For a man so revered, so attentively studied, so plentifully adorned with silverware, Pep Guardiola cuts a remarkably dour figure. Constantly anxious and frequently frustrated on the sidelines, he gives the impression of a tortured soul in a quixotic quest for footballing perfection, bored by the accolades routinely showered upon him by his peers and painfully aware of the dispiriting gap between his teams’ concrete performances and the Platonic Idea of football. 

Alex Ferguson expressed bewilderment at Guardiola’s decision to leave Barcelona after just four years at the helm. Where could you possibly get a better set-up, he asked. But on reflection, the Bundesliga offered Guardiola a context in which beating other teams was a peripheral formality, so that real work could focus on methodically chipping away at that gap between his team and greatness-as-such. 

It turned out to be difficult, however, to squeeze the best out of players who had to wait until March for their first consequential match. Hence Guardiola’s subsequent attraction to the Manchester City job: although in a worrying state of entropy when he took over, City represented a tabula rasa with no entrenched football culture, a more competitive environment and virtually unlimited material means for accomplishing Guardiola’s quasi-spiritual ends. 

The squad needed rebuilding, but softening the challenge was all that crude lying underneath Abu Dhabi. Therewith City could offer Pep six new players upfront, at a collective price tag of around £160m – about half the annual budget of the nation of Tonga (which has to take care of 100,000 citizens, mind you, not a 25-man squad). 

But spending money is easy; how to approximate aesthetic perfection is a true metaphysical conundrum, one that calls for a figure like Guardiola, a man who in the mid-90s would go to poetry readings after Cruyff-led Barcelona training sessions and who still last year was reciting Catalan poetry in literary soirées at Munich’s Literaturhaus, focusing on the work of Miquel Martí i Pol, whom as a young man he had met at those Catalan poetry readings. It is this fiery competitor with a poetic soul, if anyone, who could tackle the conundrum of pitch-sized aesthetic perfection. 

But what is aesthetic perfection? According to the German philosopher Friedrich Nietzsche, it arises from the careful equilibrium between, on the one hand, the “Apollonian” principles of order, symmetry, composition and calm control, and, on the other hand, the “Dionysian” spark of untamed creativity, bursting ecstasy and unbridled vital forces. At bottom, ‘Guardiola football’ is but the application of Nietzschean aesthetics to the realm of football. If the 2011 Barcelona side was the greatest ever, it is because Guardiola managed to create a perfectly balanced Apollonian framework, sagaciously conducted by Xavi and Busquets, within which he let loose Messi’s and Iniesta’s Dionysian genius. 

It is this subtle choreography that Pep saw an opportunity to recreate upon the virgin cultures of Manchester City Football Club. His messianic mandate was simple: Moses extracted water out of a rock; Jesus turned water into wine; Pep shall transubstantiate Gulf oil into footballing poetry. 

Clearly, Pep’s first year in England was not a success. Yet spirits are high around the Etihad. Neither fans nor institutional structure seem particularly worried. To understand why, we need a more anatomical analysis of last season’s progression. 

Average managers take for granted their teams’ general baseline of competence, seeing their task as that of pulling their players’ game-time performance as far above that baseline as possible. More sophisticated managers consider it part of their mandate also slowly and patiently to drag their team’s competence baseline ever higher. This season, it sometimes seemed that Guardiola put exclusive focus on this more nuanced task, treating the more basic one almost as an afterthought. 

The reason for this is that Guardiola’s style of football implies training in a whole new set of football habits and instincts. Known as juego de posición, it requires players – in Guardiola’s model – to assimilate a mental schema that divides the pitch into twenty modular sub-pitches of differing dimensions (see figure below). Players must rewire their brains to play differently in each sub-pitch and coordinate their spatial position across clusters of zones depending on where the ball is.


Assimilating a new set of athletic instincts takes time, and it involves a period in which the new and superior instincts have not yet settled in but the old, tried and true ones are starting to disintegrate; in this period one plays worse. Between 2004 and 2007 Roger Federer won 4 Wimbledons, 4 US Opens, and 2 Australian Opens – but not once at Roland Garros. Deciding that his only remaining wish was to complete a career grand slam by winning on Parisian clay, he reworked his game to adjust it to that peculiarly bouncy surface. And indeed, on 7 June 2009 he beat Robin Söderling in straight sets to secure his coveted career grand slam. But this had been only his second grand-slam victory in 21 months! While injuries were also a factor, clearly the project of rewiring his brain to play clay-court tennis took its toll. 

Learning Guardiola’s style of football was always going to exhibit the same dynamic. It requires considerable training and a non-negligible threshold of intelligence, but it also requires a real capacity for concentration, a cognitive resource notorious for its fast rate of depletion. It is a constant of Pep’s career that his teams play their best football in the first 30 to 35 minutes of a game. That is precisely the human brain’s natural capacity for sustained concentration, as cognitive psychologists determined some years ago. 

That’s why it had always been so crucial for Pep’s Barça and Bayern to score in the first half hour. In 2016-17, City played 56 matches in all competitions. Of the 27 matches in which they scored by minute 35, they won 23; they took home 52 of 57 possible points in league matches in which they scored by minute 35. In contradistinction, City won only 9 of the 29 matches in which they did not score by minute 35, taking home 26 of 57 league points. That’s 85% wins in matches with goals inside the concentration threshold against 31% in matches without. 

