Who can play in the No 6 role?

Martin Zubimendi's decision to stay at Real Sociedad has raised questions over alternative options over at Anfield.

Reinforcements for defensive midfield will not arrive this transfer window - not from the Basque region, and not from anywhere else, according to James Pearce, who has reported that the failed move will be Liverpool's last and only approach for the position this summer.

With incomings unlikely, therefore, Slot and his coaching staff will be forced to search for replacements from within. From what we have seen during pre-season, there seem to be four candidates; Mac Allister, Jones, Gravenberch and Endo.

PlayerTop flight games played as a No 6Games played as a No 6 for Liverpool (all comps)
Mac Allister3724
Jones01
Gravenberch131
Endo11943

Stats via Squawka

In terms of an out-and-out defensive midfielder, Endo is by far the most experienced, playing as a holding midfielder for most of his career. He was utilised exclusively there by Klopp, racking up 43 appearances last season and becoming a mainstay in the team.

Mac Allister has experience in every midfield position, and has clocked up an even spread of appearances along the midfield spine since arriving in the Premier League, with 38 appearances in attacking midfield, 35 appearances in central midfield and 36 appearances in a deeper midfield position. Klopp came no closer to uncovering Alexis' best position than the Argentine's previous coaches, deploying him 22 times in central midfield and 24 times in a deeper position last season.

Jones and Gravenberch have accumulated only a handful of appearances in holding midfield, with the vast majority of the Dutchman's minutes in that position coming at Ajax. Jones is equally inexperienced in the No 6 role, only racking up one start in a league cup match in 2021.

In any traditional system with a sole holding midfielder, Endo or Mac Allister would take the cake. They boast far better defensive numbers per 90 minutes than Jones and Gravenberch, which can somewhat be explained by their deeper position last season.

PlayerTacklesGround duels wonInterceptionsBall recoveries
Mac Allister3.55.71.36.8
Jones2.14.40.65.6
Gravenberch2.15.21.26
Endo2.64.41.15.7

Stats via Squawka

Slot's system, however, uses a holding midfielder that is supported by another player in a double pivot. This more defensive midfielder must recover the ball quickly and intercept passes of course, this has not changed. But Slot seems to have placed more emphasis on playing out from the back and hitting the opposition with quick transitions.

It seems, then, that Slot may look for other attributes in a holding player. In stats per 90 minutes in the league such as passing and press resistance, Jones and Gravenberch make more of a case for themselves.

PlayerPass accuracy (%)Take-on success (%)Possession lostTimes tackled
Mac Allister87.86011.91.7
Jones91.56010.13.4
Gravenberch83.451.113.82.9
Endo88.466.711.21.5

Jones displayed excellent passing accuracy in midfield last year, recycling possession and keeping hold of the ball adeptly in that half-turn position behind Diaz. His passing decisions led to fewer possession losses than Mac Allister, although he was caught on the ball far more.

Gravenberch struggled for form in the league somewhat, but still put up good passing accuracy and tackle evasion. In the FA cup, his take-on success rose to 75%, while his possession lost and times tackled dropped to 10.5 and 3 respectively.

In this season's opener, however, the Dutchman put on a superb display as part of the double pivot. He was tidy in the tackle, intercepting opposition passes and contesting several ground duels.

Crucially, however, he was brave and willing to drive the ball forward. He made nine passes into the final third and two through balls, one of which Díaz should arguably have converted into a goal.

Gravenberch made a brilliant start to the season, and Slot has demonstrated that he is strongly considering his countryman for the No 6 role, but only time will tell whether this was a signal of intent or a one-match experiment.