger or Pep Guardiola team. However, the examples of Mourinho, Conte, and Simeone provide all the proof needed of how effective a defense-first system can be.
Indeed, Simeone’s example is particularly compelling as his Atletico have been competing for La Liga and Champions League titles for most of the past five seasons despite enormous financial disadvantages relative to their main rivals thanks to the fact that they are so absurdly difficult to beat.
To be sure, there are some structural things Spurs can do to improve results. If Victor Wanyama could get better at splitting the center backs that would go a long way. If Pochettino could be more flexible in-game and figure out ways to support his front four when his fullbacks can’t get forward, that would be significant.
But ultimately the attack we’re seeing this season is probably more-or-less what we should expect from Pochettino teams. The big question, then, is not “how can we improve the attack?” but “how can we eliminate unforced errors when defending?”
http://cartilagefreecaptain.sbnatio...io-pochettino-defensive-manager-diego-simeone
(I've removed all the images)
When Spurs first hired Mauricio Pochettino I argued that this was basically Daniel Levy saying that Andre Villas-Boas’s ideas were right; AVB just wasn’t the one to actually execute those ideas at White Hart Lane. Pochettino was a manager who would again try to develop a distinctive tactical system that allowed Tottenham to punch above its financial weight and compete with richer domestic rivals. That had been the plan under AVB and it would continue to be the plan under Pochettino.
The funny thing in year three of the Poche era, it looks more and more like we did more than just find a manager to, like AVB, develop a distinct tactical system that allows us to compete with wealthier rivals.
We may have actually found a manager who is developing AVB’s tactical system that allows us to compete with our richer domestic rivals. Specifically, we have found a defense-first manager who sees football in essentially attritional terms and sets up his team to grind out 1-0 results.
This is still perhaps a bit counter-intuitive with Pochettino. His reputation in England, as best I can tell, has always been more Guardiola than Mourinho. Certainly his Southampton teams were praised for being entertaining, as was last year’s Tottenham team. Moreover, Guardiola himself has raved about Pochettino on multiple occasions.
But here’s the funny thing: When you take another look at the actual evidence, Poche doesn’t look like a Guardiola disciple, despite the Catalan’s high praise of him. Instead, he looks more like a Mourinho-style defense-first manager with some limited Guardiola influences.
Put another way, Poche looks a whole lot like AVB.
What was AVB’s style?
Here are some numbers on AVB (NOTE: It’s hard to get super great advanced stats going back this far):
- In his lone season at Porto, his team conceded 16 goals in 30 games.
- In his 27 games at Chelsea, his team conceded an average of 1.2 goals per game. However, if you remove the extremely derpy 5-3 defeat to Arsenal and the two matches in which they conceded three to Sir Alex Ferguson’s Manchester United, you’re left with an average of .875 goals per game. So the system was not actually failing as badly as some think; they just didn’t adapt to elite opposition. But that specific failure in those three games is more about AVB’s judgment as a manager than his system, I think.
- In his last season at Spurs, his team averaged .8 goals scored per match prior to the City game and conceded .54 goals per game prior to the City game. (After the 6-0 set back against City all hell basically broke loose for a month under AVB and then continued in much the same way for another five months under He Who Shall Not Be Named so I am not going to worry too much about those stats.)
- The one big outlier is here: In his full season at Spurs, we averaged 1.7 goals scored per match and conceded 1.2 goals per match. These are much more “decent-ish Europa League EPL team” numbers than the others, which are much more obviously what you expect from a manager who values defensive shape and system.
Pochettino has always been a defense-first manager.
We’re now in the fifth season of Mauricio Pochettino managing in England. And here’s what we know: Poche is a lot more like AVB or Mourinho than he is Guardiola or Jurgen Klopp.
To demonstrate this, let’s begin by working backward to Pochettino’s Southampton days. Upon arrival in January of 2013, Pochettino inherited a Southampton team that knew how to attack, but... well, defending wasn’t a priority for them.
Here’s how they did in the first 22 games of their first season back in the Premier League played under Nigel Adkins:
- They averaged 1.36 goals per game.
- They conceded 1.82 goals per game.
Teams that defend like Adkins’ Southampton did don’t last long in England’s top flight.
In 16 games under Pochettino and with the exact same players, those numbers transformed dramatically:
- Southampton scored 1.2 goals per game, which represents a slight drop.
- However, they conceded only 1.25 goals per game—slashing their goals conceded per match by ~33%.
Here’s the impressive thing: The only teams in 2013-14 with better defensive records in England were Jose Mourinho’s Chelsea and eventual champions Manchester City.
