Moving into midfield
Looking for midfielders is usually quite fun, as they’re often tasked with more than defenders or forwards. But, lots of the above problems persist when looking for progressive players. Working with the figures I have, it’s harder to differentiate between players who look to break lines and those who hit hopeful long balls. Similar problems exist when looking at defensive activity. Players could rack up huge numbers because they’re playing for less dominant sides. On the other side of that coin, a player may be an important part of their team’s shape and structure, but not put up high individual figures for winning the ball back. A lot of the problems are general drawbacks of event data. It has a limited scope, only focusing on a few actions a player makes on the ball with no context.
But anyway, let’s try and see if there are any interesting players around.
Not only do the same problems as above persist, but my same dilemma does. I’m never sure whether to focus more on the figures for passes into the final third or progressive passes. I prefer the idea of the latter but I’m not too keen on WyScout’s definition. But the final third boundary is quite arbitrary and players can passively collect figures. If a full-back has the ball wide in the final third and goes back to a midfielder, who then takes a look around, notices nothing’s on and goes back to the full-back, that’s a pass into the final third. Maybe seeing how players differ in each one is where things could be interesting. The graph below shows some of this difference, looking at accurate progressive passes and accurate passes into the final third.
Like Mapi dominated progression from defence, her teammate Patri Guijarro does here. 43-year-old Formiga has the most passes into the final third per 90, but it’s Patri who shines in both categories. I decided to add the line to make the difference between the two axes more apparent. It’s not a y=x line, but those above the line are the players who make more progressive passes compared to final third passes than the ‘trend’ and those below are the reverse. So Formiga makes a huge number of passes into the final third, but not as many progressive passes in comparison.
To try and see some difference I watched Formiga’s passes into the final third against Issy. They’re second from bottom so I doubt it was the best choice, but Formiga attempted a huge 32 passes into the final third. Yet she only attempted 11 progressive passes. But watching the clips, there didn’t feel many passive passes into the final third that didn’t progress the play. At least not around two of every three. In the glossary, if a pass starts and ends in the opposition half, it has to move the ball 10 metres closer to goal to be progressive. Lots of Formiga’s passes went wide, which could make it harder to reach the threshold. Even though, in the context of the move, PSG can move up the pitch with the pass wide.
There are a few more midfielders from the top clubs in a similar boat to Formiga for this. Barcelona pair Kheira Hamraoui and Victoria Losada are under the line along with Lyon’s Saki Kumagai and Wolfsburg’s Ingrid Engen. Despite this, there’s an argument that they may not less progressive. Instead, they may have more opportunity for non-progressive passes into the final third than players from other clubs.
This makes for an interesting contrast with those who are above the line and from more elite clubs. Patri is the standout, but then there’s Lyon’s Amandine Henry, Chelsea’s Sophie Ingle and another Lyon player in Dzsenifer Marozsán all some way above the line. (There’s also Atlético Madrid’s Mesi. But I can’t decide if they should be counted among the top clubs this season or not.) These players are often moving the ball closer to the goal but not as often into the final third. But the final third is an arbitrary boundary. They may be moving it to the 64% line of the pitch each time they move it forward and just missing the cut-off. Although that is unlikely.
After only focusing on progression from the back before, I thought I’d try something different when looking for midfielders. Part of the reason is that, at least in the data, there’s more variety in midfielders than centre-backs. Also, I don’t feel like looking at another graph with progressive passes on at the minute.
