Match Tottenham Hotspur v Brentford | Saturday 21st Sept | Tottenham Hotspur Stadium | 3pm

[SeeSpotRun]

Interesting points, I really like both those coaches but I'd have rewatch a De Zerbi Brighton game, try and catch a Marseille game or a Slot game to try and weigh up the differences/nuances in build up to reply properly.

I do think Ange's ethos is pretty similar in intentions. Suck in, play through, hit speedy wide forwards in quick transition. The FB's invert into midfield to try and aid that process and create constant out ball options and triangles to help play through the press. I think the problem is, some teams will just set up in a deep block - we saw Woolwich (and most teams) do exactly the same to City that they did to us. I'm pretty sure both Brighton and Liverpool use inverted wide forwards. Guardiola likewise usually - although Doku can go both ways - and has also used inverted FB's, and has rarely used overlapping FB's.

I definitely think Brentford's tactics played a part, but we seemed to tactically prepare for this (re Maddison post match) by moving our pieces around situationally. We saw Porro pop up wider than normal, Son and Maddison interchange, with Maddison popping up in that wide left area more. And we were very intense. Even when they tried to shell into a deeper block, we found ways past and round it, largely because we were so aggressive high up, despite them playing with a back 5,

I think against more passive, reactive deep blocks are still going to cause us problems, partly due to our lack of quality, specifically 1v1 problem solving forwards, but also because the deep block frustrates most proactive teams, that's why teams do it.

Click to shrink...

A few good posts on this page.

Our main attacking model seems to be:

1. Controlled build-up, gradually pushing the opp's back deep so we can set up our preferred attacking structure which tends towards a 1-1-5-3
2. Probe around the horseshoe for a while looking for a weakness
3. An incisive pass and then some problem-solving to unpick the defence and maybe create a chance
4. When 3 fails, press like crazy to force a high turnover
5. Quick passing sequence from the turnover to get in behind and create a big scoring chance

As others have pointed out in this and previous match threads, a lot of our goals are coming from stage (5).

When we spend too long doing (1) and (2), and don't have enough conviction / creativity / risk-taking moving into (3), then the result is a high possession % but low number of chances created (sound familiar?). It also creates less fatigue for the opposition since a low block is relatively low intensity compared to repeated transitions.

Our passes per attacking action and xG for our 5 EPL games so far have been LEI 20.7 (1.41), EVE 12.7 (2.61), NEW 8.8 (1.63), ARS 13.6 (0.79) and BRE 9.1 (3.89). Generally the better we've protected the ball, the lower the xG for the match.

I think there's a sweet spot around 60% possession and it you're above that (as we were for Leicester (70%), Everton (70%), Newcastle (65%), Woolwich (63%) and Coventry (70%)) then you need to take more risks in step (3). Maybe after you go a goal or two ahead you can start being more controlling.

Our possession vs Brentford at 48% was our 4th lowest ever under Ange, only pipped by our last 2 games vs City and the 9-man mayhem vs Chelsea. I think that was partly more risk-taking by us and partly the way Brentford chose to play, but regardless it worked well for us in terms of the number and quality of chances created.

 

[Klaus]

I wonder how xG adjust for the fact that something like 1 in 5 matches are decided by random events ie luck.

Click to shrink...

Should note that xG isn't a flawless metric and as such isn't the "whole story" for a one off game, but over the course of a whole season it averages out and is a pretty accurate metric.

For example, it doesn't take into account game state, as in, when during the game the chances came. If, for instance, you take the recent United 0-3 Liverpool game, the xG scores had them at roughly even (1.5xG-1.5xG), so the game should have been a draw, right? Well no, because United only started creating chances when Liverpool were already three nil up; they could afford to soak up the pressure at that stage. You'd have to imagine that had United scored one or two goals from their chances, Liverpool would have opened up more and created more xG of their own.

That's one of several examples of where xG is flawed. Another is that it only notes an xG for when a shot is taken, but that ignores plenty of great chances that don't result in a shot (think of England's cross to Gazza in the '96 Euro's semi vs Germany - 0.0xG there).

You mention luck, luck is obviously a huge factor in football because it's such a low scoring game (compared to say basketball) and the margins are so fine. And luck can determine the winner, but the point of xG is to try and take luck out of the equation and tell us who "deserved" to win (again, it's not flawless, so as an estimate).

For example, one of those Mourinho Spurs vs City games where they peppered our goal and destroyed us on the xG but we ended up winning with our only shot: that was luck. But you'd imagine if we played that game 100 times City would have won 90 of them.

And over the course of the season is where xG comes into its own. Luck from individual games evens out, and you're ultimately left with an xG (or xGA or xPoints) value that is remarkably similar to your goals scored. We are only 5 games in for instance and already our xG is only 10% higher than our actual goals scored.

