Atletico Madrid’s recent loss to RB Leipzig in Champions League meant that they ended their 2019/20 campaign trophyless and 17 points off rivals Real Madrid, many would suggest that Atletico Madrid are better than this. This loss came days after Jan Oblak told Spanish newspaper AS that he will ‘assess the season when the last game is over’. This will worry Atletico Madrid fans as Oblak has become one of the best goalkeepers in the world as he made 17 clean sheets last season, while only conceding 27 goals.
Oblak is reported to have a release cause of £110 million and this is likely to put off many suitors however Chelsea have been heavily linked to the Slovenian as they look to replace Kepa Arrizabalaga and their recent transfer embargo means that they have funds saved to spend this summer. This data analysis will take a look at possible replacements for Oblak if he were to leave.
This is the player profile of Jan Oblak and the statistics highlight the skills and traits of Oblak. Primary skills (goalkeeping) will feature more heavily in this analysis than secondary skills (technical) in the search for replacements as Oblak doesn’t use his technical skills that often compared to the median goalkeeper. Oblak is rarely involved in Atletico Madrid passing transitions in the defensive third, he only made 12.57 passes per 90 last season and 57% of his passes were long.
He is shown to excel in shot-stopping ranking highly in save percentage and having one of the lowest goals to shots on target ratio in the database. He also has a tendency to prefer to catch the ball rather than punch having a claim/punch ratio of 7.0. Lastly the data shows that he is deployed in the goalkeeper role rather than the sweeper-keeper role having only made 1.01 exits per 90.
For this data analysis, nine leagues were included in the database and the idea behind this is to build a greater database which will give us more choice in the shortlist to find the very best. The leagues included were the European top five leagues and five other interesting leagues which were Liga NOS, Pro League, Eredivisie, Brazilian Serie A and Argentine Primera División.
Goalkeepers had to play a minimum of 50% of their teams’ minutes last season to warrant being involved in the data analysis as they need to be able to compete in both La Liga and the Champions League, two leagues of high quality.
Those older than 30 were not included in this data analysis because Oblak is aged 27-years old and we assume that Atletico Madrid will want a goalkeeper that is at a similar stage of their career. This data analysis will mainly look at the data which displays the goalkeeper’s primary skills however we will look briefly at their secondary skills.
The first test will analyse the difference between the goalkeepers’ goals conceded and expected goals conceded, this will indicate strength in shot-stopping if the goalkeepers are able to save more goals than expected.
Oblak conceded 0.05 fewer goals than expected per 90, Oblak performs well in this test having conceded only 0.67 goals per 90. He performs well but not outstanding and this is partly due to Atletico Madrid’s defensive resolve where they limit the opposition’s shots on target, those that are on target are of high quality making it hard for Oblak to save, hence the 0.05.
Athletico Paranaense goalkeeper Santos performs particularly well here having conceded 0.70 goals per 90 and conceded 7.65 less goals than expected, his strength in saving shots helped his side get out of trouble on a number of occasions as they finished fifth last season.
Esteban Andrada and Predrag Rajković also conceded less than one goal per game at 0.33 and 0.93 goals per 90 respectively. Their expected goals minus goals per 90 was at a difference of +0.34 and +0.30 goals per 90 respectively. Rajković played for Stade Reims last season and conceded below one goal per game, impressive given Ligue 1 is in the top 5 leagues and his side conceded less goals than European giant PSG.
Jordi of Vasco de Gama conceded 0.41 less goals than expected per 90 which is extremely impressive given that Andrada ranks second to him at a difference of +0.34 and Jordi did not have one of the strongest defences in the league in front of him when on loan at Centro Sportivo Alagoano.
It is worth noting Aitor Fernández and Janis Blaswich, they both conceded more than one goal per game however had an expected goals minus goals value of +0.31 and +0.34 respectively.
Ability to save
This test is similar to the previous. It can tell us how successful each goalkeeper is in saving shots, however, it can also tell us the goalkeeper’s concentration levels.