So, as a quick, fun measure, I’d thought I’d look more at who’s above a certain percentile for a group of metrics and who has similar outputs to different players. Jumping forward a bit and looking more at those who impact further up the pitch, there are only five players who are above 80th percentile for accurate progressive passes per 90, progressive runs per 90, accelerations per 90, deep completions per 90, successful passes into the box per 90 and xA per 90. (Some of those are similar, so could be overkill.) These five midfielders are:
The bottom three are more interesting, if not just because they’re younger and the top two are more expected. 19-year-old Lena Oberdorf moved from SGS Essen to Wolfsburg in the summer. At Essen, looking through the data, she played in a variety of positions but mostly at centre-back. She did play in midfield and attack too and even managed an xG + xA of 0.40 per 90 in her final Essen season. At Wolfsburg, she looks to have settled into a midfield position. I liked the double pivot of Oberdorf and Engen in the Champions League quarter-final first leg against Chelsea. It’s a shame Engen has announced she’s leaving, as they could have been a duo to build the team around for the long-term.
@StatsBombES tweeted out a radar of Oberdorf not long ago, which reinforces the above. She’s been above average for progressing the ball as well as creating. She also has a strong defensive output that wasn’t considered above.
I haven’t seen much of her but she looks to be one of the most promising young players around. She’s still incredibly young but already looks to be a key player for one of the top clubs. Her presence in midfield was a big loss for Wolfsburg in their second leg against Chelsea.
Lena Lattwein I know nothing about. I’d like to see more of the two Lena’s included here, which makes it more annoying there’s no Bundesliga videos. Lattwein’s figures also come from fewer minutes. One area where Lattwein stands out is through passes. She attempts the third-most in the midfielder sample. I’ll talk about through passes, because there’s one player with crazy figures, a bit later on. A lot of the time they’re not great metric to use, but might be useful in trying to gauge a player’s style.
The last name is Roma’s Manuela Giugliano. I remember liking Giugliano at the World Cup. She was only 21-years-old then, with strong technical ability and an expansive passing range. She often looked to play teammates in behind or try to make something happen. From the figures and what I’ve watched of her, there looks to be little difference at club level.
I didn’t watch loads of her. I watched her more expansive passes, whether they were long balls, through balls, ‘smart’ passes or key passes from this season. It’s not enough to comment much, but it’s a big part of her game. Of all midfielders in the sample, she attempts the 3rd most long passes per 90. She’s completed the 10th most progressive passes per 90, so it’s not for nothing, but she also doesn’t look to be the most efficient. That’s not a massive drawback though, a lot of it looks to be built into Roma’s system from what I saw. A lot of the long passes came from her collecting the ball from the defence and hitting it long towards the flanks. Roma look to have two wingers who stretch the play, as well as attacking full-backs. If she can pick them out, they can be sent on their way. If not, the Roma midfield, in the clips I saw, did a great job of squeezing and winning the ball back quickly and quite high up. This wasn’t a massive surprise after seeing the defensive figures but was impressive when watching. It allows Roma to try and get the most out of Giugliano’s passing from deep as well as creating situations to win the ball back and look to hit the opposition in transition.
Despite being so expansive on the ball, she’s not the Roma player with the most passes into the final third this season. 33-year-old Swiss midfielder Vanessa Bernauer has more than the young Italian. Whereas most of her clips were deeper, in recent weeks, in the 2-0 win against San Marino and 4-3 win against Inter Milan, she’s played more as a #10. It’ll be interesting to see where she’s deployed more in the long term. Having her be like a quarterback is useful, but she also looks strong when drifting into the channel and combining with the wingers. She has the footwork to gain herself a yard if needed and an eye for a pass in behind. It may depend on the opposition. When Roma expect to have more of the ball, she can come deeper and look to dictate the game. If Roma want to play more on the counter, she can play further upfield and look for pockets of space to receive the ball, before turning and trying to find runners in behind.
Regardless of where she plays, she’s an exciting player to watch and looks to be in a team that is trying to get the most out of her passing range.
Looking a bit further back in midfield, I thought it’d be fun to see players with a similar output to Lia Wälti. She hasn’t stood out too much in the data yet, but she’s one of my favourite players to watch. A lot of what I like about her isn’t captured in this data, like her ambidexterity (52% right-footed on fbref. There’s a good chance she’s the player closest to 50/50 across most leagues), her press resistance and positional discipline. She not only helps get the ball upfield and offers a counterweight to the attacking midfielders Arsenal have, but she’ll also cover at centre-back if someone like Leah Williamson brings the ball forward out of defence. If I do write about Arsenal at the end of the season, Wälti will feature a lot, so I won’t say much here.