That's why it winds me up when people start talking about how we'll be battering teams as soon as we start taking our chances. We are taking our chances, largely. We're just not creating remarkably good ones. It's unrealistic to think we'll suddenly start banging in 3-4 goals a game from matches where we create 1.5xG, it simply doesn't happen.

 

[SeeSpotRun]

I wonder how xG adjust for the fact that something like 1 in 5 matches are decided by random events ie luck.

Click to shrink...

Try this at home:

1. Take a normal 6-sided die.
2. Roll it 12 times, noting how many times you roll a 6. Each 6 is a goal for the home team, the rest are misses, blocks or saves.
3. Roll it another 6 times, noting how many times you roll a 6. Each 6 is a goal for the away team.
4. Record the result
5. Repeat steps 2-4 until you get bored, noting how many times you get a home win, away win, and draw.

The above basically simulates of a match where the home side has 12 shots with and xG of 0.167 each, and the away side has 6 shots with a similar xG of 0.167, for an aggregate xG score of 2.0-1.0.

If you repeat it 10 times, home side will win about 6, away side 2 and 2 draws. The chances of the result being exactly 2-0 (ie matching xG) is only about 12%, but this is still the most likely score.

Of course in a real game each shot has a different probability rather than all be 1 in 6, but the principle is the same. There will be a spread of probable results either side of the most likely score.

Then there are all the factors Klaus mentions where the match situation impacts the style of play, and where goal scoring opportunities that don't result in a shot don't count as xG.

 

[Klaus]

Top quality post and really highlights a lot of the issues I have with our style, specifically:

Our main attacking model seems to be:

1. Controlled build-up, gradually pushing the opp's back deep so we can set up our preferred attacking structure which tends towards a 1-1-5-3
2. Probe around the horseshoe for a while looking for a weakness
3. An incisive pass and then some problem-solving to unpick the defence and maybe create a chance
4. When 3 fails, press like crazy to force a high turnover
5. Quick passing sequence from the turnover to get in behind and create a big scoring chance

As others have pointed out in this and previous match threads, a lot of our goals are coming from stage (5).

When we spend too long doing (1) and (2), and don't have enough conviction / creativity / risk-taking moving into (3), then the result is a high possession % but low number of chances created (sound familiar?). It also creates less fatigue for the opposition since a low block is relatively low intensity compared to repeated transitions.

Click to shrink...

This is exactly what it looks like to me. Controlled build up, often in a sort of 2-1-2-5 formation (but that can even look like a 2-1-7) where the entire opposition is forced back into a narrow deep block to deal with our advanced "7". (Incidentally, to me we look a lot better when the fullbacks are deeper alongside Bissouma and we look more like a 2-3-5, because this draws out the opposition press while also protecting for the counter attacks, but that's beside the point).

So, we have our 2-1-7, we have Romero usually with the ball, advancing with the ball up to the half way line, and we have the opposition dropping into a deeper and deeper block. To me this is much more reminiscent of a Guardiola style "suffocation" style of play than a De Zerbi "we'll pass it around in our penalty box and draw you out in our own half, and then quickly pass between the lines".

The difference between us and Guardiola is (1) he keeps back a third centre back, his formations look like more of a 3-2-5 or a 3-1-6, partly why they're not nearly as defensively vulnerable as we are, and (2) Pep's attacking sequences once they have the opposition pushed back into their own penalty box are frankly miles better than ours.

Firstly, his wingers are actually 1v1 specialists, so that's one option he already possesses that we dont: get Doku or Savio 1v1 with their fullback.

Second, the fact his wingers are dribble specialists means the opposition have to double up on them or they're fucked (see City's 1st goal vs Chelsea when they provide Doku with too much space - he makes a yard, passes it into the box and before you know it it's 1 nil). In Brennan and to a lesser extent Son we just don't have that 1v1 threat so opposition fullbacks can give them more space, and consequently pack the middle of the pitch with their extra man. City have more space to pass it between the lines through the middle precisely because their wide players are threats it forces the opposition back line to spread out.

Third, Pep's passing sequences out wide are much better than ours. If Doku is being double marked, he always has a runner in behind the lines (often De Bruyne) and Pep has drilled them to make the pass, which they do multiple times a game and which results in loads of chances. Son you will sometimes see making that pass to Maddison but I'm not sure I can ever recall Brennan doing so, not least because he's right footed so his angles are all wrong.

And finally Pep understands that when the opposition are sitting in a deep block half the battle is getting it out to the wingers quickly, and forcing the opposition back line to quickly shift from side to side, which causes gaps to open up. A long-passing specialist like Rodri is crucial for this, and it's precisely why I was nervous about losing Hojbjerg with no replacement - for all his failings, he was the one midfielder who could consistently see and hit a long pass to the wingers, and it's no coincidence that when he was brought on as the #6 we would suddenly have a lot more luck in opening up a deep block.

