Featured

Using data and video scouting to find a replacement for James Tarkowski

The main source of Burnley’s continued spell in the Premier League has been built on a rock-solid defence, which gives them a platform upon which to build their attacks on and allows them to stay competitive in games. At the heart of this, in recent years, has been James Tarkowski, a man bought for absolute peanuts (315,000 peanuts to be exact) in 2016 as back-up, and then the replacement, for Michael Keane. However, given his form, I would not be surprised if he were to move on from Turf Moor, especially given Burnley’s need for investment in the rest of their squad. Therefore, I thought it would be interesting to use data and video scouting to try and find a cheap(ish) player who could come straight in and start for the Clarets next season.

Before I go into finding a replacement, however, let’s have a quick look at what Burnley could potentially be losing.

A quick dashboard showing off some of Tarkowski’s key skills.

I know that pure statistics are not the best way to assess defensive quality, especially for a centre-back, but I feel like this visualisation highlights some of his best qualities. As the team set up in a low block, at least comparatively to other PL teams, it requires the centre-backs to do quite a lot of what I like to call ‘backs to the wall defending’: winning loose balls, defensive duels, balls in the air etc. This is because, once the ball has made its way into the final third, Burnley retreat to the edge of their own area, remaining compact and removing a lot of the space for players to receive in between the lines. This removes a lot of the decision making required by the centre-half pairing, although this can be undone by hitting them quickly in transition, preventing Burnley from recovering into their defensive shape. More information on Burnley’s defensive structure can be found in this Athletic article by Andy Jones from the end of last season (https://theathletic.co.uk/1903834/2020/07/03/burnley-defence-clean-sheets/), or in this video from a few years ago by Tifo Football.

Going back to Tarkowski, he excels at winning duels across the pitch, performing above average for defensive duels, aerial challenges and blocking shots. On top of this, he adds some ability to progress the ball up the field, although he is by no means Burnley’s main option in this regard. Therefore, his replacement needs to be someone with a similar ability to win the ball back in their own penalty area, whilst moving the ball forward to the strikers.

To find our James Tarkowski v2.0, I decided to scout exclusively in the Championship for players with over 1000 minutes this season. This is because it’s a division that Burnley have taken players from before with reasonable success (Chris Wood, Charlie Taylor, Josh Brownhill to name a few), and therefore I felt that it would be the most realistic option for them. I considered also looking at League One, but I felt that the players I could find there might not be ready for immediate game time in the Premier League, assuming they stay up, which I don’t think Burnley can afford when losing one of their key players. I also decided to only look at players under the age of 27. Burnley’s squad profile is enough to give anyone on Football Analytics Twitter an aneurysm, and therefore dropping the average age might not be a bad idea. It also might give them another sellable asset to make money on in a few years (in other words – good squad building!).

Only 3 players under the age of 28 have played over 50% of minutes for Burnley this season.

To start, we’ll look at the aerial ability of our centre backs. This is useful, not only for dealing for defending crosses, free-kicks and corners, but also gives Burnley a potential threat at the other end. So far this season, Burnley are the second-lowest scorers with 18 goals, so giving them another avenue from which to attack would be very helpful. Therefore, I included the number of headed goals scored as part of the visualisation.

Aerial duels per 90 vs Aerial duel win %, with size representing the number of headed goals scored.

From the first look, the players that stand out are Nottingham Forest’s Scott McKenna, Stoke City’s Harry Souttar and Derby County’s Matt Clarke, although none of these have registered a headed goal yet this season. If we were more interested in this metric, then the main names are Cardiff’s Curtis Nelson, Barnsley’s Michal Helik and Swansea’s Ben Cabango, all of which are also frequent aerial duellists.

Moving on to the next area, we want a defender who is strong in ground duels, specifically defensive ones. This is because Burnley’s centre backs are not covered by a player in front of them, and therefore will sometimes have to step up and try and win the ball, or at least slow down the attack. I also looked at the number of fouls committed per 90, as being too aggressive might give away chances for free kicks in dangerous positions.

Defensive duels per 90 vs Defensive duel win %, with size representing the number of fouls committed per 90.

We can see again that the players that stand out are McKenna, Clarke and Souttar, with Joe Rodon (now of Spurs), Ben Davies (now of Liverpool) and Helik also doing well. However, it’s fair to say that Helik does appear to give away a few more fouls than the other players mentioned.

The last area that I wanted to look at is the players’ passing ability, especially long range. Burnley play a very direct brand of football, aiming to isolate the opposition’s centre backs by playing long balls to their physically threatening strike force, hoping to force errors or win the second ball. Most of this progression is done by Ashley Westwood, although as I mentioned, giving Burnley a secondary passing player might be a good idea to prevent their game plan from being stopped at source. Therefore, I looked at the number of progressive passes made per 90 against how accurate they were.

