Window Shopping: Liverpool

At the top of the table the January transfer window tends to be disappointing. With value hard to come by, most big moves tend to feel like panic buys as opposed to good signings, or rumours that have been floating around for months usually fizzle out into nothing.

Just a week into the January though and Liverpool have pretty much thrown that out the window. The long running ‘sagas’ and rumours involving Virgil Van Dijk and Philippe Coutinho have both been quickly resolved. Liverpool have smashed a record in making the Dutchman the most expensive defender in the world, while Coutinho becomes the most expensive Premier League sale.

Two huge transfers in almost as many days has left lots to unpack and resolve for Liverpool, with lots of questions raised.

Now, this piece was initially just going to look at the Liverpool defence, whether it’s bad as made out and whether Virgil Van Dijk was the answer. Just as I was happy with what I wanted to cover, however, the Coutinho news broke – which is impossible to leave out of a Liverpool piece.

What I’m going to try to do is assess both Liverpool’s attack and defence. I’ll look at the previously mentioned questions to with the defence, as well what the Coutinho departure means and how/where they should reinvest the money.

Before getting started, a lot of articles helped guide and inform this and a lot of points made are going to sound similar. I thought it’d best to link all of these:

Another quick note before starting, my data isn’t synced up. My StrataBet data is up to date as of the 23rd of December, but my other data is up to date as of the 7th December. This isn’t ideal but my other data involves me putting in player minutes by hand and it’s not an enjoyable task.

Goalkeeper and Defence

At this point the fact that Liverpool don’t concede many chances but when they do they tend to be of high quality is starting to feel like a cliche. So much so that if The IT Crowd was made now Moss would be talking about that, rather than Arsenal always trying to walk it in.

Just to reinforce the point, however, Liverpool conceded the 2nd to least chances per game in the Premier League this season, but on average these chances are of the highest quality. In fact, their xG per chance against is the 2nd highest in all of Europe’s top five leagues, behind only Malaga.

Now, this has been mentioned many times that it’s a by product of their style. They press high so if a team does break their press they’re likely left with a good scoring oppurtunity. They’re almost the anti-Burnley (or Burnley are the anti-Liverpool). Burnley’s shots conceded tend to be of lower quality because they have so many players between the ball and goal, while Liverpool’s tend to be of higher quality because they have so few players between the ball and goal.

If this was just a by product though you’d expect other teams who press to have a similar struggle, but this isn’t the case. Mauricio Pochettino’s Tottenham have the 3rd lowest xG per chance in the Premier League, while Pep Guardiola’s Manchester City have the 10th lowest – with their 0.132 being pretty almost smack on the Premier League average of 0.134.

This becomes even more interesting when considering how Liverpool’s pressing game doesn’t appear to be so apparent this year. What Happened to the Klopp Press? by @WillTGM shows this, while looking at their passes per defensive actions (PPDA) also shows it.

For some reason, I decided to be awkward and have my PPDA as only passes and defensive actions in the opponents half, but it makes the same point. This season Liverpool’s PPDA has been 17.58, still the 7th highest in the Premier League, but is a sizable change from last years 13.14.

In Raphael Honigstein’s book ‘Klopp: Bring the Noise’ Liverpool assistant Zeljko Buvac talked about how Liverpool must improve in possession due to their struggles against low-block teams. Given that teams started to catch on that ‘gegenpressing’ was the ‘best playmaker’ they tended to sit back and not let Liverpool counter press.

This could have impacted the move to not pressing as highly, possibly allowing the opposition to bring the ball further out in attempt to disrupt the low block, create space in behind and possibly opt for more of a mid-press rather than a high one.

Watching some clips of Liverpool though, it felt as though the opposition could progress the ball through the midfield with minimal fuss. Granted it may not be best making this judgement looking at counter attacks and chances conceded, but it felt as though there were fairly big gaps between both horizontal and vertical lines, something you wouldn’t associate with a Klopp side.

I thought I’d have a look at some numbers to try and see if there was any truth in this.