The top five players with most similar output are:
|Maite Oroz||Real Madrid||23|
Formiga’s inclusion makes a lot of sense. Both spend most of their time as the pivot in a possession-heavy side. Engen’s inclusion is more interesting, just because she’s often played in a two. Though there’s no reason why that should make a big difference, it might mean she shares the workload with Oberdorf, which could have altered her output compared to a single pivot. But I guess not.
Being unable to watch clips of the Bundesliga trio means the most interesting option for me to look into is Maite Oroz. I haven’t watched a lot though. I only watched her actions in the 1-1 draw with Sevilla. She was quite active in the game and it wasn’t against one of the bottom sides, so seemed a good choice. Things like this feel prime for confirmation bias. Watching a player through the lens of them being similar to another makes me think I’m watching, waiting for them to do something similar to Lia Wälti and then going “hey, yeah they are similar. Wow, this similar output thing is cool.” There were some Wälti-isms in there though. Some nice press resistance, being quick to sort her foot out or play out of trouble. As well as nice, quick combinations to play out from the back.
Maite Oroz has even been more progressive and defensively active than Lia Wälti this season. The Swiss international makes ~7.5 progressive passes and passes into the final third per 90. The young Spaniard makes 8.9 of the former per 90 and 7.9 of the latter. Maite Oroz also makes 11.7 recoveries per 90 (not the WyScout number on the player page but successful defensive duels + interceptions per 90). Maite Oroz does make fewer long passes. She also doesn’t look as ambidextrous, which might give her a smaller ‘potential’ passing range than Wälti, but she still looks like a promising player.
Of the inclusions, judging by a few of the numbers, Freiburg’s Janina Minge looks the most surprising inclusion. Engen and Formiga make sense as defensive-minded midfielders on stronger sides. I’m not sure about Chantal Hagel, but Minge stands out because her passing figures are much lower than the others. She’s made just 3.91 progressive passes per 90. Even if Freiburg see less of the ball, she’s also the only player of the group whose progressive passes make up less than 10% of their attempted passes. I can’t remember what it was now, but she popped up when I was looking at something else (I can’t remember what though, maybe another similarity or something along those lines) so I’d like to see more of her at some point.
Doing another quick similarity thing, the next player I thought I’d try is Manchester City’s Sam Mewis. Mewis has had a big impact on City, but perhaps one that doesn’t always look as impactful in the figures. (Though her fbref Scouting Report looks pretty good.) She’s a bit of an all-rounder. She’ll contribute to the build-up but isn’t one of the main progressive players. She’ll contribute defensively but isn’t just a ball winner. She even carries attacking threat and gets into the box to be a goalscoring option. It’s this last part that she stands out most among other midfielders and I’d guess is one of the bigger impacts on the similarity search. Few midfielders have better figures for expected goals and touches in the box for her. I’d imagine it’s also the reason why the five most similar players all come from elite clubs. The top five are:
|Sydney Lohmann||Bayern Munich||20|
|Caroline Weir||Man City||25|
Four of the above five play for teams that still have 100% records. The other is Mewis’ teammate. With how Barcelona dominate almost every metric, it’s no surprise to see their players pop up here. Of the two, Bonmati has the better figures for expected goals, but Putellas for moving the ball forward, winning it back and creating.
Bonmati’s figures for goalscoring and getting into the box are great. She has 0.53 npxG per 90, almost identical to Mewis’ 0.54 per 90, but has more touches in the box than the American. Mewis makes more recoveries per 90 (5.22 vs 4.90) but Bonmati has higher figures for progressing the ball and creating, though only slightly on the latter. Putellas then has better figures than Bonmati for progressive passes, passes into the box and xA per 90. She has more touches in the box than Mewis, but less than Bonmati, and carries less goal threat with an npxG of 0.36 per 90.