Oblak faced 2.96 shots on target per 90 and those who have faced a similar amount must show that they can save a high percentage, indicating their strong concentration levels to make saves even when inactive for large periods of the game. Those that face more than Oblak can also show their strength in saving shots if they save a high percentage and face a high number.
Andrada performs particularly well in this test having faced 3.36 shots per 90 and he saved 90.28% shots that he faced.
Marco Bizot and Rajković follow for those who faced less than average shots having saved 83.58% and 81.25% of their shots faced respectively. Both players’ and Andrada’s concentration and discipline levels could be seen as strong, a skill the next Atletico Madrid goalkeeper will need.
Another strong performer is Santos who saved 84.16% of his 4.42 shots faced per 90. Compatriot Jordi also performed well having faced 6.02 shots per 90, the third most of the goalkeepers and still saved an impressive 78.95% of shots which shows his strength in saving shots.
It is assumed that the quality in the top 5 leagues is the highest in the world and Gerónimo Rulli, Alessio Cragno and Fernández all perform well considering they play for a team in the top 5 leagues. Rulli faced more shots than Oblak and saved 77.87% of his shots faced, a similar save rate to Oblak, he has experience in La Liga and this is a plus towards his cause.
The next test will analyse how likely the goalkeepers will concede given that the shot they face is on target. This is key to understand the goalkeeper’s reliability, if the team can trust the goalkeeper then they have more confidence within the game. Atletico Madrid will invite the opposition to attack and the Atletico Madrid defence will need to trust their goalkeeper if they are to perform best in their defensive shape.
Oblak performed extremely well in the test, he conceded 0.22 goals per shot on target last season and this allowed his teammates to excel and defend aggressively as they trust Oblak to save shots.
South American goalkeepers Andrada and Santos impress again having a goals/shots on target ratio of 0.08 and 0.16 respectively. They have performed outstanding so far, however, they play in the Argentine and Brazilian leagues respectively and this has to be taken into account as the attackers in these leagues may not be as efficient in front of goal as La Liga.
Jordi, Fernández and Cragno all perform at a similar level to Oblak in this test as they faced a relatively high number of shots on target per 90, Jordi conceded 0.20 goals per shot on target, Fernández and Cragno both 0.22 goals per shot on target.
Tiago Volpi outperforms Oblak as he had a goal/shot on target ratio of 0.20 last season and faced a similar number of shots on target per 90 at 3.51 shots.
Robin Olsen, Juan Musso and Bernd Leno performed at a similar level to Oblak while facing more shots than him. Leno competed in the Europa League and it would be interesting to see how he fared in the Champions League.
Claiming the pressure
Jan Oblak is characterised as liking to catch the ball rather than punching, he had a claim/punch ratio of 7.0 last season as he only made 0.07 punches per 90 last season and made. Claiming rather than punching is effective as it relieves the opposition pressure and prevents the opposition gaining possession from the punch and scoring.
Diego Simeone will likely want a goalkeeper with a similar style of play to Oblak. The claim/punch ratio also indicates the goalkeeper must be strong at catching having claimed so many balls per punches.
Mattia Perin and Joel Robles outperform the others significantly as they have a claim/punch ratio of 22 and 21 per 90. This is a trait that Simeone would like as they help to relieve pressure from the opposition and potentially create a counter-attack which Atletico Madrid enjoy to do at times, this is dependent on their technical ability, analysed in the next test.
After Perin and Robles, Zack Steffen and David Soria prove to have the claiming trait having had a 12.0 and 11.33 claim/punch ratio per 90 respectively. Steffen played at Fortuna Dusseldorf last season and now returns to his parent club Manchester City however Soria is currently contracted at Getafe and helped them have the fourth-best defence in La Liga last season, his experience in La Liga would allow for a smooth transition if he were to sign.
It is worth noting Jordi, Nick Pope and Alfred Gomis. Their ratios were lower than 7.0 showing they like to punch as well as claim however they made the most claims per 90, this is likely to be because their teams set up in a defensive shape therefore inviting the opposition to attack allowing for more opportunities for the goalkeeper to claim. Jordi made 2.01 claims per 90, Pope made 1.73 and Gomis made 1.51. Jordi made the most claims but only had a claim/punch ratio of 3.80 per 90.