In short: Ange plays like Pep, with the slow and controlled "boa constrictor" build up, but with none of the attacking patterns, wingers, or threat in the final third to make it work. We'd have a lot more success were we to move to a more De Zerbi or Slot system which keeps more of our players deeper and actively looks to bait the opposition press, and quickly play through the lines in behind to our forwards. It happened a few times against Brentford but I fear that's the exception rather than the rule.

Our passes per attacking action and xG for our 5 EPL games so far have been LEI 20.7 (1.41), EVE 12.7 (2.61), NEW 8.8 (1.63), ARS 13.6 (0.79) and BRE 9.1 (3.89). Generally the better we've protected the ball, the lower the xG for the match.

I think there's a sweet spot around 60% possession and it you're above that (as we were for Leicester (70%), Everton (70%), Newcastle (65%), Woolwich (63%) and Coventry (70%)) then you need to take more risks in step (3). Maybe after you go a goal or two ahead you can start being more controlling.

Our possession vs Brentford at 48% was our 4th lowest ever under Ange, only pipped by our last 2 games vs City and the 9-man mayhem vs Chelsea. I think that was partly more risk-taking by us and partly the way Brentford chose to play, but regardless it worked well for us in terms of the number and quality of chances created.

Click to shrink...

Interesting analysis.

City have shown that the 70% plus possession can work, but you need the attacking patterns in the final third, otherwise it just becomes a whole lot of passing around the centre circle for little gain.

Brighton have also shown that high possession can lead to high chance creation, but if you don't have City's players you perhaps need to be a bit more risky in your build up, and pass it around your own penalty box to bait the opposition. I fear we don't have the ball playing goalkeeper required to play this style. If Vicario was tasked with doing what RDZ required Jason Steele to do last season we'd concede 4 goals a half. (And that's an issue with recruitment - Vicario is a great shotstopper but why Ange didn't buy a possession specialist goalkeeper if he knew this was the style he wanted to implement I'll never know).

I do agree with you though - Brentford did press us quite a bit, and they also tried to pass it out from their keeper quite a lot, both of which played right into our hands. No idea why Thomas Frank set them up that way when there is a clear blueprint for how to beat us. I don't think we'll get so lucky going forward.

 

[bus-conductor]

I wonder how xG adjust for the fact that something like 1 in 5 matches are decided by random events ie luck.

Click to shrink...

Irrelevant. XG is the measurement of likelihood of a scoring from a given chance based on various criteria.

 

[bus-conductor]

A few good posts on this page.

Our main attacking model seems to be:

1. Controlled build-up, gradually pushing the opp's back deep so we can set up our preferred attacking structure which tends towards a 1-1-5-3
2. Probe around the horseshoe for a while looking for a weakness
3. An incisive pass and then some problem-solving to unpick the defence and maybe create a chance
4. When 3 fails, press like crazy to force a high turnover
5. Quick passing sequence from the turnover to get in behind and create a big scoring chance

As others have pointed out in this and previous match threads, a lot of our goals are coming from stage (5).

When we spend too long doing (1) and (2), and don't have enough conviction / creativity / risk-taking moving into (3), then the result is a high possession % but low number of chances created (sound familiar?). It also creates less fatigue for the opposition since a low block is relatively low intensity compared to repeated transitions.

Our passes per attacking action and xG for our 5 EPL games so far have been LEI 20.7 (1.41), EVE 12.7 (2.61), NEW 8.8 (1.63), ARS 13.6 (0.79) and BRE 9.1 (3.89). Generally the better we've protected the ball, the lower the xG for the match.

I think there's a sweet spot around 60% possession and it you're above that (as we were for Leicester (70%), Everton (70%), Newcastle (65%), Woolwich (63%) and Coventry (70%)) then you need to take more risks in step (3). Maybe after you go a goal or two ahead you can start being more controlling.

Our possession vs Brentford at 48% was our 4th lowest ever under Ange, only pipped by our last 2 games vs City and the 9-man mayhem vs Chelsea. I think that was partly more risk-taking by us and partly the way Brentford chose to play, but regardless it worked well for us in terms of the number and quality of chances created.

Click to shrink...

I slightly disagree with your theory of Ange’s preferred methodology. I think he initially wants to press high and hard, force turnovers and create overloads or spring the wide forwards in behind. Try to get 2/3 goals to the good, then sit deeper, suck teams out to chase the game, play through their press and ping the wide forwards in behind.

Ideally, he’d prefer to play a more proactive team every week. The opponent ended up pushed deep is a) because they set out that way (eg the first four games) or b) because our forwards can’t capitalise on situations and the opposition default to that due the heavy press or c) the opponent score first and shell.

 

[Klaus]

I think he initially wants to press high and hard, force turnovers and create overloads or spring the wide forwards in behind.