If in 2017-18 concentration could be sustained more reliably and City could convert early domination into goals and leads, things could start looking much better. And if Guardiolista football automatism can be trained deeper into Kevin De Bruyne, David Silva, Leroy Sané and company, indeed maybe even (a healthy) Vincent Kompany, we may yet see in north-west England scenes reminiscent of Catalonia and Bavaria. 

This, at least, seems to be the current thinking at City. There is no sense of panic, because there is an appreciation that Guardiola will need time to drag the team’s competence baseline to a level where it has the chance to implement that desired equilibrium between Apollonian structure and Dionysian creativity.

In fact, a close study of this season’s twists and turns reveals a constant search for the kind of system most likely to enhance this implementation. As we shall see, Guardiola played six or seven basic systems in 2016-17, each devised in response to a constitutive limitation he discovered in the squad he’d more or less inherited. The story of City’s evolution this season is a story of continual adaptation, of a frustrated but ever hopeful master chiseling the prime matter at his disposal in search of its inner potential. 

Traces of Guardiolesque football could be witnessed at the Etihad within a month of Pep’s arrival. City started the season playing the kind of 4-3-3 that covers the pitch with overlapping triangles, making it very difficult to lose the ball as long as everyone knows where they’re supposed to be on an imaginary shifting grid whose geometric zero point is the ball.

By the end of his first month of competitive play, Guardiola exuded his familiar air of disgruntled success. City won all five of their August matches, scoring an average of three goals per game and letting in only three all month. Like a man who made billions in investment banking, sitting on his imposing yacht and brokering mega-deals of geopolitical consequence, but all the while wallowing in existential dissatisfaction and yearning for more grace and insight in his life, Guardiola felt that while results were going their way, City’s performances left much to be desired. In Brian Clough’s memorable words, he wanted to win, but win better. 

Only two performances that month were to Pep’s satisfaction: City’s 5-0 schooling of Steaua Bucharest in the away leg of the Champions League play-off, which Pep billed as City’s most important match of the year, and a 3-1 home defeat of an injury-debilitated West Ham. In both matches one could see signs of the flowing-yet-controlled mark of Guardiola football, creating complete domination of a disoriented, overwhelmed opposition.  

As Guardiola was well aware, though, City hadn’t faced a substantial test up till then. Their first arrived on September 10 with the Manchester derby at Old Trafford – a first English instalment of the Guardiola v Mourinho modern epic. With two weeks to prepare, Pep had clearly outthought Mou: despite losing Sergio Agüero to a three-game ban and an atrocious debut from Claudio Bravo, City erupted from the gates to take hermetic control of the ball for the first 40 minutes, playing from the back with complete poise and reducing Paul Pogba, Henrikh Mkhitaryan and Jesse Lingard to witnessing bystanders in midfield. The 2-0 lead after 36 minutes barely told the story of a half in which City had 66% possession and passed the ball 300 times to United’s 139. Although Mourinho did recover at half-time, managing to smother City’s backfield initiatives with a high press and to impose a much more muscular and chaotic play throughout the pitch, ultimately City came much closer to doubling their lead than United – determined but uninventive – did to overcoming it. 

It is after this match that Guardiola could finally be detected cracking a smile, telling reporters in the sweep of subdued euphoria, “football is like this; when you win you’re happy.” (This was Pep’s eighth triumph over Mourinho, who had beaten him only three times.) It took only four days for Pep to sound a dourer note again. The occasion was City’s 4-0 Champions-League cakewalk victory over Borussia Mönchengladbach. “We will have to play more attractive football,” he reflected ruefully apropos of the patchy crowd in the stands. 

After two more humdrum league wins, on September 24 City met Swansea in what was perhaps the most significant match of the Pep pre-programming era. Guardiola divided the game into two halves, the first playing Silva in the middle and De Bruyne on the right wing, the second playing De Bruyne in the middle and Silva at left wing. This served as what Newton called an experimentum crucis: an experiment that pits rival hypotheses against each other by isolating a single variable against a backdrop of constants, with the hope of decisively confirming one hypothesis and disconfirming the other. Pep was essentially asking who would be the better Dionysian engine of his Apollonian apparatus: the clever Spaniard, master of space creation, or the fast-thinking, space-swallowing Belgian. 

The results of the experiment were unequivocal. City were stilted in the first half, having trouble unlocking Swansea’s combination of zonal marking and disciplined press, going in at the break with the score at 1-1. In the second half things started to open up. De Bruyne’s speed forced more Swansea midfielders to stay back, allowing City’s central defenders to be more adventurous. City scored two more goals to secure a 3-1 away victory. 

But just as these experimental findings came in, De Bruyne was injured in the twilight of the match. He was sidelined for the following three weeks, then took months to regain his early-season level of sharpness and efficacy. 

In their next match City were finally held, though just barely, to a 3-3 draw at a sizzling Celtic Park. The Celtic manager Brendan Rodgers came up with an original game-plan that actually worked: rather than thinking of how to contain City as a defensive exercise, Celtic stormed the pitch from minute one and attacked continually, overwhelming City for the first ten minutes and making it very hard for Guardiola’s men to settle in possession and control proceedings. Only in the game’s final half-hour did City establish control and systematically poked around Celtic’s defence in search of loopholes. But with De Bruyne out, a certain spark was lacking and the City machine, though compelling at times, ultimately proved insufficient.

What Celtic did for 50 minutes, Tottenham did four days later for 75. With a feverish and relentless high press they suffocated City’s backfield, truncating City’s attack (Agüero was barely to be seen) and handing Guardiola a first and sobering 2-0 defeat. We learned that Spurs’ attackers defend a lot better than City’s defenders attack. But in truth the model was laid out already by Mourinho at the start of the second half in the Manchester derby. With the model now demonstrably bearing fruit, other teams were sure to adopt the same approach. 

