First off: I’m not a huge believer in statements such as ‘team X never does well in one part of the season, (Arsenal in November)’ but I found the one made by the Yellow Wall Pod’s Lars Pollmann about Leverkusen “always sucking” in the Hinrunde interesting enough to look into it deeper and it shall serve as the opening of our analysis.
It turns out that the BVB columnist was making a hyperbole, as evidence suggests that Leverkusen have been excellent in the last five years:
You can see that there are two second place finishes, a third, fifth and a sixth place before the ninth place Hinrunde finish this year. (Yes, technically there is a game still to be played for everyone and Leverkusen could jump ahead of Freiburg (hosting Bayern) if they beat Hertha, but that’s as far as they can go after 17 matches). Die Werkself went on to finish fifth in 11\12 (improving one spot in the Rückrunde), third in 12\13 (dropping one position), fourth in 13\14 (going down two places) and 14\15 (down one more) and third last year (up one spot) ,in the spring that is to say on average Leverkusen actually are TWO POSITIONS WORSE in the Rückrunde….
So Roger Schmidt and co are in unfamiliar territory, especially considering that a few experts and one coach (Pál Dárdai of Hertha BSC) actually picked Bayer to win the Bundesliga. In the following post we are going to take a look at some of the factors that have contributed to a dismal Hinrunde and ask the question whether Leverkusen can get themselves out of this hole.
Raw stats, i.e. shots, shots, shots!
The first thing we notice is that shot volume is down by a ton, per Footcharts: 12.5 shots per game ranks them eighth in the league and well-below their numbers for last year (14.25) the year before (16.29) or 2013\14 (14.53) 12\13 (16.35) and even their 2011\12 season featured 12.71 shot attempts!
The shots against numbers are from 2011\12 to 2016\17 in order: 13.52, 11.94, 13.29, and as the Roger Schmidt effect kicks in: 8.79, 10.65, 10.69 – so on that end Bayer are seemingly doing OK, but this is beginning to look a lot like the Pre-Roger Leverkusen!
Shots on Target – from 11\12 to now – were 5.38, 6.41, 5.44, 5.65, 5.44, 4.75 while
SOT against Leverkusen were : 4.38, 3.62, 4.32, 3.76, 4.21, 4.44
So in short while in the previous five season Leverkusen always had NET advantage of at least PLUS ONE SHOT ON TARGET per game, that number is now down to .29!!!
The advanced stats
A quick glance at their advanced stats shows that they are two and a half goals ahead of XG in terms of scoring, and have conceded four more than their expected value of 20. That 5.4 expected goal difference would rank them fifth behind Bayern, RBL, BVB and Cologne. Instead they are -6.4 in terms of XG difference and actual goal difference (5.4 plus negative one), which is only above Ingolstadt and Gladbach in the Bundesliga. 8.7 shots needed for a goal is OK and right in line with the challengers to Bayern’s throne, though it’s noteworthy that Hertha and Mainz who spent very little on their attack are on either side of the SEVEN shot needed for a goal mark. Per the Challengerspod we find that Bayer’s .127 expected goal per shot value is second best behind Leipzig, which is just lots of fun, even if we suspect that number to be skewed by Joel Pohjanpalo and Chicharito’s hot-streaks. And it turns out to be true: That XG per shot started out at .155 in late September after five games, then dropped to .137 by October 20th, was at .133 in late November, and is now at .127 or 12.7% if you prefer it that way. So it looks like Leverkusen’s advanced offensive metrics are pretty good, despite producing just 23 goals and part of that could just be down to awful luck in one spot: Leverkusen for one have already missed ALL FOUR OF THEIR PENALTIES, per Transfermarkt. To put that into context, they were awarded just three all last season (converted two) and the closest team in the last five years to do this poorly was Greuther Fürth in 2012\13 who missed three of four! The four misses probably could have given Bayer some points:
- Chicharito missed his first one vs Frankfurt in the 87th minute, costing at least a point in a game that finished 2-1 for SGE
- In the very next game it was Charles Aranguiz’s turn to miss in the 72nd minute in a scoreless draw vs Augsburg that could’ve also easily yielded three points.
- Hakan Calhanoglu missed another, up 2-1 vs Leipzig in a match that they lost 3-2, so there’s a decent argument for three points there.
- Finally, Hernandez failed to score once again from the spot, once again in the 87th minute in a 1-1 draw against Freiburg that Leverkusen dominated so much that Freiburg keeper Schwolow was named into the team of the week.
If you want to look at it optimistically that’s EIGHT points which would have them on 29 pts, tied with Frankfurt for fourth….
Of course it doesn’t quite work that way, but it’s a fun little exercise and coupled with the solid advanced stats – Footballintheclouds has them above the 10.2% league average conversion rate at 11.7% overall, with a 32% shooting percentage (29% is avg), while converting 18% of shots in the box (avg. is 14.4) among others – it’s enough evidence to take the blame off the offense.