One way to look at it was to just see the completion rate of passes that move the ball from the opposition’s half to Liverpool’s. In this Liverpool do well with the opposition having the 4th lowest completion in the league.

Another way of looking at it was to look just at the vertical passes moving the ball from the opposition’s half into Liverpool’s. Only Newcastle allow the opposition to attempt more of these passes than Liverpool in the Premier League this season. While the completion of these passes is pretty much the average (45.3% vs 44.9%), meaning the opposition complete the 5th most vertical passes into Liverpool’s half per match.

Given vertical passes exclude 15% either side of side of the pitch to focus on the centre, this could be a worry for Liverpool. Again, comparing to Tottenham and Manchester City, these numbers are much higher. Liverpool allow a vertical pass from the opposition’s half to their half 5.733 times per game, Spurs 3.267 and City 2.333 – with the completion rates of these passes also being significantly below average, 35.3% for Spurs, 31.5% for City.

Liverpool do seem slightly better at dealing with counter attacks than last season, but is still something they should be looking to improve. When they lose the ball it feels as though the opposition are only a pass away from running at Liverpool’s centre-backs left exposed.

Having Joe Gomez play right back can help with this, naturally a centre-back he seems more conservative than an actual full-back, and often makes it look like a back three when the opposition break.

It still feels as though they’re missing something in midfield, though, to manage the space in front of the centre-backs and offer more protection (more on this in a bit though). It feel as though Jordan Henderson is drawn towards the ball too much, while Emre Can doesn’t have the mobility to cover the spaces.

Below are some clips of opposition counter attacks this season. Now, these are only a few examples and lots don’t even lead to shots, but highlights how it feels Liverpool’s defence seems as though it’s left exposed.

This all leads to two questions.

  • Given how few chances Liverpool concede, is the high xG per chance that costly?
  • If it is, how do they fix it while maintaining their style of play?

To answer the first, this may not be that costly in the grand scheme of things given how few chances they concede, but coupled with some other defensive deficiencies it could start contributing to why they slip behind a title challenge and are left in a battle for Champions League qualification.

Liverpool have conceded 1.05 open play (non penalty or direct free kick) goals per game this season, which isn’t too bad – particularly when taking into account the 9 goals conceded against Manchester City and Tottenham. However, going back to 1999/2000, only 2 sides have won the title conceding more than 1 goal per game (not just non penalty or DFK goals) – both Manchester United in 2013 and 2000.

How they can lower the xG per chance without losing their style of play is hard to answer, though.

While it’s not a way to lower the xG value of chances, one area that has come under a lot of criticism and could be an area to invest in is in goal. Both Simon Mignolet and Loris Karius have been somewhat unreliable in their time, and a top class goalkeeper could help Liverpool shave a few goals off their season total and possibly even under perform their xG against.

This following bit is going to borrow a lot from Finding Liverpool a Good Goalkeeper by @thefutebolist.

There doesn’t feel like a great way to measure goalkeeper performance, save percentage doesn’t take into account the difficulty of the save while the difficulty of the save may not always correlate with the xG value of the shot.

It feels as though having the shot placement information would help with this, but with that not available I’ve opted to just throw a bunch of other things together and hope for the best.

@colintrainor has recently been posting lots of interesting images of goalkeepers post shot xG (including shot placement) and actual goals conceded which will be a better guide than the below.

Firstly, the danger zone. Chances within the danger zone are ones inside the width of the 6-yard-box and the height of the 18-yard-box. You can read the above linked piece by @thefutebolist for a diagram.

Following the common theme, Liverpool don’t concede many chances from within the danger zone. Their 1.37 per game is the 3rd lowest in the Premier League, their save percentage from these shots, however, is the 3rd lowest in the Premier League.

Having a quick look at the numbers and it seem as though to restrict goals from the danger zone teams do one of two things (except Napoli who seem to do both). Allow less shots from the danger zone and cut it off at the source, or make the shots in the danger zone much harder (through more numbers behind the ball or defensive pressure).