Faced with this challenge, Guardiola’s response was to do away with full-backs. Pep opened the next three league matches with his own variant of the modish three at the back. But this was no catenaccio-style three-at-the-back, the kind that smoothly morphs into five at the back as wing-backs draw into the fold out of possession and double as full-backs. No, this was the kind of three at the back he’d pioneered at Bayern that deserted the defensive flanks and sent seven players into enemy territory, a kind of 3-2-2-3 that makes it too risky for the opposition to invest in pressing the defence – a simple ball over the press would create an instant cardinality problem on the other end. 

The first half hour of this new concoction from Pep’s lab, at home to Everton, was intoxicating: City enjoyed 76% possession and essentially conducted the whole game in Everton’s half. And yet this dominance seemed not to translate into much in terms of chances. Pep’s men were ineffectual in the final third, and the match ended in a 1-1 draw. And something else seemed particularly disconcerting: Ronald Koeman, Pep’s erstwhile flatmate, showed that the high press was not the only way to contain City, a simpler-minded form of defensive compactness could do the trick equally. There just wasn’t enough unpredictability in this City side.

After three more uninspiring performances (a home draw against Southampton, a League Cup exit to United by City’s B team and a win over West Brom), Pep retired the new system definitively. Although he would still occasionally play three at the back, it was always the more traditional variety featuring wing-backs with real defensive responsibilities. 

The team went back to a 4-3-3 for Pep’s first visit to the Nou Camp as City manager (playing no full-backs against Messi and Neymar being a nonstarter), though featuring De Bruyne as a false 9 (leaving Agüero on bench). The idea was to flood the midfield and leave Barcelona without enough of the ball to wreak havoc. The plan actually worked and City were in control until a Fernandinho slip left Messi in front of Bravo and soon thereafter in front of an empty net. A devastating Bravo mistake led to a 4-0 defeat, but perhaps not a defeat reflective of the game-plan. The reverse was heavy, but there was reason for optimism about the imminent rematch at the Etihad. 

The closest Shakespeare came to writing an autobiography was his last play, The Tempest. Its protagonist, Prospero, is ostensibly an island-stranded magician who raises alone his perfect daughter Miranda and controls the island’s forces of nature through two slaves, Ariel the angel of music and atmosphere and Caliban (not quite Cannibal, but almost) the wine-worshipping brute. The subtext has been clear to four centuries of Shakespeare scholars: Prospero is the inventive playwright who exercises deliberate control over his magnificent creation and does so by choreographing the activities of two complementary forces, one angelic and the other animalistic (the Greek god of wine, of course, is Dionysus, that of music, Apollo).

It is those same two complementary forces that Guardiola tried to coordinate for Barça’s visit to the Etihad on November 1. After scoring against the run of play toward the end of the first half to level the score, City took the match by the horns in the second half with a dogged display of energy and vitality, fighting for every ball and stretching the pitch along the flanks. İlkay Gündoğan and Silva provided the angelic notes, but it was De Bruyne’s animalistic tirelessness that shepherded City to a 3-1 victory. 

Commentators hailed the match as ushering in a new era for Manchester City, the era of belonging to Europe’s big boys. Yet in truth it was speed, directness and strength that won the day, not the subtle elegance of Pep’s presumed long-term project. Tellingly, City enjoyed only 34% of possession in the game. It could be argued this was an admission that City did not yet belong at the top table. They have to find circuitous and inventive ways to keep up with a side like Barcelona; they cannot just play their game and expect to impose themselves.

In 2016-17, however, no Premier League side embodied football animalism better than Antonio Conte’s Chelsea. With a band of blistering turbo runners like Eden Hazard, N’Golo Kante and Victor Moses dashing and charging across the pitch, Diego Costa dishing out his home blend of football instincts and controlled sociopathy and Conte on the sidelines yelling and howling in the grip of ecstasy, Chelsea found a compelling identity that catapulted them to boundless self-belief. 

On the back of seven straight league wins and sitting atop the table, the Blues visited the Etihad on December 3. After a series of underwhelming results in which City reverted to the 4-3-3 (though once – at Burnley – with an inverted, forward-pointing midfield triangle), Guardiola attempted to mirror Chelsea with three at the back and the wing-backs Jésus Navas and Sané expected to track back and help contain Hazard and Pedro. Meanwhile, Conte made only one change to his standard starting eleven, fielding Cesc Fàbregas instead of Nemanja Matić. His plan here was for Fàbregas to play well-weighted long balls just over City’s unreliable high line. 

After 20 minutes of relative disarray, City’s midfield took control of the game and dictated proceedings for the following 40 minutes. Nonetheless Guardiola’s men found it hard to translate this into real chances and only a Gary Cahill own goal gave City a half-time lead. City started the second half brilliantly, with effective counter-attacking that created two golden chances De Bruyne failed to convert. But in the process the midfield was forsaken and chaos started slipping into the game. Meanwhile Diego Costa was clearly told to drop deep and draw City’s high line even higher, creating forsaken landscapes in front of Bravo. It is through these unoccupied spaces that Conte’s men were to exact punishment on Guardiola’s team with three goals in the game’s final half hour. 

