With football returning to Germany and the Bundesliga ready to take centre stage, I have put myself in a small, Werder Bremen-shaped pair of shoes. With Die Werderaner 4 points off escaping the automatic relegation places, and also hopeful of avoiding a playoff, the 4-time champions are going to have to pull a rabbit out of the hat. Luckily, Bremen's very own magician, Milot Rashica, has helped keep the side afloat with 7 goals and 4 assists this season, contributing to over 40% of their goals. Staying up or relegation bound, one thing that is looking more and more likely is the sale of Rashica, most likely to title contenders, RB Leipzig. With at least 25 million euros potentially becoming part of the clubs budget, I looked to evaluate where that money could be best spent.
One spot where Werder Bremen have been weak is their defence, particularly the last line of defence. Jiri Pavlenka has been the man between the sticks this season, but it's been a poor showing. The 6"5 Czech Republican has drastically under-performed this season in terms of his post-shot xG conceded, in comparison to actual goals conceded, with a PSxG +/- of -0.51 per90, the worst of any 'keeper across the top 5 leagues (minimum 900 minutes). Essentially this means that, based on the quality of shots on target against them, for every goal the team is expected to concede each game, they're conceding 1 and half. This has really added up, in terms of goals conceded, and points lost, through the season.
This analysis is all about finding a goalkeeper that can replace Pavlenka for next season. Based upon the assumption that Rashica can help keep Bremen in the Bundesliga, and is then sold for a large sum, this analysis aims to find a top quality keeper that fits the Werder way. The requirements set are that the 'keeper plays in the top 5 leagues, has a profile that suits the club and offers value for money. This allows the analysis to be a very applied example, and ticks the three key boxes in recruitment (in my opinion): quality, compatibility and value.
Inspired by Mark Carey (@MarkCarey93), I ran a Principal Component Analysis (PCA), on every goalkeeper that fit the criteria. The PCA aims to reduce a large set of metrics into broader components, so that correlating variables are compounded into key factors. The important thing with a PCA is that it doesn't pigeon-hole certain variables into just 1 component, therefore it allows us to identify which players show skillsets across different profiles. A PCA simplifies profiling as it allows us to analyse across a few key styles, as opposed to lots of individual metrics. As shown below in the component matrix, the PCA split the goalkeepers into 4 key playing styles, based on the 15 event metrics: Aggressors, Line Defenders, Sweepers and Protectors. The darker the green, the greater correlation between each metric and component, and the red represents negative correlation.
Each metric, normalised as per 90 minutes, and standardised using z-scores. All data from f-bref.com and wyscout.
Factor 1: Aggressor
These type of goalkeepers typically play long when in possession and like to be the start of attacks. They have to deal with a lot of crosses and a fair few shots, but do so well.
Factor 2: Line Defender
These 'keepers face a high volume of action and do not like to move too far from the safety of their 6-yard box.
Factor 3: Sweeper
Sweeper's tend to complete a lot of actions outside of the box, position themself far from the goal-line, and also like to play when they have the ball.
Factor 4: Protector
This style of goalkeeper is a 'saving first and the rest is next' style. They're top quality shot stoppers and pretty good at dealing with crosses.
Profiling goalkeepers using this method allows us to find a goalkeeper that is of high compatibility to what is needed. In this case, for Werder Bremen. After analysing Pavlenka's play style, as a predominantly line defender GK, and assessing the needs of the club, I decided that Werder should look for a line defender 'keeper that has protector traits. This is because, due to the low quality of Bremen and the likelihood that they're not going to dominate games and the goalie may often see a lot of action, a line defender will identify players that are used to being bombarded, and that they're key strengths are not in-possession. To improve the quality of the replacement GK, having protector traits will allow us to find a goalkeeper that first and foremost, stops the ball going past him. Below shows the goalkeepers that have made the list to further analyse as they meet the style criteria, and are also younger than 28. (Also highlighted is Pavlenka, a line defender with no protector qualities).
Approaching a deeper analysis, looking at shot quality, dealing with crosses and passing ability, will give us a more rigorous approach to identifying a keeper that offers the best quality and best value. Firstly looking at the sweeping and passing ability of the group. As a team, Werder typically like to play short and on the deck, and they don't stretch the opponents back line often. A keeper that shows capabilities with their feet would be ideal as an 11th player, in a side that likes to look after possession when they can, keeping 48.8% of the ball per game, more than any other of the bottom half teams. Looking at the graph below, Rulli is the Montpellier man that plays a lot of passes, suggesting a comfortability on the ball. Juan Musso, shows up as a player that doesn't leave his line much, but is also pretty good with his feet.
In terms of dealing with crosses (graph below), the 'keepers that face a similar number to that of Bremen's current 'keeper (10.91 p90) are Gomis and Larsonneur. And it's Gomis of Dijon, standing at 6"4, that is the stand out with his ability to deal with them.
Finally, a look at these goalkeepers and their shot stopping ability (below). Walter Benitez, Alfred Gomis and Gautier Larsonneur are the best shot stoppers, with over 0.3 PSxG +/- p90. All three play in Ligue 1 in France. Its Gomis and Rulli who stand out as the top stoppers, and Walter Benitez also looks good, considering the amount of shots he faces every game.
Based upon this analysis, contemplating each of the three key aspects, quality, compatibility and value, I would suggest that the three top targets to replace Pavlenka are:
Alfred Gomis
Geronimo Rulli
Gautier Larsonneur
Gomis, Dijon's 26 year old Senegalese stopper recently signed in the summer from SPAL, and looking likely to survive this season, they may be reluctant to let go of him. Transfermarkt has him at 2 million, but this could well rise come the summer window
Montpellier's Argentinian, Rulli, on loan from Real Sociedad for the season, could come with a larger price tag at 8 million. However the Blue and Whites may be looking to shift him permanently in the summer with current sweeper, Remiro, having a strong season. Rulli could be a more experienced keeper that offers more dominance and agility between the sticks.
Larsonneur of Brest, the youngest of the selection at 23, has been in superb form, helping drive Stade out of danger in Ligue 1. At 5"11, a shorter keeper, offers brilliant shot stopping ability, and looks a player when the balls at his feet. With a 5 million price on his head, and a great profile, could Larsonneur be the man to kick Werder on?
Looking at each of their profiles (above) in terms of style, the perfect fit looks to be Alfred Gomis. Although there may be a lack of lethal strikers testing Gomis in France, as a goalkeeper that leaves his line fast, is solid in 1v1's and dominant in aerial situations, Gomis would be my choice to step in to Pavlenka's shoes at Bremen. He has a lot to prove still at the top level but is young enough to do so and the challenge of the Bundesliga will suit his style. Gomis shows obvious quality and compatibility as well as good value for money at, what will be just crumbs left from the Rashica departure and I believe that, from this analysis process, a top goalkeeper has been identified.
Thanks for reading, any feedback would be brilliant! Stay safe.
Jackson - @JacksonHuxley
Comentarios