That blame should go on the defense according to advanced metrics:
- opponents need just SEVEN shots to score vs die Werkself, who if you remember are conceding 10.69, so that’s quite a hole do dig yourself every game.
- The expected goal value of a shot against Leverkusen is .118 or 11.8%, the fifth worst mark in the league!
- They have just three clean sheets – one against ten man Schalke for 86 minutes, the other versus an FCA side that has scored the second fewest goals in the league, with the masterclass versus Tuchel’s Dortmund (the boy who cried FOUL) the lone outlier.
- the opponent conversion rate is staggering: 13.5% against is the worst in the Bundesliga.
- The location of those goals is interesting: shots in the box and open play conversion rates vs Bayer are like 10% higher than normal, but HOLY SET PIECES, BATMAN: 16% converted against them on set pieces – when the league average is 6%!!! and only Wolfsburg are in double digits!
Individual blame aka defensive scapegoating
Bernd Leno should bear a fair bit of criticism, especially early on in the season. He is ranked tied for eighth by Kicker, just 12th of 18 by Squawka and 11th of 17th at Whoscored among goalkeepers who have played at least half the games. His howler against Bobby Wood is the obvious pick here, though I would like to ask what in the world Ömer Toprak was doing…
Speaking of Toprak, who has been one of the outstanding defenders in the Bundesliga for the last few years, there two interesting developments:
1. BVB have signed him for the cheap price of 12 million Euros, which might be the first time ever that a player (even if his contract expires this summer) gets sold below his Transfermarkt value (20 million).
2. he’s rated as the third worst CB in the league by Whoscored in the company of Alexander Milosevic of Darmstadt and the immortal Johan Djourou
Kicker has him as 42 of 86, which sounds more reasonable, but you have to wonder why Leverkusen are not renewing his contract at age 27, even if they did buy Aleksandar Dragovic for 18 million Euros from Dynamo Kyev this summer. Moreover, Dragovic has played just 624 minutes, though more of late, as Toprak missed two games twice with minor injuries. Dragovic had one outstanding game against Freiburg and was pretty good against Wolfsburg, but failed to impress in the last couple of games against Ingolstadt and Cologne.
Is Toprak no longer good enough for Bayer, or is he too injury-prone? Does it make sense to sell him to one of your biggest rivals, if his high-priced replacement has not exactly fit in. I guess the answer to those questions is Jonathan Tah, who is one of the best CB talents in world football.
The giant former HSV defender is 194 cms and 95 kgs, and despite turning 20 in November, he is closing in on 5000 Bundesliga minutes and has solidified his spot in the Werkself starting XI. The opinion on his Hinrunde performance is divided to say the least: he is ranked as the fourth best defender by Kicker (who somehow think Philipp Lahm is number one??), but is just rated average by Whoscored with 20th of 39 qualified CBs and 42nd of 94 defenders by Squawka, ironically a spot below Dragovic. The discrepancy might be a result of weighing errors differently: Squawka says that Tah leads the Bundesliga (alongside Djourou) with FOUR defensive errors, though I only recall the opening day gaffe that put Andre Hahn in to score Gladbach’s first goal.
On the wings, Leverkusen appear set with one of the bright spots of the season in Benny Henrichs and Wendell, who are young, hungry and tackling hard, as confirmed by our friends at Footballradars.
— Football Radars (@FussballRadars) December 27, 2016
Defenders are of course not solely responsible for defending, as midfield problems could amplify problems, and that’s the case in an extreme high risk\high reward pressing and gegenpressing system like Leverkusen’s: once opponents break it, it’s most likely a great spot to score (thus driving down the shots needed per goal ratios and increasing XG per shot values)
When you couple that system with two extremely active wing backs (see the high tackling numbers for Henrichs and Wendell) the burden on the central midfielders to cover ground, intercept passes and recover enough to stop attacks is enormous. Leverkusen looked to be quite set at that position before the season, as they were ostensibly adding two new players in Charles Aranguiz (signed in the previous summer, but missed all but the last couple games of 2015\16) and Julian “the heart and soul of Austria and Mainz” Baumgartlinger to an already deep core (Lars Bender has been a well-respected CDM, while all-around midfield lynchpin Kevin Kampl, and the attack-minded Hakan Calhanoglu was also on the team). So what happened?
Lars Bender has played 335 minutes in the league, and 90 of them in two stints (Dortmund, HSV) at right back as he struggled to come back from the Olympics before going down for seven games with heel pain in late November. Kevin Kampl and Calhanoglu were forced to play every single position, due to the devastating Kimo Bellarabi injury in week 2, the early slump of Julian Brandt, the disastrous Kevin Volland signing and the Chicharito slump (without a goal since October). I really mean every position:
Kampl has probably been Leverkusen’s most consistent player
The correct answer to my Cazorla similarity riddle is:
KEVIN KAMPL. Bayer Leverkusen 16-17. Can play on the wing. Great passer + mover. pic.twitter.com/nzgeG4ANJB
— Ted Knutson (@mixedknuts) December 1, 2016
and Hakan, while still taking too many long shots (when 67% of all your shots are from outside the box and it’s a career low, that’s probably not ideal) has four goals and four assists, basically equalling last year’s production of three goals and five helpers in half the minutes played (1037 to 2265). Brandt overcame his post-Rio blues to notch five assists in the last seven games for Leverkusen.