By the end of the match City were in meltdown mode, with Agüero and Fernandinho seeing red cards ahead of important face-offs with Arsenal and Liverpool. In his post-match press conference, Guardiola put on a brave face, graciously congratulating Chelsea and complimenting his own players: “I’m proud of the way we played… We played quite well… I don’t have regrets…”

Still, after 14 games City had only one point more than at the same stage the previous year, under the supposedly hapless Pellegrini. They had been beaten by the only two top-five teams they had faced (Tottenham and Chelsea). Their defence seemed not a shade better than on week 1. But at least City had the best manager in the world, no? Well, in fact Guardiola seemed to have been outfoxed, in different ways, by both Pochettino and Conte, two fellows likely to stand in his way to silverware for years to come. 

Among the many curious characteristics of American Football stands out the following oddity: every NFL team is in fact composed of two completely independent and unrelated squads, one for offence and one for defence. These two teams are never on the field at the same time, share virtually no players and no tactics, and are subject to separate training regimes. You could in principle recombine any NFL offence with any defence with complete freedom – the sporting activities of the one have no direct relevance to those of the other. The only connection between the New England Patriots’ offence and their defence is the fact that they are owned by the same organisation – a legal connection. 

In consequence, the two squads making up one NFL team also have two different coaches: there is the offensive coordinator for the one and the defensive coordinator for the other. Granted, there is also a head coach overseeing the coordinators, but it is the coordinators who take most game-time decisions and who are in charge of most workaday training. And nobody becomes head coach before having a long career as a coordinator. Head coaches are more often defensive coordinators in origin, but every successful head coach is an excellent coordinator of both types.

In City’s next league game, against ineffectual Leicester, we learned a terrible secret about Guardiola: he is a poor defensive coordinator. The reason Leicester had been so ineffective is that after storming the league with long balls over the top to a buzzing Jamie Vardy, the league understood that the way to stop this one-trick pony is to eliminate space between keeper and central defence. In consequence, Vardy had not scored in 16 matches until facing City in early December. On that day, however, Vardy scored a hat-trick and led Leicester to a compelling 4-2 win as Pep’s boys took a hell of beating: down 2-0 after less than five minutes and 3-0 at half-time despite 76% possession and seven corners – but with zero shots on target. 

Almost none of City’s stars played well on the day; Gündoğan, Kolarov and Sterling were all poor. But the real reason for the debacle was that Guardiola played the first high line Leicester had faced in months and months, helmed by a bumbling John Stones to boot. It took Leicester all of two minutes to stroll through this apparatus for a tone-setting goal. 

Presumably, Guardiola knows what everybody else knows about the defending champions – that they will hurt you mercilessly, but only if you give them space behind the defence. And yet he shows up at the King Power and does just that. What could possibly explain this?

The only believable explanation is that Guardiola is just not very good at defence. It matters little if this is due to deliberate lack of intellectual investment or down to a genuine cognitive limitation. At Barcelona and Bayern, Guardiola’s mediocrity as defensive coordinator was obscured by the fact that his teams dominated proceedings so hermetically that opponents simply did not have enough input into the shape of the game to expose whatever defensive design flaws were lurking there. 

It is precisely because in football offence and defence are so intimately enmeshed that one can boast a strong defensive record purely on the strength of one’s offensive game-plan. In 2010-11 Guardiola’s Barça conceded 21 goals in La Liga while Bayern let in 17 in his final season at in the Bundesliga. (For comparison, Mourinho’s Real Madrid and Inter teams both conceded at least 32 in a season.) 

These defensive successes rested on a sound philosophical precept: no one can score against you if they don’t have the ball. The problem is that this single precept can govern your entire defensive approach only if your offensive game is dominant enough. For the first time in his managerial career, at City Guardiola struggled to dominate games to the point that his shortcomings as defensive coordinator became ever more apparent. 

At Leicester Guardiola deployed a complicated tactic featuring three at the back with Zabaleta in midfield in possession, but with the Argentinian almost dropping to be a fourth defender when out of possession. In the wake of the Leicester fiasco, Pep dropped Stones and reverted to a steady 4-3-3 formation, playing Kolarov and Otamendi in central defence. The experimental caprice would have to find expression in a series of false-9 auditions: Nolito did not convince in a 2-0 win at Watford; Sterling was out of place in a disheartening first half at home to Arsenal, but De Bruyne (sometimes switching with Sané) did better in the second half to ensure a much needed 2-1 victory; Nothingness was the surprise candidate in an uninspiring 3-0 win at Hull featuring something like a 4-6-0 formation. None of this would do. Something new was needed if Pep were to adapt. But what? 

More than 200 years before James Milner was converted by Jürgen Klopp into a decent left-back, the trailblazing thinker James Milne converted his own name to James Mill. It is under that name that he would later make his greatest contribution to humanity: the siring of John Stuart Mill. Little John Stuart wrote English and ancient Greek by age of three. At eight he learned Latin so he could read Euclid’s Elements, but it also stood him in good stead when at twelve he dedicated himself to the study of medieval logic. As a grown man, his most important ideas pertained to moral philosophy. His book On Liberty is still the quintessential statement of the fundamental principles of liberal democracy and the open society. In this and many other areas, Mill’s ideas were remarkably faithful to his father’s. 

One of Mill’s most controversial breaks with his father was over a crucial detail of utilitarian philosophy. According to utilitarianism, right and wrong should not be determined by oblique social and religious duties, but by the soulful precept that all action should try to enhance “the greatest happiness of the greatest number”, as Mill Senior’s friend Jeremy Bentham had put it. The controversial idea introduced by JS was to insist that although nothing is intrinsically valuable but people’s joy, there is nonetheless better and worse kinds of joy – some joys are inherently and objectively “higher” than others. “Human beings have faculties more elevated than the animal appetites,”
he wrote. 

