Liverpool’s unbeaten start in the league came to an end last week, which prompted old insecurities in the supporters. A lot of finger-pointing and scapegoating has taken place since. When the starting line-ups were announced there were only two ways it was going to go – the Reds don’t lose, and nobody cares, or City beat us, and we all blame the midfield and Dejan Lovren. The Croatian is a topic for another day. I decided to look into how the team performs with different midfield set ups. I’ve been writing similar articles for a few seasons now but in case you haven’t read them here is how this works:
- The offensive and xOff ratings of a player are calculated based on goals scored/xG generated from the team while the player is on the field per 100 minutes
- The defensive and xDef ratings of a player are calculated based on goals conceded/xGA conceded from the team while the player is on the field per 100 minutes
- Opponent quality is based on points per game won by the opponent, points per game won by the opponent in the 5 games prior to facing LFC, home/away points per game won by the opponent depending on where the game is and points per game home/away in the last three games before playing the Reds. It is normalised, so that the easiest game has opponent quality 0, while the toughest has 1.
I will only be looking at the midfielders with over 500 minutes (sorry, Adam Lallana). Bobby Firmino is also not included as a midfielder even though he is at times. So we are down to the five Reds in the table below:
Gini Wijnaldum has played 83% of all minutes this season. He is by far the most reliable of the group. As a result, his numbers are right at the team average. Fabinho has been on the field for 70% of the minutes since making his debut, an encouraging number. It’s also important to note that some of Milner’s minutes are at fullback. Two numbers jump right at you when you glance at the table – Fabinho’s offensive rating and Naby’s defensive rating. The Reds are scoring three goals per 100 minutes with the Brazilian on the field, granted against weaker opposition. With the Guinean on the pitch, LFC is only conceding 0.25 goals per 100 minutes, more than two times less than with any other midfielder. The team also generates the highest xG with at least one of the newcomers playing. The obvious problem here is the low number of minutes played by both of them.
Not surprisingly the Reds “struggle” (if you can call scoring two goals per game that) with Henderson or Milner on the field. The team scored fewer goals and generated the fewest xG with the Englishmen playing compared to the other midfielders. To graph below illustrates the numbers in a more presentable way:
The vast majority of supporters had an uneasy feeling as soon as they read the starting line-ups against City. A flat midfield three of Gini, Henderson and Milner is less than inspiring. Klopp has tried it in several high profile games, and it has had mixed success. So what is LFCs best midfield?
All of Milner’s minutes here are as a midfielder. LFC only failed to outscore their opponent with the flat midfield from the Etihad. Klopp has trusted them in big games (opponent quality 0.77), and it hasn’t worked well enough for the Reds to be winning those matches. The difference between the flat three and a midfield two of Fabinho and Gini in a 4-2-3-1, which has primarily been used in theoretically easier matches is mind-blowing. 8 more goals scored in roughly the same number of minutes. Looking a bit deeper into the xG numbers:
One can argue, that the Gini, Hendo, Milner three has been relatively successful defensively. Limiting high profile opposition to 0.96 xG per 100 minutes is very good. But better teams will convert on low-quality chances, and we are left with not much to respond offensively, as happened against Guardiola’s side. Jurgen Klopp needs to find a more balanced approach in such matches. We still have three massive fixtures in the Premier League with additional ones in Europe. Judging by the second half at the Etihad, Fabinho might be the key.
Creativity from midfield
One of the big problems in the match at the Etihad was the lack of service to the front three. And even when the attackers managed to create something it was either created by themselves (the Mane chance) or the fullbacks (the goal). This has been a recurring theme this season. The only real creator from midfield from last season was the Ox, and he won’t be playing a significant role in the current campaign. Yes, I know James Milner holds the Champions League single-season assists record, but as great as he is, the terminator isn’t exactly going to create a chance out of nothing. Klopp and Edwards acknowledged the problem and brought in Naby Keita to fill that role, unfortunately, through no fault of his own the former Leipzig man has not been able to get on a run of form. The creativity numbers for our midfielders are underwhelming:
Fabinho and Gini are least involved in the build-up or finishing of attacks, which is expected as they both tend to play deeper and do no take set pieces. Milner is the most involved midfielder mainly due to him having the occasional dead ball and penalty duty. The five most played midfielders have managed five goals and three assists combined this season. Perhaps more worrying the combined expected assists of the group is just 5.03 for a total of 5602 minutes. If we exclude set pieces and look at xA from open play, it’s just 3.7 combined. The Ox himself managed 4.21 xA in about 1500 minutes last season. Naby Keita has the highest xG of the group despite having the by far the least minutes played, that just shows how important he can be for the title challenge. The system Jurgen Klopp has in place doesn’t necessarily depend on midfielders to create, as evident by our position in the league. However getting at least some creativity from the middle of the park from open play may prove to be the fine margin between a parade in May and wondering what the hell happened.