The top 10 teams for the least chances conceded within the danger zone have an average save percentage of 55%, the average for the top five leagues is 54% – these clubs seem to look to reduce the goals from within the danger zone by reducing the number of chances, as opposed to the quality of them, meaning there’s no problem conceding/saving at the average rate.

Liverpool’s save percentage in the danger zone meanwhile is 38%. The hard work in restricting these chances is pretty much being undone by having such a low save percentage.

This doesn’t take into account the difficulty of the shots, however. Using an xG model that only accounts for shots on target (as we’re judging ‘keepers) sees Liverpool concede 15% more goals than they should have from shots resulting in the danger area – the 5th most in the Premier League.

This led to trying to create expected saves(see An Introduction to Expected Saved (xS) by @sbourgenforcer). As the xG model only uses shots on target, the probability of a save is just going to be 1 – the xG value of the chance. Looking at all shots on target now and not just the danger zone sees Liverpool have the joint lowest expected save rating (Saves/Expected Saves) in the Premier League and the 2nd joint lowest in Europe’s top five leagues.

I thought it’d also be interesting to create some kind of expected shot on target model (xSOT) to try and see if a ‘keeper can influence shots and lead to more being astray than others. I’m not sure how helpful it is, but Liverpool do pretty well in this, conceding 90% of their ‘expected’ shots on target.

The above seems to state that there’s a problem with Liverpool’s goalkeepers though, and it’s an area that needs investment. While it may not be all Ederson’s doing, Manchester City signing the Brazilian is a good look at why it’s worth investing in a goalkeeper.

Manchester City’s xG is down from 0.8 to 0.63, but more impressively the xG Rating against has gone from conceding 17.9% more than ‘expected’ to 16.6% less. It’s important to point out that it may not be long enough into the season (or Ederson’s time at City) to see if this will regress to mean or is the effects of a talented goalkeeper, but it’s still encouraging.

While another case study shows Manchester United and Atletico Madrid conceding a lot less than expected in the last two seasons, most likely due to David De Gea and Jan Oblak, respectively. Investing in a goalkeeper can definitely be worth it.

The final part that I’ll look at for searching for a new ‘keeper is distribution. Another thing that Ederson has helped show in the Premier League is how useful a ‘keeper with good distribution can be. Not just being good receiving back passes and coping under pressure, but being able to pick out players between lines and effectively start attacks.

I’m not entirely sure how to rate ‘keeper distribution, so it’ll mostly just be looking at xP Rating and the percentage of their passes which are long balls – although I don’t have huge faith in using the xP Rating for goalkeepers.

As a guideline, Simon Mignolet’s passes long 44% of the time, Ederson is the lowest with 21% and the average is 60.5%.

Putting all of the above together, three interesting candidates are:

Oliver Baumann – 27 – Hoffenheim

Baumann has managed a danger zone save percentage of 66%, quite significantly above the average of 54% and the highest in the Bundesliga. This has led to Hoffenheim only conceding 80% of their expected danger zone goals and 85% of their total expected goals.

What’s encouraging is Baumann also played every minute of the 2016/17 campaign, where Hoffenheim also under performed their xG against, conceding just 79.9% of their ‘expected’ total.

Baumann’s passing numbers are pretty interesting. He goes long more than Mignolet, with 49.3%, and under performs his xP quite a bit with a rating of 0.876. Interestingly, though, splitting his xP into short and long sees his rating of short passes to be 1.238, while his long pass rating is 0.490.

There’s a very good chance this is something wrong with the model, but it makes it seem as though he’s comfortable on the ball and can pass it around the back, but isn’t too great going long.

A good age for a ‘keeper, plus good save numbers and the fact he plays in a team with a similar PPDA definitely make Baumann an option worth looking into – though looking more into his distribution would be necessary.

Thomas Strakosha – 22 – Lazio

At 22-years-old Strakosha is ridiculously young for a goalkeeper, but has put forward some impressive numbers.