Perhaps fearing the charge of elitism and arbitrariness, he offered no concrete example of higher versus lower joys. But we may illustrate what he had in mind by pointing at the difference between the joys of winning ugly and the joys of orchestrating expansive, exciting, elegant, free-flowing football. The pleasure of beating our cross-town foes, however it is achieved, satisfies something of an animal appetite – real and compelling no doubt, but no more noble for that. When we witness a series of perfectly weighted touches astutely distributed across space, however, we experience a qualitatively different, more exhilarating and more spiritual kind of delight. 

Through the first half of the season, two teams consistently offered us the higher joys of football. There was Guardiola’s City, but there was also Klopp’s Liverpool, with its chaotic, breathless, barely controlled swarming of energies that produced a league-high 46 goals. Accordingly, all eyes were on Anfield as the two teams squared off hours before the end of 2016. 

Unfortunately, the match itself was a bit of a damp squib. Klopp came out with a defensive posture, replacing Divock Origi with Emre Can in his starting XI, and with a game-plan whose centrepiece was cluttering the middle of the park so that neither team could “win the right to play” for as long as possible. Meanwhile, Pep’s own posture betrayed a measure of fear and hesitancy. He opened with Fernandinho and Touré playing relatively deep. More curiously, even after conceding to Liverpool’s first real attack, City never fully committed, playing overcautiously until the final minutes and creating essentially no real chances. 

Perhaps still feeling the Leicester sting, Guardiola refused to make significant tactical changes that might leave him more exposed in the back. As Klopp’s men managed for long stretches to devolve the game into meandering mayhem, wherein Touré and Fernandinho could not hold on to the ball and Silva and De Bruyne were cut off from the action, City looked like a sluggish brute with no insight into their own principles of action, losing 0-1. 

The immediate upshot of this dispiriting display was that, stunningly, City fell out of the top four just at the mid-point of the season. It is at that stage that Dyspeptic Pep started tossing passive-aggressive retorts at the English media, a section of which exhibited a lack of interest in football intricacies and a thirst for sensationalist leads that seemed to evoke thinly veiled derision in this continental aesthete. 

“I have to learn the rules here,” he kept saying in lieu of Mourinho-style overt whining about officials. The man who in a seven-year career had beaten Manchester United in two Champions’ League finals, as well as Arsenal (three times) and Chelsea and Manchester United earlier in knock-out stages, was suddenly bemused by the Premier League. 

The truth was that Guardiola was facing the steepest challenge of his career. There was something about the Premier League he wasn’t getting, or at least something about his system at City that stopped it from imposing itself as it had in Spain and Germany. But what? A probing investigation was needed. 

It took Guardiola less than a week to reach a diagnosis. City’s loss to Chelsea was unfortunate – they had outplayed Chelsea on the day and should have won. But against both Tottenham and Liverpool, the loss was justified on the pitch. The common theme in both defeats was Pochettino and Klopp managing to “truncate” Guardiola’s team, stopping the ball from making its way to City’s most dangerous players up front. Spurs did it incredibly low in the pitch, truncating City at the joint between defence and midfield, sometimes between keeper and central defence. The Reds did it much higher up, truncating City at the joint between holding and attacking midfielders. But the challenge posed by both was the same: to create a smoother continuity across the formation. 

Thus far in the season, City had tried a traditional 4-3-3 (with backward-pointing midfield triangle), a heterodox 4-3-3 (with forward-pointing midfield triangle), a traditional 3-4-3 (with wingbacks), a heterodox 3-2-2-3 (without wingbacks), and a bizarre 4-6-0. Once the initial run of wins was over, none of it worked. On January 6, for Pep’s inaugural FA Cup match, City showed up to the London Stadium playing De Bruyne unusually deep, almost like a holding midfielder, and Sterling unusually high and central, almost like a second striker. In other words, a diamond. Or rather, a kind of tilted trapezoid:


If this looks vaguely familiar, it’s because diamonds are forever. Consider this blast from the past:


For the first time in weeks, City were truly compelling, hammering West Ham 5-0 in the sort of shock-and-awe display that characterised the Guardiola Doctrine in Spain and Germany.

The shock-and-awe doctrine was developed at the Pentagon in the mid-90s. Both in la Liga and the Bundesliga its football analogue was highly effective: the idea was to win the game in the first 30 minutes by sapping the spirit out of the opposition. In this strategy, however, getting the first goal is crucial.

In his first year at Barça, Pep’s team got 2.65 points per game in which they scored first, as opposed to 1.5 points per game in which they did not; in his first year at Bayern, they got 2.77 points per game in which they scored first and only 1.77 points per game in which they did not. Fortunately, then, Pep’s teams scored first in 77% and 75% of their matches respectively. 

In contrast, City’s performance over the first half of the season was doubly problematic. First, although they managed 2.69 points per game in which they scored first, they scored first in only 55% of matches, which means that they didn’t capitalise as much on this healthy rate of return. Secondly, they got only 1.05 points per game in which they did not score first, which means they had harder time coming back from behind. 

Both problems were due to a level of grit in Premier League teams that Guardiola was unaccustomed to. Whatever its faults, the Premier League is special in that no matter who you are, at least a dozen of its teams believe to their core that they can get a result against you. West Ham, meanwhile, capitulated after half an hour the way so many Spanish and German teams have before them. Finally, a blueprint emerged: play the diamond and score first. 