Aiding progressive transitions
The final test will use statistics to analyse each goalkeeper’s long-distance passing and their ability to successfully make progressive passes, this has been part of Atletico Madrid’s play as Oblak made 4.71 successful progressive passes per 90 last season.
A progressive pass is one that travels a significant distance, the distance classification depends on what third the pass comes from, when a pass is made in the defensive third it has to progress at least 10 metres to be classed as progressive.
Etienne Vaessen and Lucas Cháves rank among the best when it comes to long passing distribution. They made the most long passes last season while having a long pass accuracy of 70.76% and 69.69%, both similar to Oblak’s 70% pass accuracy. They also impressively completed 9.38 and 9.92 progressive passes per 90, Cháves had a progressive pass accuracy of 88.13% and this would be key for Atletico Madrid as they look win loose balls and aerial duels in the opposition defensive third.
Gabriel Arias and Rulli may have made less than average long passes however when they did pass long they were highly accurate having a pass accuracy of 80.37% and 76.45%. Arias made 5.12 successful progressive passes per 90 with an accuracy of 92.05%, he performs well here as he would be a great asset when Atletico want to transition quickly from defence to attack.
From the data we have made a shortlist of three players who could replace Oblak. Aitor Fernández, Lucas Cháves, Santos and Juan Musso all came close to making the shortlist. It is also worth noting Jordi who would be a good addition in the future and could transition into Europe before making the jump to a Champions League team. The graphic below displays the three players shortlisted and how they compare to Oblak.
Andrada was one of the best performers in the tests concerning shot stopping, in the graph above he ranks first out of the shortlisted players in the first three tests. What is most impressive is his 90.28% save percentage and 0.10 goals/shot on target ratio, these statistics are strong evidence that he is a quality shot-stopper as he helped Boca Juniors only concede seven goals in 20 games.
It has to be taken into consideration that La Liga may well have more clinical shooters than Argentine Primera División, however, he has been capped four times for Argentina and started their last two matches which shows his quality and potential.
Another Argentine that has impressed, Rulli was loaned out to Montpellier last season however has since returned to parent club Real Sociedad where he was first choice goalkeeper for the previous four seasons. He displayed similar traits to Oblak in the data analysis as he had a claim/punch ratio of 5.0 and saved 77.87% of shots he faced.
Rulli’s experience in Europe is a bonus compared to his compatriot Andrada however he didn’t compete at the same level as Andrada for many of the tests.
The Serbian could be a great investment for Atletico Madrid given that he is a 24-year old meaning that he is likely to have not hit his goalkeeping peak yet. His statistics are impressive given his age and the quality of the league that he plays in, Ligue 1, he only conceded 0.63 goals per 90 last season for Stade Reims and ranked highly in the first three primary skill tests as he appears a strong shot-stopper.
He had a 3.63 claim/punch ratio, however, still made at least one claim per game and was exposed to 3.35 shots per 90, he displays strong concentration and discipline for a young player having saved 81% of these shots.
This data analysis has highlighted goalkeepers that could replace Jan Oblak as well as goalkeepers that could be cheaper alternatives to Oblak. Chelsea could look towards these alternatives, if Oblak’s release clause is too much money for them. It must be taken into account that skills such as reflexes, positioning and communication are difficult to quantify in statistics therefore further video analysis would need to take place to understand the players’ qualities further.
Those shortlisted all have a market value of £8 million, 10% of the estimated market value of Oblak (his release clause is greater than that), this shows that there are potential bargains for players that may have gone under the radar. As this was being written, Roma were linked to Andrada and if he were to join a European club it will be interesting to see how he fares. We now ponder, what goalkeeper will Atletico Madrid choose to buy if Oblak were to leave? Will they sign a promising goalkeeper with similar traits as Oblak or will they opt towards someone with more experience?