Lazio’s save percentage in the danger zone is 67% (Strakosha has played every minute), the joint 2nd highest in Serie A this season. He’s conceded 14% less goals than expected and saves 5% more than expected.

He does go long a lot more than Mignolet, with 56.7% of his passes, but he has a good short pass xP Rating with 1.185. He under performs long passes by quite a bit (again, I put this on the model) but his rating of 0.698 is higher than that of Baumann’s.

Also encouraging is that Lazio are in pretty much the same position on the graph of PPDA vs passes in the final third per shot, while also being below average for number of players between ball and goal for shots in the danger zone – though not by quite as much as Liverpool.

Strakosha definitely seems talented, has some great numbers and is still very young. This doesn’t necessarily mean he’s the man for Liverpool, but he’s definitely worth looking into.

Alisson – 25 – Roma

Roma have only conceded 0.53 non-penalty or DFK goals this season, around half of their 0.98 xG against per game, and a large part of that could be down to Alisson.

The Brazilian’s save percentage from chances within the danger zone is the joint seventh highest in Europe’s top five leagues this season. Then his expected save rating is the joint 3rd highest in Europe’s top five leagues, with only David de Gea and Marc-Andre ter Stegen having a higher rating.

Roma’s defence also has a lower PPDA than Liverpool’s, although like most clubs they have more players and more pressure on the shooter for shots within the danger zone than the Reds.

Alisson’s distribution also seems impressive. He has the 12th lowest percentage of passes long and a positive xP Rating of 1.013. Separating the passes to long and short, looking only at those with below average passes long, sees him have the 7th highest xP Rating for long passes (albeit behind Mignolet) in the top 4 leagues (no Ligue 1), while his short pass xP Rating of 1.18 is also impressive.

Finally, though I’m not sure how helpful this is, Alisson’s xSOT rating is the joint 4th lowest in Serie A this season and slightly better than that of Mignolet’s.

To add onto all the above, @colintrainors model also rates Alisson very highly, both for his time in Brazil and Roma.

While I’ve only had a quick look, I think I’m in agreement with @thefutebolist that Alisson could be the best for Liverpool. Still young for a ‘keeper, great numbers and looks as though he could fit the style and be a big improvement on what they have. A few clips of Alisson can be seen below:

Finally, how does £75m Virgil Van Dijk factor into all of this?

Football Data Analysis: Van Dijk and Liverpool’s Defence by @EveryTeam_Mark takes an interesting looking into the signing of the Dutch defender. If Virgil van Dijk has been signed to help Liverpool’s defence of set pieces, his aerial duel success rate in his own box isn’t much higher than current Liverpool pair Dejan Lovren and Joel Matip.

While Liverpool have also been better at dealing with set-pieces in general this season. Looking at chances where the primary or secondary assist was either a corner or free-kick see’s Liverpool conceded the 6th least xG per game from set-pieces. This is to be expected though given they concede the 8th least chances from set pieces per game, and given how few chances they concede it’s likely they also concede few set pieces.

Looking at the percentages, 16.3% of their xG against comes from set pieces (primary or secondary) assist, the 7th highest in the league and down from last year which was ~20%. More impressively, however, the quality of chances conceded from set pieces has fallen from 0.178 xG per chance to 0.114 per chance – the joint 5th lowest in the Premier League.

Moving on from set pieces, it feels as though a new centre-back is unlikely to be the answer to Liverpool’s defensive issues. Improved individual quality will obviously be an advantage, but if Liverpool continue to leave their centre-backs so exposed it’s likely they’ll make either make errors or not as look as good as they did at their old club.

Particularly with a defensive midfield player like Oriol Romeu, Van Dijk is moving from a club where he had a lot more cover in front of him and was probably less likely to be so exposed with the opposition running at him. This isn’t to say he’ll suddenly lose all defensive ability, just there’s still work to be done and other players/changes needed.

This brings us onto the midfield and attack.

