clock menu more-arrow no yes

Filed under:

What is Chelsea’s best midfield?

Statistical comparison of Chelsea’s midfield combinations

Chelsea Training Session Photo by Darren Walsh/Chelsea FC via Getty Images

Chelsea did a terrific job at filling the holes in the squad during the transfer window but still left one gaping one, midfield. Frank Lampard may or may not be targeting Declan Rice for that role, but Rice will not be a Chelsea player for the foreseeable future and we will have to make do with our current options for at least this season, the top three of whom are N’Golo Kante, Mateo Kovacic, Jorginho in some order and some combination.

N’Golo Kante has a valid argument for being the only consistently world-class player at the club but he has been blighted by injuries ever since he was forced to play the Europa League final on injections. Mateo Kovacic was the team’s most improved player last season and the club’s Player of the Year. And Jorginho is Jorginho. (In Billy Gilmour, the team has a top-class young talent too.)

However, Lampard has not settled on a first choice midfield combination due to the profiles involved, and that problem may yet continue this season unless someone takes a dramatic step forward.

But just how have Chelsea’s midfield combinations performed under Lampard? We will attempt to explore that question, though do keep in mind that performances are conditioned by other factors such as opponents, teammates, substitutions and momentum too, which are not accounted for in this guide.


Overview

Chelsea have used seven (7) primary combinations in midfield, which can be seen in the graphic above. Ross Barkley has since left on loan, but let’s assume that we can replace him with Kai Havertz, who is more active off the ball, better on it and has a similar set of physical tools.

Again, some discretion must be used when looking at the graphs. The midfield with Jorginho, Kovacic and Kante might rank poorly in a lot of metrics but this can be attributed to the quality of opponents. Similarly, the midfield with Jorginho at the base behind two attacking 8s was usually deployed against pretty average teams at home. When Chelsea used this midfield against a good team, the result was a thumping 3-0 defeat. Both sample size and quality of opponents are big factors to keep in mind.

Notes: A sample size of seven games was deemed large enough for the purpose of this study. The term “Jorginho DM” refers to the system with Jorginho playing behind two attacking number 8s. Midfield combinations such as Gilmour-Jorginho, Gilmour-Kovacic, and others have not been seen enough to be judged and as such have been excluded from this analysis.


Goal difference vs expected goal difference

Set-ups featuring Kante and Jorginho playing together or them being flanked by attacking number 8s rank the best in this regard. A similar set-up with Gilmour sitting deep has produced excellent results but the underlying numbers do not look great.

Jorginho and Kovacic are not a net-positive contribution, Kante and Kovacic together are net-positive by a very small margin.


Quality and quantity of shots taken and conceded

Jorginho - Mount - Barkley ranks excellently again and Kante - Kovacic are good from a shots taken/conceded perspective too.

Kante and Jorginho together are not the best combination in terms of creating shots or suppressing them.

Chelsea create few chances with “Jovacic” in midfield but the chances created tend to be of higher quality than average. Likewise, the chances conceded tend to be fewer in number but of higher quality too.

Kante and Kovacic make for an interesting pairing in midfield. Games involving them typically tend not to have many high quality chances. On average, Chelsea create many low-quality chances while conceding few low-quality chances.

The three games with Gilmour playing as the deepest midfielder were fun, to say the very least. Chelsea created and conceded a whole ton of high-quality chances. This strategy is a lot of things, sustainable is not one of them.


Passing, ball progression and territorial dominance

This is the section where a lot of pre-conceived notions and beliefs (including those held by tis writer) can be proven wrong.

It is surprising to see the midfield with Jorginho and Kovacic rank only in the middle. For a team with two players known for the ball progression, there just were not enough passes into the final third or penalty box. The system with Gilmour playing deeper is broadly similar, too. It appears as if the team converted an unsustainable amount of final third passes into goal-scoring opportunities with Gilmour playing as a number 6.

Chelsea’s passing into the final third and penalty box has been phenomenal with Kante and Kovacic playing together. That pivot has many flaws but passing is not one of them.

To build on their excellent final third passing, Kovacic and Kante help the team carry the ball forward too. Interestingly, Kante and Jorginho just do not progress the ball well. Even accounting for the quality of opponents, the numbers are pretty mediocre.

The “attacking 8s” setup used by Lampard last season yielded some superb ball progression depending on the player used in the deepest position.

Another way to measure progression is by territorial dominance, which builds on the assumption that teams that progress the ball well spend more time in the final third than in midfield. Similarly, they take a high percentage of their final third touches in the penalty box.

The fact that Chelsea do not progress the ball well with Jorginho and Kovacic is reflected in this chart. When Chelsea pair Kante with one of Jorginho or Kovacic, the team gets the ball into the final third efficiently — though are not as good at taking the ball into the penalty box.

The three-man midfield with Jorginho, Kovacic and Kante are perhaps the biggest surprise here. Despite the strength of the opponents, the numbers are superb.


Defensive activity

As always, it must be noted that defensive activity and quality are not the same. This chart is designated to indicate how intense Chelsea are when winning the ball back in midfield and the final third.