This renewed sense of hope was shattered when City fielded the exact same diamond formation at Goodison Park and were humiliated 4-0 by Everton. City opened well, with the diamond doing precisely what it was meant to, frustrating Everton’s high press and ensuring 71% first-half possession. But Guardiola’s undoing was his persistence with the high line, which proved porous over and over. With all their domination of the pitch between the boxes, City did not create any strong chances and the match was closer to ending 6-0 than 4-1.

In one memorable second-half moment, Guardiola stopped watching the game and appeared to retreat deep into his own psyche. At the post-game press conference, he stressed that Tottenham were the only team to have outplayed City all season. This was a genuinely alarming moment, manifesting a shocking degree of stubbornness and inflexibility as though the continuous frustration of his wishes had triggered an insulation from reality. 

Three thinkers had an inordinate impact on the twentieth century: Marx set the agenda for the main thread of human history during the century, Einstein framed our conception of the cosmos’s structure and Freud revolutionised our way of understanding ourselves. Freud’s best-known contribution was his emphasis on the role of the unconscious in shaping our behaviour, but his more theoretical contribution was his understanding of the mind as the theatre of interaction between two forces, which he called the “pleasure principle” and the “reality principle”. The pleasure principle instructs us to seek pleasure and avoid pain. Meanwhile, the reality principle is supposed – in a well-functioning psychological set-up – to take into account the world’s independence from one’s will and help the mind generate sophisticated strategies for obtaining greater long-term satisfaction instead of short-term pleasure (this is the capacity for delayed gratification). 

When your two year old throws regular tantrums, it’s only his or her healthy way of dealing with the painful realisation of the reality principle, of grieving the lost paradise of sole governance by the pleasure principle. A person who never absorbs the reality principle is bound to remain at the level of cognitive functioning proper to two year olds, hopping from one pleasure-seeking act to the next in a torrent of short attention span and disdain for facts.

It is this theory that Guardiola’s post-match comments about outplaying an opponent who just beat him 4-0 conjured. It portrayed the image of a person cracking under the discovery of something we might call the manager-reality principle. The high line that worked in Catalonia and Bavaria was evidently not working in England, and Guardiola looked like someone who could not contemplate a football style made 90% of aesthetic purity and 10% of reality-heeding pragmatism. 

But Pep wasn’t cracking at all. He kept experimenting in search of a solution. His diagnosis of the problem remained constant: poor fluency in the lower part of the pitch made his team too easy to truncate. It’s just that the diamond didn’t establish sufficient continuity. Something more radical was needed: his best ball-handling players – Silva, De Bruyne, and to some extent Yaya Touré – needed to get the ball earlier and deeper. They needed to be close enough to the central defenders to bail them out. 

Thus came into being the era of City’s seventh formation for the season: a kind of 4-1-4-1 featuring Touré as axis man (the “Guardiola position”) flanked by Silva and De Bruyne playing almost like holding midfielders. It is this system that would ultimately win Guardiola’s heart. Its first test was Tottenham on January 21. The experiment was a big success: time and again City went long over the press – just as Guardiola’s Bayern had done in its first encounter with Klopp’s Dortmund – and had a dozen real chances before half-time, with none of note for Spurs. After going 2-0 up, a combination of weak refereeing and weaker defending (Bravo conceding a 16th goal from the previous 24 shots he’d faced) led to a 2-2 draw. The result was disappointing, perhaps, but the display was revitalising. 

City played the same 4-1-4-1 formation in a dazzling 4-0 win at West Ham, in which Guardiola, doubling down on the feeling of rebirth, introduced some revolutionary personnel changes to inhabit his new system: Gabriel Jesus replaced Agüero up front and Caballero replaced Bravo in goal; Sané seemed to have supplanted Nolito definitively on the left wing. Later, Fernandinho would start playing right-back (and once, left-back) in this system. 

The new system’s success was maintained against Swansea and Bournemouth, though in both matches City’s players seemed to lack urgency, certainly after the first half-hour.  And at Bournemouth Guardiola’s bad luck continued, as Gabriel Jesus fractured a metatarsal just when Guardiola thought he had finally solidified his first line of defence (this on the heels of losing Gündoğan in December, just after thinking he had found his axis man, and missing De Bruyne for much of October, just after thinking he had found City’s future Dionysian motor). 

The decisive moment for the 4-1-4-1, proving it was the right format for this City side, came with a scintillating 5-3 home win over Monaco. City had to content themselves with Agüero in the Gabriel Jesus role, albeit a somewhat reformed Agüero suddenly wise to the profits of pressing opposition defenders. This match also showcased a device we might term the backfield pentagon. As soon as Monaco’s high press started distressing City’s defense, it morphed with speedy discipline thus:


The device was not a sweeping success, as City still conceded three goals (plus an unconverted penalty) at home. Nonetheless, City’s mettle in coming back from a 3-2 deficit that seemed to doom the tie, and their spectacular offensive prowess in one of the best Champions League games in years, breathed new life into Guardiola’s project. 

Alas, a team’s shape is only part of what makes it work well. The right interrelation between a plurality of parts can enhance the whole’s performance, but the parts’ independent contribution to the overall performance is at least of equal significance. 

And just at this stage of the season, as everything started looking like it was coming together, and all three competitions were in principle live targets, City’s players seemed to have been afflicted with an inexplicable loss of energy. Superior quality still ensured laboured wins over Huddersfield at home (5-1) and Sunderland away (2-0), but Stoke proved too formidable an adversary for a lethargic City, forcing a drab 0-0 at the Etihad. 