Click to shrink...

If this is actually what Ange wanted to do mate, he'd be happy to give the opposition the ball more often, because that would give us more opportunity to press them. Very Klopp-esque; get Trent or VVD to hit a long ball over the top to one of the wingers; if we win it, great, we can spring an attack. If we lose it, great, we can spring a high press.

But that's not what Ange does. We don't get the ball forward anything like a Klopp team. We are obsessed with possession, short passing build up, rarely ever do we hit a ball over the top to the wings.

Try to get 2/3 goals to the good, then sit deeper, suck teams out to chase the game, play through their press and ping the wide forwards in behind.

Click to shrink...

Is this actually what we do, or what you want us to do? I don't notice us change the way we play at all. To me this is the antithesis of Ange; he plays one way independent of game-state, and he is very open about that.

 

[bus-conductor]

If this is actually what Ange wanted to do mate, he'd be happy to give the opposition the ball more often, because that would give us more opportunity to press them. Very Klopp-esque; get Trent or VVD to hit a long ball over the top to one of the wingers; if we win it, great, we can spring an attack. If we lose it, great, we can spring a high press.

But that's not what Ange does. We don't get the ball forward anything like a Klopp team. We are obsessed with possession, short passing build up, rarely ever do we hit a ball over the top to the wings.

Is this actually what we do, or what you want us to do? I don't notice us change the way we play at all. To me this is the antithesis of Ange; he plays one way independent of game-state, and he is very open about that.

Click to shrink...

I don't think it's that simple. We've all seen/heard Ange get angry with players for playing safe. Klopp's team didn't really click until year 3-4 and some outstanding recruitment had put a pretty ideal 11 together. Ange doesn't have anywhere near the quality and compatibility of pieces Klopp had in place. There's no TAA smacking 30 yard passes, There's no Mane and Salah and VDV to clean up TAA's mess at the back.

And Ange doesn't always get to dictate game state. Some teams will park and bus and force us to to have lots of the ball in front of them, in exactly the same way they did to Klopp's Liverpool, Pep's City (like Arse did on Sunday) etc.

 

[SeeSpotRun]

Top quality post and really highlights a lot of the issues I have with our style...

Click to shrink...

Thanks, likewise. Nice to have a grown-up conversation without it descending into hyperbole, name-calling or seeing who can "win" the discussion.

City have shown that the 70% plus possession can work

Click to shrink...

I still think that's situational though. City's possession numbers are higher in games they won convincingly. Without extracting their possession numbers in the early parts of those games (before they went ahead) it's impossible to tell if it's chicken or egg first, but I suspect their possession % goes up after they establish a lead.
In the 10 matches that City drew or lost last season, their average possession was 60.5%. In the 21 matches they won by 2 or more it was 67.2%. Again chicken / egg debatable but it's consistent with a possession sweet spot of around 60% (via more risk-taking in attack) until you establish a lead.

In short: Ange plays like Pep, with the slow and controlled "boa constrictor" build up, but with none of the attacking patterns, wingers, or threat in the final third to make it work.

Click to shrink...

I don't think it's our preferred mode; we mainly revert to it if the opposition have retreated quickly into a defensive overload, which has become a popular tactic in "anti-Angeball". I feel that early in this calendar year we were still trying to play quickly through deep-dropping teams, but our forwards generally coughed it up too easily in the face of a defensive overload, and we became exposed to a lot of counter attacks. Around February I seem to recall the Boa becoming a more dominant feature of our build up, and I think it was a tactical response to conceding too many big chances on the counter. However, as per your four points (actually 3 really since the second is a consequence of the first), the lack of 1-v-1 wingers, limited effectiveness of passing sequences, and lack of long-pass specialists, has rendered our boa constrictor fairly toothless.

Recruitment over the summer is obviously looking to start to address this. I liked the look of the Odobert-Maddison combination down the left vs Everton; it's in stark contrast to say a Sarr-Johnson combo down the right. Kulu at R8 is more effective than Sarr although the patterns are a bit chaotic and don't utilise width very well; he almost reverts to a false 9. I'm hoping Bergvall can grow into that R8 role by the end of the season.

Anyway, despite some progress, it's pretty clearly that the summer recruitment is still several levels short of having us challenging for the premiership this season.

Yet Ange persists with a strategy that requires better players to be effective against low block & counter setups.

A pragmatist might revert to more Slot / de Zerbi setups as you suggest, keeping more players deeper, baiting the opposition press, playing more quickly through the lines. Or perhaps even just taking more long shots and swinging in more "put it in the mixer" type speculative high crosses. And arguably those tactics might be more effective in the short term.

But I'm ok with Ange's more idealistic approach, as long as the endgame is that the improvements through repetition (by sticking with this strategy) make us more effective by the time we do have the necessary players (through recruitment and/or development).