Progressive passes per 90 vs Accurate progressive pass %.

The only player of the main three we’ve seen on every visual that stands out here is Harry Souttar, who not only plays a lot of progressive passes, but is also very accurate with them. Another honourable mention has to go to Ben Davies, who shows why Liverpool picked him up to solve their centre back crisis in January, although this means he is obviously not available for Burnley. Yoann Barbet of QPR and Paddy McNair of Middlesbrough also show their abilities here, but performing poorly in other areas has stopped them from being a real option for this role.

After going away and doing some video analysis, the standout player for me, and therefore my choice, is Harry Souttar. Although there were other options who did well in certain aspects, Souttar was the only one who performed consistently well in the areas I was focusing on. He was also helped by the fact that both Clarke and McKenna are both primarily left footed and would therefore be better replacements for Ben Mee.  

Souttar currently plays CB for Stoke who primarily use him on the right in a back three alongside James Chester and Nathan Collins. He’s currently 22, meaning he could be part of Burnley’s defence for the foreseeable future and, although he’s only playing in his first full season for Stoke after going on loan to Fleetwood last year, he’s already registered more minutes than anyone else in their squad this year, as well as earning himself a call-up to the Australian national team. As for his value, Transfermarkt has his price at around £3.15 million as of December 20th 2020, although it could be less as he’s going into the final year of his contract in the summer.  

The first thing that stands out about Souttar is his height. He stands at a gargantuan 200cm which obviously helps him aerially and makes him difficult to force off the ball. He uses this to his advantage when defending one-on-one, often trying to get himself between the opposing player and the ball, rather than making a challenge. The obvious disadvantage of his height is his lack of speed. Whilst he isn’t slow and can often keep up with attackers once he gets going, he can be found slightly flat-footed upon occasion. This is mainly an issue when he steps up to pressure an attacker, as a quick release of the ball will often leave him beaten.

The timing of his challenges is good and he has good range due to his long legs, often helping him block crosses at the byline if he slides across. His positioning is also generally good although he can often get caught out, especially if he has to react to things quickly. A good example of this is Stoke’s away game against Brentford this season.

Souttar is initially picking up Toney on the edge of the area. The Brentford player then sends a ball to Mbeumo who is making a run between the other two Stoke centre backs.

Although he stays with Toney, once Brentford win the header, he is slow to react, allowing Toney to move into space and get a shot away, leading to a Brentford goal.

These defensive lapses could be an issue if he makes the step up, especially since he won’t have the protection of an extra central defender in Burnley’s system, although I would have faith in Sean Dyche improve his defensive game if given the chance.

His last major strength is his long passing, which is absolutely fantastic, especially for Championship level. He has usually played one of two passes from what I’ve watched of him this season: a slightly lofted ball between the opposition centre back and full back, looking for a striker running into the channels, or a flatter cross-field switch, looking to stretch the play. The first pass especially is typical of what we’ve seen from Burnley in recent years, and I have no doubt that he would improve their ball progression significantly. In terms of shorter passing, he is reliable, if unspectacular. He can play a little too safe at times, which can put his team-mates under pressure, and I’d like to see how he does against teams that would exert much more pressure on the ball, as I feel he is too often unchallenged when making a pass in the Championship.

Despite these problems, I still think that Souttar is the best option for Burnley should Tarkowski leave. His positives far outweigh any of his competitors at Championship level, and the combination of his age and potential price would potentially give Burnley a bargain with high profit potential should he continue his form to the top tier.

All data taken from/video scouting done with Wyscout

Featured

Who should play in CM for England?

One of the many decisions Gareth Southgate will have to make before this summer is who will he wants to take to this summer’s European Championships (which as of yet have not been cancelled – fingers crossed), and out of these players, which make up his strongest line-up? Most of these will be challenging, given the strength in depth of some and the lack of depth in others, but personally the most interesting area for me is that of central midfield. Not only is there a lot of strength in this position, but it is one of only two roles in Southgate’s preferred 3-4-3 formation that requires more than one player to fill it. This means the conversation is not just about who the best player is, but which two players form the best unit to control the centre of the pitch.

What does Southgate want?

Let us start by looking at what is required from this role and the way England set up. Like I mentioned, the midfield consists of two players forming a double pivot that assist in build-up, progress the ball forward and help shield the defence. As England like to push their wing backs up high to support attacks, this requires the midfield to be even more defensively solid and positionally disciplined than in other formations that use a double pivot (such as the 4-2-3-1), as pushing up too high themselves might leave their centre backs vulnerable. It is worth noting however that against teams that sit back defensively, such as Iceland, Southgate has played a more attacking midfielder in the pivot to help support the front three. They also need to be comfortable with the ball, as there is a risk of being man-marked by teams that operate with a 3-man midfield, which could damage England’s ability to play out from the back.