When you look back at his normal XI during that season, that defensive record makes more sense. Pochettino played a 4-2-3-1 system that typically looked like this:
Boruc
Shaw, Fonte, Lovren, Clyne
Wanyama, Schneiderlin
Lallana, Davis, Rodriguez
Lambert
A few things should stand out about that lineup:
- Like his Spurs teams, this Southampton leaned on its attacking fullbacks to support the relatively limited attack.
- Morgan Schneiderlin, who has been used in a purely destroying, deep role when he can even get a game at Man United, is the most progressive midfield player.
- Steven Davis is part of the attacking four rather than more attacking players like Gaston Ramirez or Dani Osvaldo, both of whom Poche banished to Siberia after only a few appearances for the club.
Indeed, we can push the point further: His record at Southampton showed that Pochettino’s system requires a great deal of running and work, but that this running and work is closer to the work demanded by Diego Simeone than Klopp.
In an interview I did with him at Just Football, Into the Calderon editor Robbie Dunne said that Simeone believes in running and a high work rate because he sees football as being basically attritional in nature. That, of course, sounds an awful lot like both AVB and Pochettino.
How does Pochettino’s style explain results this season?
As at Southampton, Pochettino’s style at Spurs has been to generally play a more conservative midfield two with one player used in an almost exclusively defensive role. The first priority is to be defensively solid and to not concede bad goals. The attack must be built on this platform.
For this reason, Pochettino does not typically attack with numbers in the way that Klopp, Schmidt, Guardiola, or Thomas Tuchel do. Recall the numbers highlighted in this piece by Ricardo Tavares. Amongst the top teams in England last season, Spurs were the only ones whose main passing combinations were between defenders.
Rather than slow, patient possession, there are three main ways that Pochettino teams create chances:
- They rely on long, direct passing from the back.
- They use transition sequences in the attacking third which are often created when those long balls are not completed and the ball pings around a bit before Spurs win control and can attack an out-of-position defense.
- They grind teams down and score goals late in the game on the counter when they are still fresh and the opponent has been worn down. (Like, you known, the winner we scored just this past weekend against Burnley. But this description also fits many of the late goals scored in Pochettino’s first two seasons at Tottenham.)
Moreover, once teams get wise to this style and begin sitting back themselves, you typically end up being badly out numbered. Long ball attacks from Spurs often begin with four Tottenham attacking players going up against five or six opposition defenders but against teams like West Bromwich Albion earlier this season it was more like seven, eight, or even nine defenders.
Those numbers begin to change a bit once the fullbacks get forward, but as we saw in the Champions League, if you can keep Spurs’ fullbacks pinned deep, the attack sputters.
What happens when teams eliminate Tottenham’s primary attacking outlets?
This brings us back to AVB. When you play defense first and rely on direct attacks and counters to create scoring chances, it is relatively easy for opponents to say, “Fine. We’ll sit in a low block ourselves and then you can’t create chances.”
The result of this is that Tottenham games against weaker opposition begin to all look more-or-less the same: Spurs have a high amount of possession, but they have no ability to break down a packed defense because their players either don’t know how to open up such a defense or lack the ability to do so.
As a result, they end up taking lots of shots from distance and other lower quality chances. This is, in fact, a common problem with defense-first managers.
Here are a few sample xG maps from just this season for Poche and other defense-first managers:
Simeone’s Atletico earlier this season: 16 shots against Las Palmas, only one xG.
xG map for Atletico Madrid - Las Palmas.
8:34 PM - 17 Dec 2016
Here is Mourinho’s United against Watford earlier this season. 15 shots, 1.3 xG:
xG map for Watford - Manchester United. I've been expecting something like this to happen to United all season.
3:54 PM - 18 Sep 2016
That brings us to this season’s Tottenham. To be honest, the number of examples I could choose from is kind of terrifying. But here’s a small sample.
This is last weekend’s clash with Burnley. 30 shots, 1.7 xG:
xG map for Tottenham - Burnley. Spurs do it with volume. It's not pretty, and it's pretty risky.
Here is Bournemouth almost two months ago. Note that Erik Lamela played in this game so don’t think this is necessarily a “we-miss-Lamela” problem. 16 shots, .7 xG:
xG map for Bournemouth - Tottenham. Excellent defensive performance by Bournemouth, and they were unlucky this ended 11 v 11.
It’s a problem in Europe too. Here is the win against CSKA in Moscow. 23 shots, 1.3 xG
xG map for CSKA - Tottenham.
Another very Poche Spurs performance. Control midfield and take your time to unlock a set defense.
Aggregate xG Data for 2016-17
To make the point a bit more clearly, I went through and reviewed xG maps for all six top English teams so far this season. I took a very, very basic approach to the question: Let’s find out each team’s average xG/match, average shots/match, and average xG value/shot.