The main takeaway from this chart is the performance of the Kante - Jorginho pivot. Using a traditional defensive midfielder allows Kante to do what he does best: follow the ball and use his intelligence and tackling ability to win the ball back.

Kante and Kovacic are both excellent tacklers and it is reflected here. That they both prefer to follow the ball rather than sitting deep is perhaps why Chelsea’s pressing is not as successful as with other midfields.


Opponent final third passing, ball progression and territorial dominance

So far, we have only looked at how Chelsea do on the ball with different midfield combinations. Football, however, is a game with two teams and the performance of Chelsea’s midfield affects the opponent’s performance too.

Chelsea played three easy games at home with the Jorginho - Barkley - Mount midfield, which is reflected by the low tally of opponent passes into both the final third and the penalty box. It is a surprise to see how permeable Chelsea’s Jorginho - Kovacic - Kante midfield is, too.

Chelsea struggle to stop opponent passes into the penalty box with a midfield pairing of Kante and one of Jorginho/Kovacic, although they do a good job of stopping final third passes as a whole.

Playing with a designated defensive midfielder, be it Kante or Jorginho, has produced excellent results for Chelsea in terms of preventing opponent ball progression. The idea that Jorginho and Kovacic together are poor defensively is nothing new, but this further reaffirms the notion.

The mobility of Kante and Kovacic helps Chelsea prevent opponents from carrying the ball while having Jorginho next to Kante allows Chelsea to prevent passes better.

Kante and Jorginho playing together is perhaps the weakest combination for Chelsea in terms of allowing opponent territorial dominance. Opponents convert plenty of their midfield possession into plays in the final third and move the ball from the final third into the penalty box with ease.

Surprisingly, Chelsea do well in terms of territorial dominance when Jorginho and Kovacic play together. While Chelsea concede plenty of high quality chances with the two in midfield, the time spent by opponents in dangerous areas against this midfield is better than average.


So, what is Chelsea’s best midfield?

At this moment, Chelsea do not have an ideal midfield that could be used against every opponent. The best Lampard can do right now is use different midfields against different teams, based on stylistic and qualitative match-ups.

Keeping that in mind, here are Chelsea’s midfield combinations. ranked in descending order of metrics, quality and balance, where 7 is the least impressive and 1 the most impressive.

7. Jorginho - Kovacic:

In a lot of ways, this midfield sums up the Arsenal-ification of Chelsea. While they look to be doing a lot of things well, the reality is far different. If the midfield had to be described in one word, it would be superficial.

Should not be used outside emergencies.

6. Jorginho - Kovacic - Kante:

While this midfield has been used mostly against strong opponents and consequently has poor numbers, there is just not enough balance among the three. There is not enough match-winning quality and while that is not always a problem (see: Liverpool’s midfield), the midfield is too ball-dominant to not contribute anything going forward.

Should not be used often and if it is used, it should only be used in big games.

5. Gilmour lone holder in 4-3-3 behind two attacking mids:

While this midfield yielded three positive results, including an aggregate 6-0 win over the Merseyside teams, their numbers are just not sustainable enough. While the sample size is way too small to make sweeping judgements, the initial signs are not great in terms of underlying metrics.

4. Jorginho lone holder in 4-3-3 behind two attacking mids:

The numbers posted by this midfield are phenomenal to say the least. However, how much is down to the fact 3 of the 4 opponents were cannon fodder at home? The fact that Chelsea got thumped in the only game against a moderately strong opponent is not particularly promising. Again, the sample size is pretty small.

Against weak times at home where Chelsea can be expected to have 70% of the ball, this midfield is a solid option.

3. Kante lone holder in 4-3-3 behind two attacking midfielders:

In terms of results and underlying metrics, Chelsea have done excellently with this midfield. However, Lampard seems to have discarded this options, and as a consequence, it cannot be viewed as a serious alternative unless Jorginho and Kovacic get injured.

This midfield should work in theory against all kinds of teams, and has in reality too. The wins against Manchester City, Leicester and Watford underline its versatility.

2. Kante - Kovacic:

These two are arguably Chelsea’s best midfielders but they do not mesh particularly well. Both prefer following the ball at the same time instead of taking turns to sit deep and this leaves acres of space in midfield. However, Chelsea’s ball progression numbers with the two playing together are promising.

This is a reasonably good option against teams with slow and unathletic midfielders.

1. Kante - Jorginho:

On paper, this is Chelsea’s most balanced midfield. Jorginho is capable of sitting deeper, which allows Kante ro roam box-to-box, which he is best at. However, Jorginho is the biggest defensive liability among Chelsea’s midfielders and sooner than later, this weakness will be exploited by opponents.

This midfield should work against teams that will not press much.

Tottenham Hotspur v Chelsea FC - Premier League Photo by Darren Walsh/Chelsea FC via Getty Images

Agree or disagree, feedback of any kind will be gladly welcomed as always.

Chelsea FC: Champions League

Chelsea 4-0 Malmö FF, Champions League: Tactical Analysis

Around Europe

Even Chelsea B are beating Tottenham Hotspur

Chelsea FC: Injury News

Romelu Lukaku expected to miss ‘at least two to three weeks’ — report