In build-up, rather than dropping back to progress the ball, the midfield is tasked with finding gaps in the opposition’s defensive lines from which they can receive the ball from the centre backs and move the ball up the field. Once they have the ball, they can then turn and find one of the two inside forwards, the striker or the wingbacks who should have pushed up. If these options are not available, they can either find their midfield partner or recycle the ball back to the defence.

The Contenders

Now that we have looked at what Southgate wants from his midfield, let us consider who he might call up. I’ve considered players that have played regularly (over 900 minutes as of 14th January 2020) in a top 5 European league as either the deepest midfielder or as part of a midfield double-pivot at some point this season. Note that this does exclude several players who have been called up by Southgate recently (most notably Winks and Bellingham), but I feel that it would be difficult for me compare these players, and for Southgate to call them up, when they aren’t even playing regularly for their respective clubs.

Once these filters have been applied, we are left with 9 players: Jordan Henderson, Mason Mount, Declan Rice, Kalvin Phillips, James Ward-Prowse, Ben White, Isaac Hayden, Josh Brownhill and Ashley Westwood. Some of these, if we are being honest, might be longshots at best but, nevertheless, we will give them a chance to claim their place in the squad.

Passing

We first want to look at how good these players are when they have the ball. As I mentioned before, a key part of this role is to get the ball up to England’s more attacking players, so we will look at their progressive passes p90 and their passes into the final third p90 compared with all of the other players in the Premier League who meet our initial criteria.

A plot of accurate passes into the final third p90 vs accurate progressive passes p90, with the players of interest marked in red.

As you can see, Henderson is by far the best progressive passer for this position in the league, let alone for England, making him a strong candidate for selection. Other notable players who perform well here are Leeds’ Kalvin Phillips, Chelsea’s Mason Mount and, rather surprisingly, Burnley’s Ashley Westwood. On the other end, both Hayden and Brownhill really struggle in this area, which would rule them out of contention given the need for more of an all-rounder. It is also worth mentioning Ben White who, although he appears rather low at first, is handicapped by a lot of his league games coming at CB for Brighton, making it difficult to pass into the final third regularly.

Defence

Another area to look at is how well these players control the area in front of their defence, and to do so we will look at their number of successful defensive actions p90 against their duel win percentage, which encompasses all three types: offensive, defensive and loose ball.

A plot of successful defensive actions p90 vs duel win percentage.

In contrast to the previous section, we see both Mount and Westwood performing poorly here, which potentially rules them out of selection in the best midfield. On the other end, the most active defenders are Phillips and Brownhill, whilst the best in terms of success rate are Brownhill (again), White and Henderson. Overall, with the exception of the first two mentioned, all of the players gave a good account of themselves here and would be able to provide a solid base for England to attack from.

Ability in possession

Whilst it is important that these players get the ball forward, it is also vital that they don’t give the ball away too easily in the middle of the pitch, allowing the opponent to attack England from dangerous areas. Southgate has also placed emphasis of keeping control of the ball and, whilst lateral passing is not amazing from a purely progressive sense, it allows England time to get their wing backs forward to support attacks and keep the width.

A plot of pass completion percentage vs passing directness percentage. Passing directness was measured as the proportion of forward passes over the total number of passes.

This visualisation probably puts the nail in the coffin of anyone who wanted to see Ashley Westwood in an England shirt. Although he is very good at getting the ball forward, this is usually done with the sacrifice of accuracy and, having watched some of Burnley’s games this season, his long, looped passes forward definitely suit their playstyle more than England’s. Another player that falls short of this requirement is Brownhill who, despite playing very conservative passes most of the time, fails to find his target. The safest passer of the group is Declan Rice, who gets the ball to a teammate just shy of 90% of the time, despite being as direct as the likes of Ward-Prowse, Mount and Phillips. He could definitely be an option for games where Southgate wants to control the ball more. Lastly, I want to point out that, despite being the third most direct player, the fact that Henderson’s accuracy is comparable to many of his English colleagues goes to show how good of a passer he is.

Passing further forward

A bonus category I want to look at is how good these players are once they get into the final third. I mentioned before that Southgate has asked of at least one of the CMs to get forward against weaker sides historically and would definitely want this option at the Euros.

A plot of passes into the penalty area p90 vs xA p90.

This is the area that Mason Mount really shines in due to him playing as a more advanced LCM in a midfield three for Chelsea, and you can see how this allows Southgate more attacking output without sacrificing ball progression. Other (realistic) options here are Henderson and Phillips, both of which have already put down strong cases for starting roles.