Midfield and Attack

The big news obviously involves Philippe Coutinho making his huge move from Liverpool to Barcelona. While Liverpool will be disappointed to lose on of their key players, given the fee and the players desire to move to Barcelona, it seemed inevitable.

My main issue with the sale would be that it came in January, rather than the sale itself, given Liverpool are either going to be scrambling for a replacement – with a good chance of overpaying or panicking – or spend half a season without replacing one of their best players.

Before diving into this though, it’s time to find Liverpool a defensive midfielder.

As stated above, I’m generally in the camp that a centre-back won’t solve all of Liverpool’s problems and money should be investing into a new defensive midfielder to protect the back four (not to say this is a quick fix either).

So, what would Liverpool’s defensive midfielder need to do? Ideally, on the ball they should be a good distributor while they should be good defensively and preferably playing for a team that also likes to play high up, so they’re used to dealing with a large space to protect.

At the moment, Liverpool’s two options for the central/more defensive midfield role – Jordan Henderson and Emre Can – both either average or below average on passing metrics.

Looking at the xP of the players, like @footballfactman did for his expected passing piece, shows Henderson and Can to be completing around the same as they are predicted – albeit below average for this group of midfielders (>=450 minutes).

Click to Enlarge

While looking at vertical passes (see Vertical Passing 2.1) both Can and Henderson under perform the number of completions they should have made.

Click to Enlarge

Then, to show what teams have similar activity off the ball (PPDA using all passes and defensive actions now) and on average how high up the perform their defensive actions, I’ve used @footballfactman’s idea of a proxy for pressing. This gives the following:

Click to Enlarge

The position that Brighton are in makes me feel as though I’ve made some kind of error, but going off the above teams in the upper left quadrant are most likely to be suitable for Liverpool. While players can definitely adapt, it’s mostly just a measure to try and reduce risk and the time it’ll take to settle in.

With all the above in mind though, who are some candidates?

Max Meyer – 22 – Schalke

Meyer may not be the first name to come to mind when thinking of defensive players, but playing a deeper role this season seems to have suited the 22-year-old.

For midfielders under 23-years-old (born after 1/6/1994), Meyer has completed the 2nd most interceptions and tackles in the 4 leagues (Premier League, Serie A, Bundesliga, La Liga) used this season, only behind Liverpool bound Naby Keita, with 5.961. Interceptions and tackles aren’t always great indicator, given it doesn’t take possession into account, but it’s still encouraging.

In fact looking at possession adjusted defensive numbers, using the method by @statlurker posted on StatsBomb, Meyer has the 12th most possession adjusted interceptions and attempted tackles p90 for midfielders in the 4 leagues used.

Meyer’s expected passing numbers are also exemplary, with the 7th highest xP Rating for the same group of young midfielders. He doesn’t make many vertical passes, however, with just 3.506 attempted p90, while his completing them at the expected rate (-0.006 extra completion ratio).

Finally, not too important for a deep midfielder, his expected goal contribution (xG value of chances, chances created, chances created with a secondary assist) is 0.208, which seems impressive for a deep player.

To conclude, Meyer as a defensive midfielder seems an interesting choice. Still only 22-years-old, some good numbers, Schalke not too far away from Liverpool for their defensive numbers plus his contract being up in the summer all make it an interesting choice.

Even if he isn’t the choice for the defensive midfield slot, he may be worth bringing in either fairly cheap or on a pre-contract deal for the summer given his contract situation, as he could be a useful addition to the squad.

Santiago Ascaibar – 20 – Stuttgart

Stuttgart seem to possess a lot of young talent, having talked about Timo Baumgartl and Benjamin Pavard before, this time it’s Argentinian midfielder Santiago Acscaibar.

Ascaibar has very similar numbers to Meyer, completing 5.206 interceptions and tackles p90, having an xP rating of 1.065 and also completing his vertical passes at around the average rate – albeit attempting slightly more p90 with 4.015.

Like Schalke, Stuttgart also occupy a similar area of the grid to Liverpool, though it’s worth pointing out Stuttgart and Schalke both make a lot less passes in the final third per shot than Liverpool do.