That’s enough good news:
here are all the key passes into the box by Julian Baumgartlinger and Charles Aranguiz for the Hinrunde! It’s safe to say that the focus is not on offense, (you can check who is responsible for that by clicking here) so they must be elite on defense?
Aranguiz who should have been the perfect box to box midfielder, due to his Chile experience with Jorge Sampaoli, remains a tantalizing, physical player with a couple of elite skills (dribbling, interceptions). Yet, he ranks 42nd of 52nd qualified players (seven games played) at Whoscored.com and keeps making dumb mistakes like the one against Ingolstadt where he got trapped, lost the ball, then elbowed an FCI player five seconds after and got sent off. 40% of duels won is also quite bad from a guy with a reputation as a tough-tackler. He also adds very little offensively, averaging 0.9 key passes and 0.8 shots per 90. It’s no wonder Kicker has him as one of the ten worst midfielders…
Julian Baumgartlinger is sadly on a lot of those lists – though not on Kicker, because he played only 515 minutes, despite there being a ton available due to the absences and performances of the players above. He didn’t make the most of his opportunity in the earlier parts of the season, as a tepid and slow display against Gladbach’s high-tempo offense (at the time, certainly not nowadays) and another impactless match against Eintracht landed him in Roger Schmidt’s doghouse.
Although he is a reliable distributor and solid passer of the ball, Baumgartlinger needs time to pick out the rare forward passes (the assist to Kevin Volland vs Sportfreunde Lotten was one of those times where the 3. Liga side gave him ample time), but he just doesn’t do them often and\or well enough under pressure. Last season at Mainz, he was able to create 21 chances, but that number is just at four at the midway point so far. On the surface, his numbers do look similar to the player that he was at Mainz, and the radars confirm this:
What’s surprising about these radars isn’t really its similarity, but the fact that despite Baumgartlinger being the leading tackler in the Bundesliga with 3.7 per game, that it was mostly a byproduct of Mainz’s overall philosophy (19.5 tackles were fifth in the league) and how little it translated into Leverkusen’s system (having to cede the lead in tackles to RBL with 19.9 after 20.5 tackles per match last year). For Julian, it appears that Leverkusen paid for the outlier season, as he attempts more or less 4.5 tackles per 90 in his career, but went off for 6.1 last season.
His success rate is lower as well, along with most other defensive metrics (which have basically haved), which is either down to having to (and so far failing to) adapt to a different system under Schmidt or Leverkusen potentially misevaluating him as a starter. It really is quite a shame for one of the more likable players in Germany (check out this rather thoughtful interview with der Standard) and it’s a decent idea for die Werkself to try to accept the 6m Euro offer from RSC Anderlecht and thus turn a 2m profit on what looked like quite a steal in the summer transfer window.
The final nail in the Aranguiz – Baumgartlinger partnership came in the Schalke game and in this tweet: https://twitter.com/BundesPL/status/807995034700222464
after which of course Roger Schmidt subbed them both out for 17-year-old Kai Havertz, and Vlad Yurchenko who had played 90 minutes on the season prior to this!
Sideways passing and the inability to stop anyone on the counter? Good god, that’s Sebastian Rode music!!!
Other random notes
- The early season injuries to Bellarabi, the two broken hands (Volland + Chicharito) and the Olympic hangover to Brandt and Bender definitely affected Bayer’s start
- They have this tendency to play down\up to the opponent’s level and struggle versus teams they should handle, and if their forwards are on (see Pohjanpalo vs HSV or Hernandez’s hat-trick vs Mainz) they can overcome them, and sometimes stuff just doesn’t go in (Freiburg\Augsburg) but there were too many games were lesser teams outplayed them: Eintracht, Werder and Ingolstadt come to mind.
- The flip side is the three games against the elite sides – BVB, RBL and Bayern, where they arguably had a great case to have won all three and were probably unlucky not to have done so.
- Admir Mehmedi, despite his three goals, offers very little and is ranked 31\35 among forwards by Kicker, and it’s hard to make a case for Chicharito being good, when he basically had three good games all fall. That said, he is still the team’s top scorer with five and it’s not like a ton of their players are undershooting their XG numbers per Alex Rathke
- The Pohjanpalo miracle story was cool, but you’ve to wonder why he’s getting a chance to play with Volland, Kiessling, Hernandez, Mehmedi and all their other midfielders who can play upfront….
- The three big money summer signings that were supposed to give Leverkusen have all been awful: Volland for 20m, Dragovic for 18 and Baumgartlinger for 4 and one could argue that neither has been a Bundesliga starter level player. None has played over 720 minutes – or half the available 1440 minutes, in fact their combined minutes total is just 1500!!
- Roger Schmidt probably deserves his own column, for some of his mistakes, but that is another day’s work.