The real let-down came on the ides of March, when City were knocked out of the Champions League with a 3-1 defeat at Monaco’s Stade Louis II. The first half saw City struggle under Monaco’s frenzied domination. It took the Monegasques less than half an hour to erase City’s two-goal first-leg advantage, with Kylian Mbappé and Fabinho enthusiastically capitalising on laxity from Aleksandar Kolarov and Sané. It could have been worse – City were on the ropes for much of the first half. They came out with more fight at the start of the second half and finally went ahead again through Sané, but when it was cancelled out by an Tiemoué Bakayoko header minutes later, the whole stadium knew City were doomed. No real chance was registered in the remaining quarter-hour. 

“We weren’t there in the first half,” Guardiola diagnosed succinctly after the match. But one was entitled to wonder: Why not? In the shadow of Conte’s feverish mania on Chelsea’s sidelines, another disappointing fact about Guardiola came into sharp relief: he has not been the master motivator, the catapult of human focus, that some of his peers have been. 

Four days after the Monaco defeat, City were effectively eliminated from the title race as well after a thrilling 1-1 draw at home to Liverpool. As in the New Year’s Eve match, Pep seemed to have no answer to the frenetic chaos Klopp created in midfield, with no team managing to “win the right to play” for long stretches. Only an hour into the match (so around 150 minutes of game time against Liverpool), and trailing after a Milner penalty, did Guardiola finally find a solution: in the 64th minute Sagna came on with complicated instructions evidently requiring a preliminary sideline seminar, whereafter De Bruyne moved to the right wing, Clichy pushed up almost to the same latitude, Sané became a second striker and Sterling drifted center into a number 10 position. With a sparsely populated midfield and a concentration of numbers around the Liverpool box, City became very vulnerable at the back for the remaining half-hour, but it took only five minutes after Sagna’s entrance for Agüero to convert a cross coming precisely from De Bruyne on the right wing. 

Still, the draw put City 12 points behind Chelsea with only 10 matches to go – effectively out of title contention. City’s last realistic shot at a trophy was in the FA Cup, where they faced Middlesbrough in the quarter-final. For the first time in the Cup campaign, City showed up in full force right from the start. The result was a comfortable 2-0 victory, with Agüero scoring his 24th goal of the season – not bad for a striker who had increasingly come to look like a back-up. (Agüero would finish the season with 34 goals in all competitions, the most prolific season of his career.)

With City out of the main two competitions they had been chasing, Guardiola could continue perfecting the 4-1-4-1 or alternatively to use the remainder of the season for further experimentation. He chose the latter.  

The seminal figure in modern military theory is doubtless the 19th-century Prussian general Carl von Clausewitz, whose book On War served as a textbook for generations of geopolitical and military analysts. It covers everything from the nitty-gritty of Napoleonic strategy to the role of personal character traits such as boldness and perseverance in a soldier’s profile. Curiously, given its subject matter, it is replete with life-clarifying insights and clever aphorisms, making it a pleasure to read even if you’re not necessarily a war enthusiast. 

Among Clausewitz’s fundamental insights is that there is nothing more misleading than to think of battle as a polar encounter between two symmetric forces, attack and defence. The mistake is to think there is symmetry between attack and defence. On the contrary, claims Clausewitz, defence is vastly superior to attack. He reasons that if attack and defence were symmetric, battle would be continuous in war. But in fact wars are long periods of inaction punctured by occasional battles, and the reason for this is that the attacking force must think a lot in advance – come up with something truly extraordinary – to overcome its inbuilt deficit against the defence. 

As a young man, Clausewitz joined Napoleon’s army in its catastrophic invasion of Russia and witnessed first-hand how the triumphant grande armée, crossing the Nemen into Russian territory with upward of 700,000 soldiers, left six months later with its tail between its legs, counting little over 20,000 men. Most striking was the limited effort the Russians had to invest to defeat Napoleon so decisively: all they did is patiently retreat further and further and let General Winter take care of the French. The asymmetry of defence and offence had thereafter become a foundation in his theory of war. 

Managers such as José Mourinho, Diego Simeone, and Antonio Conte are inveterate Clausewitzians. The constitutive superiority of defence over offence is the bedrock of their football philosophy. Guardiola probably appreciates this asymmetry, but believes in his ability to come up with the “something extraordinary” needed to overcome it, and anyway finds no joy in sitting tight and waiting for the opposition to make mistakes. (Why not play more pragmatically? “I deserve to be happy,” he answered in January.) Coming up with the extraordinary spark that would undo a defence’s inbuilt advantage – that’s what it’s all about for Guardiola. 

And yet, starting on April 2, and for the entirety of April (seven matches in total), Guardiola decided to drop the 4-1-4-1 and pay his respects to Clausewitz’s foundational insight by padding his defensive midfield and playing a 4-2-3-1 system. The premiere was away to Arsenal, featuring De Bruyne and Fernandinho as bona fide holding midfielders and Silva as a classic number 10. The 2-2 draw was not impressive, and City enjoyed only 56% possession, as against an average of 69% during the 4-1-4-1 era. 

But Guardiola stuck with this system in their next league match, a rather sad display of offensive impotence in front of Chelsea’s defense. Setting aside a fortuitous goal gifted to Agüero by Thibaut Courtois, City – this time with Fabian Delph in central midfield and De Bruyne instead of Sterling on the right wing – managed to generate no real threat for 90 minutes. After Chelsea went 2-1 up in the 35th minute, Conte decided to lock it down and keep ambitious ventures to a minimum; City looked embarrassingly uninventive for the remaining hour of football and did not produce so much as one full chance on the day. 