I'm also ok with other fans like you who can see all of the above and come to their own view that a better strategy is a middle ground somewhere between idealism and pragmatism, and that a manager change (if the right one is available and interested) sooner rather than later might be better for the club. I honestly don't know who's right.

(I don't have much time however for the "out of his depth", "found out", "naïve", "fraud" I-want-it-now-and-don't-have-it-so-toss-the-toys-out-of-the-pram types. If nothing else it's harder to have meaningful conversations with them.)

Sticking to the gameplan with an incomplete squad is not without its risks obviously. It may reduce our chances of CL qualification in the short term, which means significantly less budget for squad recruitment, which quickly becomes a downwards spiral. And it may lead to enough anti-Ange sentiment that we have to go through the turmoil of another cycle or two of manager changes.

There have been, and will continue to be, plenty of "sliding doors" moments where Ange moves significantly closer to or further from the exit door based on something which is largely chance. The dreaded Chelsea game, the luck of the draw for the FA cup 4th round, the home loss to Scum last season despite 2.73-1.28 xG, and the recent get-out-of-jail vs Coventry. That's part of what makes it fun I guess.

Tactics aside, I also rate very highly the things that Ange brings to the culture and attitude within the clubs that he's been at. Spurs need a healthy dose of that. So unless the results are spectacularly disappointing I'd be keen to see him given this season and maybe most of the next, before bringing in the "right" tactician to take us the final step to a premiership. And who knows, maybe by then we find out that the right tactician was Ange all along.

 

[SeeSpotRun]

I slightly disagree with your theory of Ange’s preferred methodology. I think he initially wants to press high and hard, force turnovers and create overloads or spring the wide forwards in behind. Try to get 2/3 goals to the good, then sit deeper, suck teams out to chase the game, play through their press and ping the wide forwards in behind.

Ideally, he’d prefer to play a more proactive team every week. The opponent ended up pushed deep is a) because they set out that way (eg the first four games) or b) because our forwards can’t capitalise on situations and the opposition default to that due the heavy press or c) the opponent score first and shell.

Click to shrink...

Yes poor wording by me probably, I didn't mean to imply that it's his preferred method overall, just that it's become our main method when we start with possession fairly deep and are facing an opponent retreating into a low block. See my previous response to Klaus for a bit more elaboration.

And indeed by taking more chances in attack once we've established our setup around a low block, we lose the ball more often and actually increase the rate at which we create forced turnovers and opportunities to get in behind.

So totally agree on your points.

 

[Richard Arlison]

Should note that xG isn't a flawless metric and as such isn't the "whole story" for a one off game, but over the course of a whole season it averages out and is a pretty accurate metric.

For example, it doesn't take into account game state, as in, when during the game the chances came. If, for instance, you take the recent United 0-3 Liverpool game, the xG scores had them at roughly even (1.5xG-1.5xG), so the game should have been a draw, right? Well no, because United only started creating chances when Liverpool were already three nil up; they could afford to soak up the pressure at that stage. You'd have to imagine that had United scored one or two goals from their chances, Liverpool would have opened up more and created more xG of their own.

That's one of several examples of where xG is flawed. Another is that it only notes an xG for when a shot is taken, but that ignores plenty of great chances that don't result in a shot (think of England's cross to Gazza in the '96 Euro's semi vs Germany - 0.0xG there).

You mention luck, luck is obviously a huge factor in football because it's such a low scoring game (compared to say basketball) and the margins are so fine. And luck can determine the winner, but the point of xG is to try and take luck out of the equation and tell us who "deserved" to win (again, it's not flawless, so as an estimate).

For example, one of those Mourinho Spurs vs City games where they peppered our goal and destroyed us on the xG but we ended up winning with our only shot: that was luck. But you'd imagine if we played that game 100 times City would have won 90 of them.

And over the course of the season is where xG comes into its own. Luck from individual games evens out, and you're ultimately left with an xG (or xGA or xPoints) value that is remarkably similar to your goals scored. We are only 5 games in for instance and already our xG is only 10% higher than our actual goals scored.

That's why it winds me up when people start talking about how we'll be battering teams as soon as we start taking our chances. We are taking our chances, largely. We're just not creating remarkably good ones. It's unrealistic to think we'll suddenly start banging in 3-4 goals a game from matches where we create 1.5xG, it simply doesn't happen.

Click to shrink...

This is a really solid post imo.

Agree with everything other than I don’t think we can say either way if our xG means we should have scored more or less, precisely because of the luck that you mentioned and the small sample size that you also mentioned.

I don’t think our xG from the Newcastle or Leicester games represented the number of dangerous situations that we created and fluffed. The scum game I think felt more accurate.

