x icon

Our knowledge is at your service

The recently concluded Premier League season is in one way eerily similar to the 2018-19 campaign when City were crowned champions, only 1 point ahead of runners-up Liverpool. The fight for top spot was undoubtedly a thrilling affair for Premier League fans again this year, with Klopp and Guardiola’s renewed clash not boring in the least. But was the result fair? In the first part of our two-part analysis, we compare the quality of attacking play, using both classic and advanced statistical indicators, as well as descriptive measures. In Part 2 we will focus on the defensive play and take a closer look at how the two teams faired in terms of off the ball movements.

Nerve-wracking end of season

While the final league standings will of course be determined primarily by points earned, it was not inconceivable that a scenario of a points draw between the two rivals could occur. In this case, the goal difference and the number of goals scored could have been the deciding factor. , It was not the case at the end, but it was necessary to play with maximum concentration until the last minute and to do everything possible to score as many goals as possible.

For a long time it looked as if Liverpool would finish the season with more goals scored, but in the last few rounds the Manchester Blues have stepped up a gear (or more). At the end of April they sent Watford home with a loose 5, only to score 4 against Leeds United. After that came Newcastle, who have shown encouraging signs under Eddie Howe, and Wolves, who completely drained by the end of the season (as usual), both conceding 5 against them. While City were scoring goals at a rate that was almost beyond words, the Reds were doing the opposite in the last few weeks. The latter phenomenon, while perhaps not surprising to say the least, has been the result of a record 63 games played this season, which has on several occasions left its mark on what could be considered the decisive meetings in the final stages. The 1-1 draw against Tottenham, for example, was seen by many as flattering for Klopp’s side, not only because Díaz’s equaliser was a shot bounded off a defender, but also because the feeling was that Conte’s will and vision had prevailed almost throughout. In the end, City won the league by almost 100 goals in the whole PL season, with Liverpool scoring 94 times in total.

Comparision of expected goals

When comparing the attacking play of the teams, it is important to consider the expected goals (xG) indicator, which focuses on the quality of shooting chances and quantifies the probability of an average quality striker scoring from a given situation based on the historical pattern. It has been an ongoing theme for years, that City definitely need a classic goal scorer up front who is relentless in converting chances. In Haaland, such a centre forward has certainly arrived, but Guardiola cannot be criticised for having been so wasteful in the past season. Nothing proves this more than the fact that they scored 99 goals on a xG of 90.88. Only rivals Liverpool (93.75) managed a higher xG, but they rattled the net just as often (94) as the model would suggest they should have. In other words, while City, who won the league trophy, outscored their expected number of goals by almost 9%, the team from Merseyside "only" scored the ‘expected’ number.

It is also worth taking a close look at how expected goalsvaried over time, which gives us an overall picture of the teams’ consistency in attack. Firstly, the 5-match xG moving average chart shows that Manchester City have been much more hectic in creating quality chances (i.e. higher variation in xG). For a long time, the Manchester Blues’ average xG hovered around 2.0, before reaching a staggering 3.74 (!) a few days before New Year’s Eve. After that, there was a significant drop down to 1.53 in mid-February, for xG to climb back to and above 2.0 again for the rest of the campaign.

The Reds’ ceiling peaked at the 3.25 mid-point in early March, while the 1.62 xG moving average in the second half of April proved to be their lowest in the season. The chart shows that Liverpool developed higher quality chances in the period up to mid-December, and City’s attacks showed more potential until the end of January. This was followed by a period of Pool dominance from the end of winter until the first half of March, before Guardiola’s side again produced the better numbers in terms of attempts on goal in front of goal in the last two months.