But it makes the partnership of Putellas and Bonmati sound both well balanced and a huge threat. Which is a redundant statement given their team’s form this season.
21-year-old Arianna Caruso from Juventus falls behind in most figures, except recoveries per 90 and getting the ball forward, but still has great figures.
Bayern’ Sydney Lohmann above looks interesting. She’s the most similar, but she’s also the only player who matches (betters actually) Mewis for accelerations per 90. The other three mentioned are some way below Mewis. Mewis makes 1.63 per 90. Though it’s not a huge number, this ability combined with her general physicality does feel like an important part of what Mewis offers. Lohmann looks to be the only player close to that from the group.
The young German has played in a few different roles in the midfield, whereas Mewis seems almost always stationed in that right centre-midfield slot, which could affect the numbers. But it’s hard to pick any faults in Lohmann’s figures. The table below shows a quick comparison with some of the metrics mentioned so far.
|Metric||Sam Mewis||Sydney Lohmann|
|Accurate progressive passes per 90||3.17||3.34|
|Accurate passes to the final third per 90||3.26||3.08|
|Progressive runs per 90||2.14||2.25|
|Accelerations per 90||1.63||1.75|
|Successful passes into the box per 90||1.11||2.25|
|Touches in the box per 90||5.65||4.84|
|npxG per 90||0.54||0.65|
|xA per 90||0.15||0.24|
|Recoveries per 90||5.22||5.42|
Despite using a similarity thing, I’m always amazed when looking at things like this and seeing how similar the figures are. It’s not to say Lohmann is the new Mewis or anything, but it looks like it’s been a great season for the youngster. With the lack of Bundes video, I only watched her clips from the two Ajax games in the Champions League, so I can’t say much else about her. But it’ll be interesting seeing how she develops and what role she plays over the next few seasons.
Finally, there are two Spanish youngsters I have to mention because they’re huge outliers for a couple of metrics.
First up is Claudia Pina. I wasn’t sure whether to include her here, as she’s often regarded as a forward. Looking through the line-ups, she’s played both as a forward and a midfielder. Even as a forward she looks like she drops deep though. She’s on loan to Sevilla from Barcelona and looks to have a big reputation. Not only does she play for Barcelona, but she’s enjoyed success at youth level with the national side. From a quick Google, she’s touted as a big talent within the Spanish set-up. But this season she has been a massive outlier for through balls. Over 15% of all her attempted passes have been through balls. (Through pass is defined as “a pass played into the space behind the defensive line for a teammate to contest.”) The graph below shows how no one comes close:
As you’re probably sick of hearing if you’ve stuck around this long, I haven’t seen much of her. After seeing the figures that make the above graph, I went and watched her long, key and through passes from this season. Then I watched some highlights on YouTube of her goals with the Spanish youth sides.
With the above in mind, I can’t comment much on her. But she’s exciting to watch. She attempts the 2nd most dribbles per 90 of midfielders, but not a massive number of accelerations. It suggests she won’t carry the ball long distances but instead will use her footwork to gain some space to play a pass or take a shot. This mostly matches up with the clips that I saw.
It’ll be interesting to see in the long-term whether she’ll be deployed as a forward/false 9 or in midfield. Imagining her as a false 9 at Barcelona makes sense, but having her play in the same side as Asisat Oshoala sounds more exciting considering her tendency to play through balls.
There were instances where she went for the killer pass despite a simpler pass looking the better option. But for an exciting, young player who looks to make things happen in the final third, it’s not all surprising she doesn’t always make the best choice. Especially not in her first season of regular, first division football. It’s also easier for these inconsistencies to have a bigger effect when playing for a more inconsistent team. At Barcelona, there’ll likely be more passing options and more opportunities on the ball. A lot of the clips I saw had her drop deep to progress the ball and look to play in behind or look to play in behind on the counter. At Barcelona, she’ll likely be able to stay further up the pitch as the likes of Mapi and Patri get the ball forward. Then she’ll be in the final third with wingers making runs wide, possibly a forward and another midfielder.