As an example being exactly what you referenced with the xG of the Gazza chance being zero. Leicester cleared the ball off the line twice (at least), one I remember clearly with Brennan right there for a 0.9xG tap in if the defender doesn’t pull the miracle off. The xG didn’t pick that up but over a season, on average the defender probably doesn’t pull off the miracle. In a single game situation or even 5 games, xG gives you the questions that your eyes need to answer. As you said exactly, over a longer stretch xG can tell you if you’ve been lucky or unlucky from a statistical perspective.

For the Newcastle game where turned them over about 7 times outside their box in the second half, I think xT might cover that over a season but I’m guessing even more on xT than I am xG tbh. And at the end of the day off you fluff those turnovers all season and don’t convert them to high xG chances, then you haven’t been unlucky anyway.

 

[Richard Arlison]

I slightly disagree with your theory of Ange’s preferred methodology. I think he initially wants to press high and hard, force turnovers and create overloads or spring the wide forwards in behind. Try to get 2/3 goals to the good, then sit deeper, suck teams out to chase the game, play through their press and ping the wide forwards in behind.

Ideally, he’d prefer to play a more proactive team every week. The opponent ended up pushed deep is a) because they set out that way (eg the first four games) or b) because our forwards can’t capitalise on situations and the opposition default to that due the heavy press or c) the opponent score first and shell.

Click to shrink...

C) has actually been really telling of our season so far. Brentford was only the second time in 5 games that we’ve taken the lead.

Newcastle weren’t half as deep at St James until they scored.

 

[dudu]

C) has actually been really telling of our season so far. Brentford was only the second time in 5 games that we’ve taken the lead.

Newcastle weren’t half as deep at St James until they scored.

Click to shrink...

Yeah I've seen some intimate that we let Newcaslte have the ball the whole game. They were actually on top u til they scored.... didn't someone get injured or there was a stoppage for some reason and after that we basically took control of the game?

 

[Richard Arlison]

Yeah I've seen some intimate that we let Newcaslte have the ball the whole game. They were actually on top u til they scored.... didn't someone get injured or there was a stoppage for some reason and after that we basically took control of the game?

Click to shrink...

Was fairly even for possession until HT but they scored pretty late.

We subbed Kulusevski on an dominated them second half.

 

[Starks]

Was fairly even for possession until HT but they scored pretty late.

We subbed Kulusevski on an dominated them second half.

Click to shrink...

Kulusevski started up top that day.

We subbed Johnson on. Which allowed us to move Kulusevski central midfield.

 

[Richard Arlison]

Kulusevski started up top that day.

We subbed Johnson on. Which allowed us to move Kulusevski central midfield.

Click to shrink...

Right my bad. I knew the Kulusevski to R8 change turned the flow and helped us press more but I had forgotten that he'd started CF.

 

[Klaus]

I don't think it's that simple. We've all seen/heard Ange get angry with players for playing safe.

Click to shrink...

We definitely have, but is that because he wants balls over the top for us to chase and press, or does he just want more urgency with our short passes into midfield to the likes of Bissouma/Maddison? We can only guess but based on the way we continued to try and play in that game at Stamford Bridge I'd assume the latter.

 

[r-u-s-x]

I wonder how xG adjust for the fact that something like 1 in 5 matches are decided by random events ie luck.

Click to shrink...

It would be included if judged over many games as it's intended. The goals used to create XG stats would have some luck involved.

 

[Klaus]

Thanks, likewise. Nice to have a grown-up conversation without it descending into hyperbole, name-calling or seeing who can "win" the discussion.

Click to shrink...

Tell me about it What a nice win does to us all eh?

I still think that's situational though. City's possession numbers are higher in games they won convincingly. Without extracting their possession numbers in the early parts of those games (before they went ahead) it's impossible to tell if it's chicken or egg first, but I suspect their possession % goes up after they establish a lead.

Click to shrink...

Honestly mate I think some games their possession is just flat out high the full 90', irrespective of game state. I've watched some games at the Etihad against relegation fodder where the possession ends up at like 75%-25% and I'm left wondering where the hell that 25% came from.

For instance, I watched their game against Ipswich from early on in the season. It was one of those 75% games and it very much felt like that watching it, from first minute to last. Here are the stats:


View: https://x.com/markstatsbot/status/1827375966139904057

And here is when the goals & chances game:

Ipswich actually scored early on from a quick counter, completely against the run of play. But the main story of the game was City hammering chance after chance after chance at the Ipswich goal. They eased up a fair bit after half time, which you would at 3-1 up at home, but it was still complete control.

What is my point? My point is that this was a game where City had complete control of the game and all the possession, and actually went behind early on, but at no point did that hinder their chance creation. The ball was constantly going out to the wide players - Doku and Savinho - who were constantly making little slide rule passes to KDB and co behind the lines. It was just an endless barrage of chance creation as reflected in the stats.