Ascaibar looks an interesting prospect, given that he’s still only 20-years-old and was signed last summer, he may not be what Liverpool are looking for at this moment, but seems to definitely be worth keeping an eye on.

Jean-Philippe Gbamin – 22 – Mainz 05

The Bundesliga seems to be home to most the young, talented defensive midfield players and Gbamin is no different.

At a club that Jurgen Klopp knows well, Gbamin has been hugely impressive this season. His xP Rating is slightly below the previous two names, with 1.056, and he does attempt less interceptions and tackles p90 with 4.67, but these are still good numbers.

His vertical passing is an area in which he performs better than the previous two. As can be seen in the below graph (taken from Vertical Passing 2.1), Naby Keita is the only under 24 midfielder to attempt significantly more vertical passes than the Ivory Coast international.

Click to Enlarge

Like the previous clubs, Mainz are also in a similar area to Liverpool when it comes to their defensive numbers, but like the other clubs they are also more direct than Liverpool.

There is also a question mark over the availability of Gbamin, given that he’s recently (November 2017) signed a contract extension, meaning he’s under contract until 2022. Again, like the above two, Gbamin is definitely worth looking into further, while the below clips show that he seems pretty comfortable sitting in front of the defence, winning the ball back and spreading it (he has also made some appearances at centre-back, which may be included in the clips):

Lucas Torreira – 21 – Sampdoria

Probably my favourite of the names mentioned, Uruguayan midfielder Lucas Torreira has been putting in impressive performances for Sampdoria over the last two seasons.

Like all of the above, he impresses with both his xP Rating (1.081) and his interceptions and completed tackles p90 (4.357). As you can see on the vertical passing graph used for Gbamin, he also excels with that, attempting around the same number as Gbamin but with a positive extra completion ratio.

Torreira also has an xG contribution higher than the above options with 0.328 p90 – mostly thanks to 0.228 x2A p90.

Sampdoria have had some exceptional youngsters over recent years and Torreira seems no different. With this in mind, it’s strange not to see more interest in Torreira.

He seems as though he’d be a good fit for Liverpool, being able to sit in front of the defence, distribute the ball and cover lots of ground – plus, going off the pressing graph above, Sampdoria even have a higher line and lower PPDA than Liverpool this season, so the transition shouldn’t be a long one for Torreria.

Having extended his contract at the end of last summer it seems as though he won’t be cheap, but being an important player in a fairly strong Sampdoria side for the past two seasons at such a young age make it seem as though he should be able to step up and fit straight into this Liverpool side.

While a lot more video analysis would be needed, Torreira would probably be my pick for a new Liverpool defensive midfielder from the above names, although the names above are also promising. Some clips of Torriera can be seen below:

Some other names include Napoli’s Jorginho or Amadou Diawara and Borussia Dortmund’s Julian Weigl, who I talked about when discussing potential replacements for Sergio Busquets. Then while I don’t have advanced stats on Ligue 1 (for passing and defensive stuff), Monaco’s Fabinho could also be worth a look.

The four names talked about above can be seen in the below graph for possession adjusted defensive actions and xP Rating – with Max Meyer particularly impressing.

Click to Enlarge

Now, finally getting around to Coutinho.

Given the front three of Sadio Mane, Roberto Firmino and Mohamed Salah, I don’t think big money needs to be spent on an attacker to throw straight into the first team. I’d favour a younger option, while putting the excess money towards signing a goalkeeper, defensive midfielder and possibly another midfielder.

Coutinho helped Liverpool break down deep defences, particularly thanks to great dribbling ability in and around the box. In fact, looking at successful take-ons in central areas in the final two fifths of the pitch, only Messi has completed more than Coutinho.

What should be encouraging for Liverpool fans is that 5th on the list is Naby Keita.

For more on dribbling, @TiagoEstv recently wrote a piece about why dribbling can be useful.