This is using only results from this season (and only results that I could find on Caley’s Twitter timeline). So the data is not perfect. Even so, the results are interesting:
Shot Quality Data
Team
xG / Match
Shots / Match
Avg. xG / Shot
Arsenal
1.7125
13.75
0.12
Chelsea
1.67
14.9
0.11
Liverpool
1.7
16.9
0.1
Man City
1.86
15.6
0.12
Man United
1.5
16.19
0.09
Tottenham
1.57
18.9
0.08
As you can see, our xG value per shot is the worst amongst the six sides with Champions League aspirations. Indeed, Tottenham’s xG value of any one shot is, on average, 50% lower than the xG value of Arsenal or City’s average shot.
One way of breaking down this data would be to identify two separate groups. I’m listing the members of each group in order from the most extreme example to the least extreme, though that is admittedly defined in somewhat subjective terms. (I’m looking for a way to do a chart like this one, but am having no luck—if someone can help me out in the comments, I can update the post with a chart.)
Higher Shot Quantity, Lower Shot Quality
- Tottenham: 18.9 shots/match, .08 xG/shot
- Manchester United: 16.19 shots/match, .09 xG/shot
- Liverpool: 16.9 shots/match, .1 xG/shot
- Arsenal: 13.75 shots/match, .12 xG/match
- Manchester City: 15.6 shots/match, .12 xG/match
- Chelsea: 14.9 shots/match, .11 xG/shot
At the other end of things, more attack-minded managers like Arsene Wenger and Pep Guardiola will generally produce higher quality chances because their teams are more comfortable attacking packed defenses and are better at opening them up.
Conte’s Chelsea, meanwhile, is sitting in the sweet spot between these two poles. On the one hand, they actually defend better than either Mourinho’s United or Pochettino’s Tottenham. But they do this without sacrificing shot quality: Their xG/shot is almost the same as Arsenal’s and City’s and is actually better than Liverpool’s, which has the most goals in the league so far this season. That is how you win a title.
It’s also worth noting that the team with the best chance at challenging Chelsea is probably Klopp’s Liverpool. They have also found a way to buck the trend by having a high shot quantity number and better shot quality than United or Spurs.
What can Spurs do to fix this?
This is the key question, of course. And, unfortunately, there aren’t easy answers. Conte’s Chelsea seems to have figured out how to retain a robust defensive structure without compromising shot quality, but then Conte’s Chelsea has N’Golo Kante in midfield, Eden Hazard in attack, and an on-fire Diego Costa leading the line.
So while Conte deserves major credit for his transformation of Chelsea, I am not sure how much Chelsea can be a template for other teams since most teams lack Chelsea’s financial muscle and consequent ability to recruit world-class players.
The more likely reality is that this sort of attack is probably more or less what we should expect from Pochettino teams and the best way to improve results is not necessarily by increasing shot quality, which may not be possible without compromising defensive soundness, but instead looking to eliminate mistakes that lead to high-quality chances for the opposition.
This past weekend’s fixture is probably as good a place as any to begin on that point: We weren’t bad, necessarily. In addition to our two goals, Dele Alli wasted a good chance in the opening minutes and came very close to scoring a curler from the edge of the box in the second half. Christian Eriksen also nearly scored from distance but for great work from Burnley keeper Tom Heaton and nearly got on the edge of a Harry Kane cross in the first half. None of these were great chances, but you also wouldn’t have been shocked to see any of those four chances find their way into the net.
What killed us is that we allowed Burnley two extremely good chances in the first half, both of which were totally unnecessary. In both cases, we had multiple chances to clear and we simply failed to do so. The second chance, which is the one that produced Burnley’s goal, also involved a couple lucky deflections for Sean Dyche’s team, but if Kyle Walker clears the ball when he has the chance to do so, those deflections never happen.
To put it another way, we probably shouldn’t expect Pochettino teams to consistently produce more than what this team is currently averaging, which is 1.57 xG/match. Pochettino teams are never going to create chances like an Arsene Wenger or Pep Guardiola team. However, the examples of Mourinho, Conte, and Simeone provide all the proof needed of how effective a defense-first system can be.
Indeed, Simeone’s example is particularly compelling as his Atletico have been competing for La Liga and Champions League titles for most of the past five seasons despite enormous financial disadvantages relative to their main rivals thanks to the fact that they are so absurdly difficult to beat.
To be sure, there are some structural things Spurs can do to improve results. If Victor Wanyama could get better at splitting the center backs that would go a long way. If Pochettino could be more flexible in-game and figure out ways to support his front four when his fullbacks can’t get forward, that would be significant.
But ultimately the attack we’re seeing this season is probably more-or-less what we should expect from Pochettino teams. The big question, then, is not “how can we improve the attack?” but “how can we eliminate unforced errors when defending?”