Now compare that to one of our games when we have 75% possession and it's a completely different story. The ball will get passed around the middle quite nicely, eventually find its way to the wings, and then... usually find its way back to a centre back. Then some more passes round the middle until Sarr gets bored and has a crack over the bar from 30 yards.

All of which is a very long winded way of saying, City's 75% possession to me looks very different to an Ange 75%. If we were to play a low block side this weekend at home--let's say Wolves--and I miss the game, and all you showed me was the possession stats, and I see we had 75% of the ball, I'm getting nervous. I'm having flashbacks from last year of horse-shoe football, passing placidly around the middle, and quick breaks with Neto on the counter. Show me Spurs possession of closer to 50% and I'm a bit more confident, because it suggests the game is more transitional which to me is the only way we create a good amount of chances.

Now, if you show me the stats of City home to Wolves where they have 75% possession, all I'm wondering is how long it took Haaland to score his hattrick.

However, as per your four points (actually 3 really since the second is a consequence of the first), the lack of 1-v-1 wingers, limited effectiveness of passing sequences, and lack of long-pass specialists, has rendered our boa constrictor fairly toothless.

Click to shrink...

Just wanted to say thanks for reading my comment in such detail you noticed point 2 was an extension of point 1

Recruitment over the summer is obviously looking to start to address this. I liked the look of the Odobert-Maddison combination down the left vs Everton; it's in stark contrast to say a Sarr-Johnson combo down the right. Kulu at R8 is more effective than Sarr although the patterns are a bit chaotic and don't utilise width very well; he almost reverts to a false 9. I'm hoping Bergvall can grow into that R8 role by the end of the season.

Click to shrink...

Completely agree with this. I really enjoyed the Odobert/Maddison link up. (It just concerns me we've bought Odobert as a RW option when he's clearly more comfortable on the left, which again raises questions over our recruitment in attacking roles).

Johnson had a much better game against Brentford but his lack of combination play with Porro, Sarr, Kulu etc is a concern. Agree about Kulusevski; I rate him but something seems off to me about when we play him in midfield; he almost wants to operate too high, almost as a false #9, and often drifts over to the right flank, and our midfield can end up looking pretty empty as a result. And it forces Maddison to drop deeper to compensate which I don't particularly like; I want him closer to the goal where I can hurt the opponent.

Early days with Bergvall but he clearly has the ability, and to me it seems he knows instinctively how to play as a midfielder--which spaces to pick up, and where to move based on where Maddison & Bissouma are--more than Kulu does, who just tends to just do his own thing and float right or into the forward line.

My short term solution would be to move Kulu back out to the right wing, use Brennan as a super sub, bring in a proper B2B (so, Sarr or Bentancur) to play alongside our #6, and push Maddison higher where he can hurt the opponent. Our chance creation might suffer slightly, but to me we'd be a lot more balanced in the middle and more solid defensively.

But I'm ok with Ange's more idealistic approach, as long as the endgame is that the improvements through repetition (by sticking with this strategy) make us more effective by the time we do have the necessary players (through recruitment and/or development).

Click to shrink...

What worries me about this strategy is our recruitment. In theory I am fine giving the idealism more of a shot, but it completely rests on the recruitment being top notch, so that we do end up giving Ange the players he needs to make his idealism work.

But to date our recruitment has been anything but, especially in the attacking positions. We've spent £80m combined on Brennan and Odobert, neither of whom are anywhere close to the required standard or even the right profile of right winger we need.

(Odobert has high promise as a LW but that's not the problem he's been brought in to solve. Brennan has ability as an inside right forward for a counter attacking team, but at this stage it should be obvious to everyone he's not a touchline winger on a possession team tasked with cracking a low block).

How have we ended up in this position? It's an inordinate amount of money to have wasted on one position, and speaks to a failure of player profiling, either in the manager, or in the recruitment department, or some combination of the two. The fact our alternative option on the right--Pedro Neto--is (1) absolutely nothing at all like Odobert and (2) horrendously injury prone just leaves me scratching my head even more.

Where is the strategy here? And who's to say if we don't give Ange and this recruitment team another £80m to sort out that problem right wing position they don't just spaff it on more players who don't fit the profile?

I'm also ok with other fans like you who can see all of the above and come to their own view that a better strategy is a middle ground somewhere between idealism and pragmatism, and that a manager change (if the right one is available and interested) sooner rather than later might be better for the club. I honestly don't know who's right.

Click to shrink...

Neither do I to be honest. I'm aware of my biases, and I'm aware I'm impatient when it comes to managers. I'd have probably sacked Arteta back when he was floundering a few years back. The flip side to that is that not all managers are Arteta--ultimately most are AVB's, Nuno's, and Mourinho's, for whom the only issue in hindsight was that we didn't sack them sooner.