Now, with Naby Keita coming in in the summer, while not a like for like replacement for Coutinho, they do have some very similar numbers. The exception is Naby has a huge defensive output and less attacking output than the Brazilian.

Keita’s 9.963 attempted tackles and interceptions p90 is pretty crazy, and obviously a lot more than Coutinho’s 4.202 p90. Interestingly, Keita also attempts and completes more take-ons than Coutinho p90 too, while their vertical passes are pretty identical. Keita attempts 7.417 compared to Coutinho’s 7.861, though Coutinho is marginally above the expected completion number and Keita marginally below.

Where Coutinho shines is in and around the box, attempting 5.828 passes into the box p90, over double Keita’s 2.546. Coutinho’s xG contribution is much larger than Keita’s, with 0.959 compared to 0.424. Interestingly, though, Keita’s open play xA per 100 key entries is larger than that of Coutinho’s with 1.039 to Coutinho’s 0.604.

Liverpool both gain and lose in switching Keita for Coutinho, Keita seems a more all-round midfielder, more than capable of helping defensively, carrying the ball into dangerous areas and chipping in with goals and assists, but doesn’t seem to excel in attack in the same way as Coutinho. This could make Liverpool more balanced, and better defensively, but they also don’t want to lose that ability in and around the box.

The problem is, not many players have that ability. The below graph shows completed take-ons in advances areas against completed passes into the box p90:

Click to Enlarge

One midfielder that could be of interest to Liverpool is Allan from Napoli.

@TiagoEstv wrote a great piece on Allan, that’s a lot better than anything I could say about him.

On the above graph, he doesn’t jump out, but he’s in the same kind of spot as Riyad Mahrez, below Goncalo Guedes and Dries Mertens. His numbers for the above graph obviously aren’t going to be as good as Coutinho’s, but he’s still incredibly impressive.

He only completes 1.142 passes into the box p90, and 1.142 take-ons in advanced areas p90, yet his xG contribution of 0.726 p90 is hugely impressive – mostly thanks to lots of secondary assists, of which is x2A p90(using the xG value of the resulting shot) is the 2nd highest in Serie A this season.

His total xG contribution is the 24th highest in Serie A, which may not sound too amazing, but given all the forwards/wingers/attacking midfielders to consider, it’s hugely impressive. While only scanning the names, he seems to be the highest central midfielder in the league.

His vertical passing is also impressive, though he doesn’t attempt many (3.235 p90), he does complete 24.7% more than expected.

In the graph showing xP Rating and possession adjusted defensive numbers Allan also shines, having the 5th highest possession adjusted attempted tackles and interceptions p90 for midfielders this season.

A key part of an impressive Napoli side, he should be able to jump straight into this Liverpool side. Whether or not he’d be open to leaving is another matter, but given he only has 18 months left on his contract and Napoli have Piotr Zielinski waiting in the wings it could be an avenue worth exploring.

With the search for a more attacking midfield option not proving too great, let’s move on to the front three.

As stated above I don’t think Liverpool need to go out and spend big on someone like Thomas Lemar or Riyad Mahrez, unless they want them to operate centrally. With that being said the players who I think could make interesting purchases will also cost a pretty penny.

The two players are Leverkusen’s Leon Bailey and Julian Brandt. Liverpool have been linked with Brandt in the past and him and his team-mate have really kicked on this season.

Firstly, Brandt.

I mentioned Brandt as a possible Alexis Sanchez replacement in my piece about replacing Mesut Ozil and Sanchez, citing some impressive creativity stats. These creativity stats have improved even further for the first half of this season. Looking at the xG value of shots created by Brandt (through the primary or secondary assist) in open play sees him have the 2nd highest xG contribution in Europe’s top 5 leagues with 0.797 p90 – behind only Kevin de Bruyne.

His xG of 0.203 isn’t crazy, but (including set-pieces now) it means his total xG contribution is the 6th highest in the Bundesliga this season, while still only being 21-years-old.