Key players in attack – Manchester City

Although the team-level aggregate xG index for the whole season also has a lot to teach us, it would be a mistake not to take into account the individual productivity rates broken down by player. Kevin De Bruyne (KDB) has been one of the most influential players for the team that won their 6th PL title this season, both scoring the most goals (15) and providing the most assists (8). It should be added that City’s top scorer had a very strong season, but his underlying xG (5.39) was exceeded by his goals scored to an unprecedented degree. Instead of the 0.2 goals predicted by the model, he recorded 0.57 goals per 90 minutes, a nearly threefold improvement. Nothing shows the amazing potential of City’s attack better than the fact that 8 of their players have scored at least 7 goals. Raheem Sterling (13) and Riyad Mahrez (11) scored as many goals as fans could have hoped for based on the expected goals model, while Gabriel Jesus did not make the most of his opportunities (only 8 goals on a xG of 11.18).

The xG+xA parameter, calculated over 90 minutes, is perhaps the best indicator of the players’ end product, with the Algerian (0.8) and Brazilian (0.72) strikers performing best, but the aforementioned Sterling (0.69) and the equally impressive Phil Foden (0.62) did not disappoint either. The Belgian genius (0.5) and the excellent Ilkay Gündogan (0.54), who made a crucial contribution in the last game of the season, also reached the 0.5 mark. However, Jack Grealish, who has just arrived from Aston Villa for €117.5 million, has not had the brightest season yet (3 goals, 3 assists). The demands of Guardiola’s system are not easy to master, so we will probably see more of him in the coming years. His forward stats don’t look too bad though, as the 26-year-old English attacker has managed to produce 5.89 xG and 5.00 xA. He, however, certainly has room for improvement in terms of finishing.

Rodri, who is unstoppable at the No. 6 position, has not only been brilliant at stopping opposition’s attacks, but has also made a massive contribution to his side’s title success both in terms of putting himself in great goalscoring position (3.2 xG) and scoring (7 goals). Joao Cancelo has also had a great season, although 7 assists alone would not necessarily predict him to be named in the team of the year by many experts. What is certain is that only Ederson has played more minutes in the PL than him, meaning that Guardiola sees the Portuguese all-rounder as an indispensable cog.

Of course, it is not only goals and assists, as well as xG and xA, that characterise the extent of a player’s role in attack, but there are other important parameters. One of these is the number of progressive passes per 90 minutes, of which Zinchenko (10.93), Cancelo (9.94) and Laporte (9.15) were the top performers among those with at least 1,000 minutes. In terms of key passes - passes that create an immediate clear goal-scoring opportunity for a teammate - the Ukrainian winger (0.88) is the fourth with KDB (1.03), Grealish (0.92), Sterling (0.91) in the top 3. Finally, Laporte (11.43), Walker (10.2) and Rúben Dias (9.51) lead the way in terms of passing balls into the attacking third.

In terms of the time spent on the pitch by the players responsible for scoring goals and organising attacks, the Spanish genius has done a great job of rotation, with only Bernardo Silva (3072) playing more than 2500 minutes, and most of the others between 2000 and 2500 minutes. Mahrez (1,642) has had slightly fewer opportunities this season than his post rivals, but his numbers give him every reason to be hopeful that he can carve out a more dominant role for himself next season. The fact that Cancelo was able to play 3,477 minutes was a huge plus for City, without his presence the attacking play would hardly have been as punchy and unpredictable.

Key players in attack - Liverpool

If it was true to say that City’s goals were fairly evenly distributed between the team’s players, then the opposite is true of Liverpool. No less than 57% (!) of the Reds’ league goals came from the Jota-Mané-Salah trio, and while the midfield was also a source of goals for the Mancunians, ’Pool were essentially unable to rely on anyone else when the trio were not having a good day. Except perhaps one, as Luis Díaz, bought from Porto for €47 million in the winter, has picked up the pace as smoothly as if he had been playing in the PL for years. The Colombian’s output is 4 goals and 2 assists, not bad at all for a debut half-season, especially considering he managed to do this in 1017 minutes of play.

Returning to the top scorers, perhaps the first to be singled out is Salah, who, although was unable to match his autumn season numbers after the Africa Cup of Nations (AFCON), still earned the title of the League’s joint top scorer (with Son) with 23 goals and nobodymanaged to provide more assists than him (13). His full-season xG+xA/90 of 0.93 is without exaggeration ‘world class’, although the lack of consistency means that this statement should be treated with caution.