Ultimately all these decisions are based on a gut feeling, and my gut feeling is that Ange is more AVB than Arteta, and we'd be better off ripping off the bandage now and starting afresh with someone with a better track record at the top level. I hope I'm wrong, not least because it seems pretty clear we're in this for the long haul with Ange, for good or ill.

the home loss to Scum last season despite 2.73-1.28 xG

Click to shrink...

Just a slight quibble with this one--this is a prime example for me of when the xG doesn't tell the whole picture. Woolwich were 3-0 up at half time; of course they were going to sit back and soak up the pressure, and Spurs were going to rack up the chances and the

xG. Not dissimilar to the United vs Liverpool game earlier on this season. Chances created at 3 nil down aren't the same as chances created at nil-nil, and that game Woolwich probably deserved the result for a smart & streetwise performance despite it going against the xG narrative:

 

[SeeSpotRun]

What is my point? My point is that this was a game where City had complete control of the game and all the possession, and actually went behind early on, but at no point did that hinder their chance creation. The ball was constantly going out to the wide players - Doku and Savinho - who were constantly making little slide rule passes to KDB and co behind the lines. It was just an endless barrage of chance creation as reflected in the stats.

Now compare that to one of our games when we have 75% possession and it's a completely different story. The ball will get passed around the middle quite nicely, eventually find its way to the wings, and then... usually find its way back to a centre back. Then some more passes round the middle until Sarr gets bored and has a crack over the bar from 30 yards.

All of which is a very long winded way of saying, City's 75% possession to me looks very different to an Ange 75%.

Click to shrink...

I think this aligns well with my initial point:

Our main attacking model seems to be:

1. Controlled build-up, gradually pushing the opp's back deep so we can set up our preferred attacking structure which tends towards a 1-1-5-3
2. Probe around the horseshoe for a while looking for a weakness
3. An incisive pass and then some problem-solving to unpick the defence and maybe create a chance
4. When 3 fails, press like crazy to force a high turnover
5. Quick passing sequence from the turnover to get in behind and create a big scoring chance

As others have pointed out in this and previous match threads, a lot of our goals are coming from stage (5).

When we spend too long doing (1) and (2), and don't have enough conviction / creativity / risk-taking moving into (3), then the result is a high possession % but low number of chances created (sound familiar?).

Click to shrink...

I think we're in agreement that we're not attempting enough slide-rule passes, take-ons and flicks, and choosing safe options too often. There are a few components to that including:

• individual ability
• individual and collective confidence / conviction
• off-ball runs to create opportunities for defence-splitting passes
• pre-drilled attacking patterns
• fear of consequences of coughing up possession
• instructions from manager

The symptoms are pretty clear: high possession but low xG. And it's pretty squarely on Ange and his coaching staff to be improving all of the above. We're still a long way from City levels but that's not really that much of a surprise. But we were pretty horrible early this calendar year and I like to think we've improved so far this season but just not clicking yet (witness the number of cut-backs to a vacant penalty spot, and number of fizzing low balls in front of the keeper when our striker has checked back to the penalty spot).

Honestly mate I think some games their possession is just flat out high the full 90', irrespective of game state. I've watched some games at the Etihad against relegation fodder where the possession ends up at like 75%-25% and I'm left wondering where the hell that 25% came from.

For instance, I watched their game against Ipswich from early on in the season...

Click to shrink...

My thesis is that there's a shift in risk/reward depending on the score, which mean it's better to attempt more risky defence-splitting passes when trying to establish a lead, which is going to mean lower possession stats because they don't always come off. Then after establishing a lead, risk/reward favours safer option when in possession, which leads to higher possession%. So let's test the hypothesis on City's last 10 or so EPL games:

Match	Possession when not leading	Possession when leading	Difference
Luton (H)	N/A (scored 2nd minute)
Brighton (A)	64.2	64.8	+0.6
Forest (A)	71.1	65.1	-6.0
Wolves (H)	49.6	59.0	+9.4
Fulham (A)	60.4	64.8	+4.4
West Ham (A)	N/A (scored 2nd minute)
Tottenham (A)	45.9	47.7	+1.8
Chelsea (A)	52.2	51.7	-0.5
Ipswich (H)	69.4	75.9	+6.5
West Ham (A)	66.6	67.6	+1.0
Brentford (H)	43.2	59.9	+16.7
Arse (H) (first half only)	66.1	61.7	-4.4

(Data from whoscored match centre; there's a slider at the bottom that lets you see stats for any time slice)

So it's a bit meh statistically but maybe there's something there. Obviously how the opponents decide to set up both (a) initially when scores are level and (b) after they go behind, is going to influence how possession pans out.

Anyway, coming back to the main point: Tottenham's matches with high possession% and low chance creation (Leicester, Newcastle, Coventry) are symptomatic of insufficient risk-taking when in possession, too often opting for safe passes around the horseshoe, and not enough attempts at defence-splitting passes, take-ons, disguised flicks etc.