Brandt also completes 1.609 passes into the box p90, behind only Marco Asensio and Bruma this season for under 23’s. The only area that is slightly weaker is his xP Rating, which is marginally below 1 at 0.989.

Now, Leon Bailey.

Impressively with Bailey, his xG contribution’s are all around the same, an xG of 0.380 p90, xA of 0.347 p90 and x2A of 0.335 p90. This all amounts to an xG contribution of 1.062 p90, the 5th highest in the Bundesliga, marginally above Brandt.

Bailey’s numbers for passes into the box (1.311 completed p90) and xP Rating (0.866) aren’t quite as high as Brandt’s, and you’d hope the xP Rating in particular would improve a bit, but Bailey’s play this season has been impressive, even more so given he’s still only 20-years-old. Watching his runs and finishes it feels as though he could be a great option out wide for Liverpool.

See some clips of the pair below:


Bringing all the above together, Liverpool find themselves in an interesting position.

Now, if I was in charge of Liverpool’s transfers (and their fans will be grateful I’m not), my list would be a goalkeeper, defensive midfielder, a central midfielder and a young winger.

While Virgil van Dijk is a good defender, I’m unsure if him alone will solve Liverpool’s defensive problems. It feels as though they should also look to recruit a new goalkeeper and defensive midfielder to help their defence further – though I’m not sure who should be in January and who should be left until the summer.

Their current goalkeepers save and xG numbers aren’t great and signing a great ‘keeper can go a long way in out performing xG against – as seen with Manchester United and Atletico Madrid. The best choice seems to be Allison from Roma.

Moving into midfield, I feel slightly underwhelmed by the options. The attacking options don’t have impressive stats as Coutinho, while the defensive options still seem fairly young and raw. Given Liverpool already have a lot of midfield players this may not be a bad thing, as it means they won’t have to be thrown straight in, but it may reduce the immediate impact they can make.

From the names I went through above, I’d probably go with Lucas Torriera as the defensive midfielder, while also look to bring in Allan as a central midfielder. Max Meyer on a pre-contract deal would be hugely tempting also, though I’m not 100% sure on what role would be best for him.

With Emre Can rumoured to be leaving, Adam Lallana and James Milner both being over the 30 mark at the end of the season (although Lallana would have just turned 30) and Jordan Henderson having some below par numbers, it feels as though freshening up the midfield could be a good idea for Liverpool.

With that being said, if Liverpool view Alex Oxlade-Chamberlain as a central midfield player they might not have to buy an additional central midfield player in Allan, and just have Keita come in, preferably trying to see if they can bring him in early as opposed to the summer.

Making the midfield slightly less attacking, and more of a functional unit, could help improve the defence while still having Mane, Firmino and Salah as a dangerous front three – while Naby Keita and Allan/Chamberlain are more than capable of contributing to goals and dribbling in advanced areas.

In attack the choice of Julian Brandt or Leon Bailey depends on what they’re looking for. Bailey seems like he’d fit the Mane/Salah mould more, while Brandt seems more like a creator, possibly similar to when Coutinho played out wide albeit with less dribbling. Either one would be a huge asset and it’d mostly just be a stylistic choice.

I’m not sure if the Van Dijk deal was done with the Coutinho deal/money in mind, or whether Klopp has money to spend, but given Liverpool have had a positive net spend over the last two seasons I’d imagine Klopp has some money to spend.

If he could bring in these 3-4 targets I think Liverpool would have an impressive squad. While it may not possess the star quality of Coutinho, it could be a strong, functional, team as seen below (I went with Bailey for the diagram, but Bailey or Brandt).

Regardless of what happens, it’ll be interesting to follow the actions of Liverpool for the remainder of the window, whether it’s seeing how Coutinho’s replacement performs or, if they don’t sign anyone, how they manage without the Brazilian.

This article was written with the aid of StrataData, which is property of Stratagem Technologies. StrataData powers the StrataBet Sports Trading Platform, in addition to StrataBet Premium Recommendations.

Leave a Reply

Your email address will not be published. Required fields are marked *