It’s not to absolve her of any blame for poor decisions, but that these decisions can be easier to make in a more dominant, possession-heavy side. At Sevilla, from the clips I watched, she only had the right-winger and forward making runs ahead of her. If she wanted to continue pushing forward those were her only two options. She’d try and pick them out but wasn’t always successful.
She also hasn’t put up a huge xG output, but she is Sevilla’s 2nd most productive player in front of goal. Ana Franco is the only Sevilla player with a higher npxG + xA per 90 but has around a third of the minutes in the data available.
Pina looks like an exciting talent. Exciting both because of the potential she has, but also just exciting to watch. She’s got good technique, good footwork and often looks to make things happen. It’ll be interesting to see if this loan spell has done enough to give her a chance in the Barcelona first-team set-up next season.
The next Spanish youngster is Rosa Márquez from Real Betis. The 20-year-old has been a huge outlier for dribbles and accelerations among midfielders. It’s not as bad as the previous graph, as some players come close in either category, but she’s on her own for both.
I watched a bit more of Rosa Márquez than I did Pina. Not loads, but her clips from a handful of games this season and her dribbles and accelerations from this and last season. Like Pina, she’s an exciting player to watch and one that looks to be positive a lot. She’ll often drop and receive the ball from the ‘keeper before looking to turn the player marking her or playing a quick one-two. But this does carry some risk. @StatsBombEs tweeted out her radar not long ago and it shows this. She has a great number for dribbles, a good number for deep progressions, then a horrific number for turnovers.
Rosa Márquez. Real Betis. Primera División Femenina. 2020-21. Radar vs. la media de la liga en su posición pic.twitter.com/28Q0YICtXe— StatsBombES (@StatsBombES) February 2, 2021
If the team Pina plays on affects her figures though, that’ll be bigger for Rosa Márquez. Her Betis side are fourth from bottom and only four points clear of bottom place. But, while she may not be one of the most efficient players, she might be one of the most exciting. There’s a YouTube video showing clips of her against Valencia and Atlético Madrid that shows a lot of exciting play, particularly against Atlético. Against Valencia, she has the ball higher up the pitch but often runs into a corner, takes too many touches or fails to deliver. Against Atlético Madrid though, there’s lots of great progression from deep and resisting pressure.
It’s a silly comparison. I think it’s only present because I’m currently watching season two of Six Dreams on Amazon (which features Santi Cazorla) and I’m a Wolves fan. But my first reaction when watching her was that it’s like if someone gave a player like Santi Cazorla Adama Traoré’s change of pace (I’d say Adama’s dribbling too, but Cazorla’s also great at dribbling in tight spaces).
She hasn’t carried much threat in front of goal, but she has been an important part of her team getting the ball forward, with her passing as well as her dribbling. She has the 4th most accurate progressive passes per 90 for her side, but only makes 0.3 per 90 less than the leader, then makes the most passes into the final third, the most passes into the box and deep completions.
Like Pina, and I guess most Spanish youngsters, she’s had success at youth level and was in the 2019 U19 Euro’s team of the tournament, though her side lost to France in the semi-finals that year.
She signed a contract extension at Betis until 2023 at the start of the season, which is great news for the club. As a neutral, I’d quite like to see her play for a more dominant club. It’s a similar situation to Pina. It’d be fun to see her on a team that sees more of the ball, where she’ll have more opportunities on the ball and perhaps more runners off it. Though there still feels like a lot of rawness in her game, underlined by the turnovers and risk-taking, Rosa Márquez is one of the most exciting players to watch I’ve seen in a while. The hope is she can become more efficient, without losing the bravery and progressiveness on the ball.
With this section being a bit different, I don’t think it needs a summary in the same way as the others. Now we’re onto the last section.