Worryingly, with this defeat City garnered a grand total of two out eighteen possible points from matches against top-four teams this season (and 11 out of 36 possible points against top-seven teams). Potentially a third of the way through the project, the feeling is that Guardiola’s attack is very far from being resourceful enough to overcome the built-in advantage of Conte’s and Pochettino’s Clausewitzian outfits. 

It’s not a 3-1 home victory over Hull and a sleepy 3-0 win away to a Southampton side who were pretty much on the beach that would change that feeling. What was needed was a victory over Chelsea in the FA Cup final – as soon as free-falling Arsenal had been dispatched in the semi. 

The match opened in a familiar way: in the first half hour, a 4-2-3-1 City with De Bruyne as right winger had 73% of possession but not a half-chance to show for it. A game-ending injury to Silva and a galvanising half-time speech from Wenger meant Arsenal were in the driving seat in the second half. Again Guardiola’s men became inexplicably languid, seeming dispirited and lustreless. Although Agüero managed to convert an Aaron Ramsey error into a goal against the run of play, the lead soon vanished to lackadaisical defending, notably from Otamendi. Arsenal looked the more dangerous side in extra-time, winning 2-1 through a scrappy Sánchez goal.

With this defeat – on la Diada de Sant Jordi, St George’s Day, the national day of Catalunya and England  – Guardiola’s first season in England became officially a thorough disappointment: out of the Champions League in the round of 16, not in real contention for a Premier League title since November, a semi-final exit in the FA Cup – this is not how it was all supposed to play out. 

On the last day of April, Guardiola started the match at Middlesbrough with his ninth tactic for the year: 3-4-1-2 intended to make room for an Agüero-Gabriel Jesus duo upfront (and featuring Navas and Clichy as wing-backs). The first half was so disappointing in its toothless complacency that shortly after half-time Guardiola brought on Sané and Sterling together and reverted to more a familiar formation – just in time to rescue a 2-2 draw against the team who finished second bottom.

Guardiola reverted to the 4-1-4-1 in May and stuck to it for the season’s remaining four matches, winning them all convincingly. There was still a new experimental twist here: in March, Silva and De Bruyne were playing in parallel somewhere between Touré and the wingers; in May, Silva played rather like a number 10 while De Bruyne was given the improvisational freedom to roam across the pitch and sow mercurial fancy in the opposition half. This new variant of the 4-1-4-1, with De Bruyne as the Joker, was a direct response from Guardiola to City’s consistent inability, in April as well as earlier, to unpick sophisticated conservative opposition. 

This tenth and final formation of the season was the culmination of a long process of experimentation and adaptation. With this final gambit, Pep also returned to his Nietzschean fundamentals: a stable, finely tuned Apollonian superstructure enveloping a buzzing Dionysian element of caprice. 

Moreover, he went back to his initial intuition, brewing since the September 24 experimentum crucis, that De Bruyne is the right man in the squad to don Dionysus’s mantle. And indeed, De Bruyne ended the season with 18 assists, the most ever recorded by a Premier League player in a season (though far short of the 29 Messi dished out for Pep in the 2011-12 season for Barça). 

With Gabriel Jesus and his maniacal press back in the side (playing on the right wing when Agüero is also included); with Fernandinho growing into an almost respectable and certainly inventive right-back; with an uninjured Kompany infusing calm competence into central defence; and, it must be said, with soft and often disinterested opposition, City performed wonderfully in those final four matches: a 5-0 demolition of Crystal Palace, a 2-1 revenge against Leicester, a 3-1 total domination of West Brom (and their six-man defence) and a 5-0 formality at Watford. 

Obviously, none of this rescued the season from what it was: manifest failure. It looks even worse when we consider that Antonio Conte, despite inheriting a demoralised team that had finished tenth the previous season and getting almost none of the players he wanted, managed to win the league at a canter. He did so in large part not through flashy new signings (though his bosses did gift him, with the signature of N’Golo Kanté, the Premier League’s best player over the past two years); rather, Conte simply rebooted the known talents of Hazard, Diego Costa and Pedro and unearthed unknown qualities in Victor Moses and Marcos Alonso. This Conte side garnered 15 points more than Guardiola’s, scoring more and conceding fewer, with a fraction of the investment and just as short a learning curb.

But Conte’s gambit is Clausewitzian, it enjoys the built-in advantage of defence that Guardiola’s mandate must steadily develop to overcome over time. Guardiola’s juego de posición style requires players accustomed to running and banging to concentrate and reimagine the pitch. Designed to produce rather than react, and produce in accordance with a very subtle recipe, it requires a long period of rewiring players’ football brains. In other words, it may be that much of what has transpired at City this season remains hidden, because it takes the elusive form of slowly growing dispositions and potentialities, ones whose full fruition will take time to emerge. 

It is perhaps for all these reasons that there is virtually no agitation around the Etihad over this disappointing first season of the Guardiola era. The hope is that the very limited success on the surface belies an underlying progress as yet unrealised – that a compelling, fluid, magnificent City team will surface next season with perfect instincts and well-equilibrated forces of organisation and unpredictability. Perhaps the opening words of the poem “Let Me Say,” by Miquel Martí i Pol, Guardiola’s favourite poet, capture best the current mood at Manchester City: “Let me say that it’s now time to forget / that it’s now time to believe in purity again / and to repaint the houses in cheerful colours.”