Mané and Jota produced somewhat lower, but still excellent numbers not only in terms of xG (15.28 vs 16.43) but also in terms of goals scored (16 vs 15), and it also shows that both of them took advantage of their shooting chances with the expected effectiveness. The following statement would probably be questioned by even the most ardent Liverpool fans on first reading, but it is still true. The Portuguese was ahead of the Senegalese in both xA (0.15 vs 0.11) and effective assists (0.14 vs 0.06) per 90 minutes, and the difference between the two in terms of xG+xA is 0.16 (0.73 vs 0.57), which can hardly be called marginal. Divock Origi, who is about to leave the club as a cult hero in the summer, produced 3 goals in 157 minutes and although his 0.95 goals/90 index is undoubtedly par for the course, the small sample size suggests to use it with caution.

While at City we noted that the midfield did their bit both in building attacks and not infrequently finishing (think of Gündogan, Rodri and Silva), the situation is quite different for the Reds. Keita contributed only 3, Henderson and Oxlade-Chamberlain 2, Jones and Thiago 1, while Milner 0 goals, which is not very solid to say the least. These 6 players have a combined number of 12 assists, although 5 of those are from the captain and 4 from the Spaniard, with the others trailing well behind. On a per 90 minutes basis, only Keita’s xG+xA (0.27) reaches the 0.2 mark, which even Fernandinho (0.22) has managed to do for Manchester City, not to mention the midfielders (No. 8’s) for the Blues.

One of the main reasons for the lack of significant help from midfield in terms of forward play is the role of the full backs. Robertson’s runs up the left wing and his trademark crosses into the middle, and Trent Alexander-Arnold’s (TAA) phenomenal vision in the right half space are invaluable to Liverpool, and their role in Klopp’s system is unique and unparalleled. The Scot and the England defender finished the 2021-22 campaign with 3 and 2 goals respectively, while the 9 and 11 assists are far more representative of their key roles in the team. For TAA’s xA/90 of 0.36 it is very difficult task to find an adjective, perhaps frenetic is the best choice to describe this performance.

TAA’s place in the starting line-up at Liverpool is not in question, he gave 13.22 progressive passes and 0.94 key passes per 90 minutes, and made 8.43 passes into the attacking third. He is unbeatable within the teamin the former two, while in the latter he is only behind Thiago (12.08) and Milner (8.84). The Spanish conductor was also brilliant in the number of progressive passes (9.5), as was Robertson, who averaged 9.61 of them. Salah (0.85) and Díaz (0.71) are also on the podium in terms of key passes.

The German coach had much more leeway in rotating the squad this season than in the previous one, thanks mainly to the fact that he had to deal with far fewer injuries. Among the forwards, Mané (3026) has played the most minutes, but Salah is not far behind (2974) and Jota (2537) has also had a busy season. Díaz and Firmino (1,062) have had roughly equal opportunities to prove themselves. In the middle, the Fabinho (2480) Henderson (2781) duo have been the regular stable points, with the ’left 8 position’ typically occupied by Thiago (1639), whose magnificent performances clearly justified it.


As we have seen from the 5-matches rolling xG average of the two teams, one of the fundamental differences was consistency. While Liverpool created high quality chances in a relatively balanced manner match after match, Manchester City started the season relatively sleepy, but as Christmas approached they created more and more valuable chances at an incredible rate. For the Reds, the Salah-Mané-Jota trio have been by far the most prolific scorers in PL matches, but perhaps few would have guessed before the season started that they would be followed by Fabinho (along with Firmino). In comparison, City’s opponents certainly had to watch several players closely, as whether Rodri, Gündogan or Silva found themselves in a goalscoring position, they were ruthless in front of goal, not to mention the Blues’ strikers. In terms of progressive passing, the full backs produced excellent numbers for both teams, while the duo of Salah-TAA for the Mersey siders and the KDB-Grealish-Sterling trio for City deserve the highest praise at giving the most key passes. Last but not least, Guardiola distributed the minutes more evenly between his attackers, with Klopp not giving much rest to his most frequently chosen front-three.

(Cover photo: )