Conclusion

Whilst there are no real shocks in terms of who should be called up, I think these plots should give you some idea of who Southgate should take as part of his squad this summer, and who fits into England’s strongest line-up. Whilst they obviously all have strengths, Westwood, Hayden and Brownhill are all too limited in key areas of their game to be even considered for selection. Ben White does slightly better, and it would be interesting to see what his metrics look like if he plays at CM for Brighton more often. However, at the moment, he is far too raw to be called up, both at CM and at CB. As for the best pairing, it is clear to me that Henderson and Phillips perform the combined best across all the areas we looked at, with Henderson being the best progresser of the ball, and Phillips being the one of the most active and strongest defenders.

If we look at the rest of the squad, there will probably be space for all three of the remaining players, depending on who else gets called up. It is hard to say who is the natural third choice for this role, with Rice and Ward-Prowse producing very similar metrics across the board. The key decider would be whether Southgate prefers Rice’s reliability in possession to Ward-Prowse’s set piece expertise, and I have a feeling that he does. As for Mount, I don’t think that comparing him to these players for this role really highlights his best qualities and, ironically, I feel that he is rather hamstrung by Southgate’s insistence on playing this formation. However, due to his versatility, work ethic and attacking output, he will almost certainly find a place into the squad.  

Data, clips and video analysis all taken from/done using Wyscout.

Is Timo Werner a good finisher?

There is has been a lot of attention placed at the door of Timo Werner recently, especially in the midst of an agreed move to Chelsea for a reported £50 million transfer fee. This has led to a lot of criticism around his recent performances, especially against Paderborn last weekend, some going as far as to suggest that Werner is not an elite finisher, and only scores goals because of the number of chances he is provided by the rest of his team. Therefore, I have taken it upon myself to test this hypothesis using the data I have available to me.

It is first worth establishing that RB Leipzig are an elite team in terms of chance creation. They are second in the league (behind Bayern Munich would you believe) for Total Shots p90, Shots on Target p90 and Non-Penalty xG, and joint-second for Non-Penalty xG per shot. They also do finish a lot of the chances that they create, being the third highest scorers in the division and joint-second for Non-Penalty Goals minus Non-Penalty xG. As for Werner himself, he is second in the league for Goals Scored, Shots on Target p90, xG and Non-Penalty xG, all behind Robert Lewandowski, and third for Total Shots p90. This makes sense, as obviously Werner is the focal point of RB’s attack, and would therefore contribute highly to the team’s shot numbers.

 It starts to get interesting however when you look at the difference between his Non-Penalty goals scored and Non-Penalty xG, a statistic commonly used to measure how clinical a player is in front of goal. Using this metric, Werner drops to 9th in the Bundesliga, now behind the likes of Erling Braut Håland, his teammate Patrik Schick and the league’s runaway leader in this metric, Jadon Sancho. Although this may seem like a sign that Werner is under-performing, it is worth noting that this value is still +3.7, meaning that he has scored nearly 4 goals more than he was predicted to score using StatsBomb’s xG model. When compared to all the players in European Top 5 leagues that have scored 10 goals or more this season, Werner is 14th out of 65, and he maintains this whilst taking the 13th most Shots p90. In fact, there are only 3 players that out-perform Werner whilst taking more Shots p90; the previously mentioned Lewandowski, Josip Iličić, a man who his having the season of his life for an attacking Atlanta side, and Lionel Messi.

Players with 10+ goals in European Top 5 Leagues compared on Total Shots p90 against Non-penalty Goals – xG

Another thing to mention is that, despite being framed as a weakness, the amount of shots that Werner takes can generally be seen as a positive, especially if they’re being taken from good positions. If we look at Non-Penalty xG per Shot, Werner sits at joint 20th in the Bundesliga with 0.17, a respectable position for someone with his volume of shots. For comparison, Lewandowski sits at joint 12th with just 0.01 npxG/Shot higher, and Serge Gnabry, who is ahead of Werner for Total Shots p90, is down in joint 47th with 0.13. He comes of even better when we compare him to 10+ goal players, finishing joint 18th, above both Iličić and Messi.

Players with 10+ goals in European Top 5 Leagues compared on Total Shots p90 against xG per Shot

 So, in conclusion, is this a fair claim to make about Timo Werner? Based on the data at hand, I would argue that not only is this assumption incorrect, but also that the value being paid for him is far less than what should be expected based on his age and ability. Not only is he getting a large number of high quality chances, both due to his team and his own excellent off-the-ball movement, he is also converting these at a rate higher than his xG would suggest. I also feel that it is fair to say that due to the amount of shots he is taking, that it would be unfair to rule out his scoring streak as pure luck, and therefore, given the opportunity, he should carry on in the same form going forward. The key will be now for Chelsea, if the transfer goes through, is making sure they play Werner in a system that plays to his strengths: his ability to run in behind and finishing ability in the box. If they can do that, then Lampard and the rest of the team will have found a striker that will score plenty of goals for them for a long period of time.

Data taken from fbref.com

Create your website with WordPress.com
Get started