Dortmund Analytics Thread

As is likely the case with most of us, I’m frustrated with the season we’ve had thus far. I think we can all agree injuries, age/maturity, and an imbalanced squad have all played into this. With that said, I wanted to put some of my pre-conceived notions to the test, and see how we’ve done from an analytical perspective – and thought I would share (I didn't want to derail any existing threads in case this doesn't interest some of you).

I don’t have direct access to Opta data (which would be great), but I managed to gather some data from a multitude of different sources (transfermarkt, squawka, espn, RN, Wahretabelle, SZ, and other blogs that have access to Opta).

Feel free to tune out or disagree – these are just my elementary thoughts.

INJURIES

We all knew we were going into the season without Reus, but the magnitude of injuries the club has endured this hinrunde has been well beyond the norm. The chart below outlines the number of man games lost in the Bundesliga this year. Dortmund has had more than double the number of man games missed compared to prior year, and more than any team other than Frankfurt.



Perhaps even more importantly it’s “who” has been missing. Other teams were primarily missing role players, although Bremen and Leverkusen were missing Kruse and Bellarabi. We on the other hand were missing key pieces such as Reus, Bender, Guerreiro, Schurrle, and Burki.

Lastly, unlike other squads who have been missing some players for extended periods – our injuries have been across the spectrum. Every player on our squad with the exceptions of PEA, Mor, Dembele, Pulisic, Weigl, and Weidenfeller have missed games due to injury. That means we have had 20 players who have missed at least 1 or more games to injury. Ginter is the only one who has only missed 1, everyone else has missed 2 or more.


FOULS

I’ll admit I don’t like Tuchel complaining to the media about the treatment of his players, but I equally don’t like the physicality the team is having to endure. I’m sure that the constant fouls lead in some way to the injuries noted above. I would argue one of the reasons for the lack of calls is the age of our players – a foul that would warrant a card on Lahm or Robben, does not apparently warrant a card on Dembele or Pulisic.

The charts below illustrate that Dortmund has been fouled more than any other team in the league – that margin is material (almost 2 fouls per game more). Additionally, we have by far the largest variance between the fouls we’ve committed (192) and fouls we’ve had committed on us (302).



With all that said, the fouls have not come without cards. During the hinrunde, the Bundesliga average was approximately 1 card every 8 fouls (that includes all teams over 16 match days). In contrast to that, teams that foul Dortmund have received on average 1 card every 6.8 fouls – meaning that despite our protestations, the referees are drawing the yellow cards on our opponents.

We could potentially argue that there are some fouls that aren’t called, or some second yellows are staying the pocket, however I think it’s a fairer assessment to say the following:

1. We lack the ability to punish teams for those fouls through good set piece execution – thereby teams know they can play more physical against us without fear of conceding.
2. Due to our youth and style of play, the fouls disrupt our flow and get our players off their game. It comes down to mentality - and if we are going to be faced with more of this type of tactical fouling in the ruckrunde, we will need to adapt our play accordingly. That means being more cynical ourselves, and focusing on set pieces.
​

GOALKEEPING

After struggling for portions of last year, Burki settled into a good rhythm at the start of this season and made some key stops for the club early on. Subsequent to his injury against Bayern, we have only won 2 of 6 Buli matches with Weidenfeller between the sticks.

I think we can all agree Weidenfeller has been terrible, but let’s quantify just how terrible that is:

1. We have trailed 0-1 in every single game he has started in the Buli.
2. He has not kept a single clean sheet.
3. Outlined below is his save % in the Bundesliga in contrast to other keepers (Buli only). At this point he is saving marginally more than 50% of the shots on goal – compared to nearly 80% for Burki, Neuer, Baumann, and Horn. If you add in non-Buli games his save % is actually below 50%.
4. Last season we maintained a clean sheet 38% of the time, this season to date it’s 19% (0% under Weidenfeller).​

With goalkeeping of that quality, it’s little doubt we can’t win any matches. It not only destabilizes the defensive line, it also completely demoralizes the team to start every match down 0-1.


DEFENSE

Our defense has seemingly been terrible this year – but it’s been tough to define the exact problem. It looks to be death by a thousand cuts rather than one key problem, although some of these drop-offs are clearly related to the departure of Hummels (as you’ll see).

The problem isn’t error rate in the sense of individual errors (turnovers resulting in goals) – which for example are killing Hamburg and BMG. We still make defensive errors, but depending on whose data you trust we have made fewer errors this year than we did at this point last year.

There are three problem areas that stick out to me thus far – the first two of which can be largely linked to the departure of Hummels.

1. Defensive interceptions (breaking up the transition play of the other squad). We have the lowest number of defensive interceptions in the league this year – by a wide margin. To put that into context, we have 243 whereas Bayern, Leipzig, and Cologne have 319, 317, and 310 respectively. This is about reading and anticipating the game (something that Hummels excelled at), and something that Ginter and Bartra don’t yet have in their arsenal.

2. Aerial Duels won. Last year we won approximately 53% of our aerial duels – which is adequate however not dominant by any means. This year that numbers has fallen to an abysmal 45% (compared to Bayern at 55%), and this was an area where we traditionally excelled under Klopp (we were nearly 57%). This again is an area where Hummels excels, and not surprisingly he has the highest rankings in the Bundesliga in the hinrunde. It’s also an area that Barta specifically seems to struggle. His timing on his jumps, and his positional awareness of when to commit are still not there right now. I believe it may come with more time on the pitch, however it’s impacting our game right now. Sokratis is quite good (albeit not great) in the air with 64% won.

I've outlined our 3 CB's along with the other top defenders below:



3. Formation. Statistically speaking our best formation has been a 4-1-4-1 with Bartra and Sokratis as the anchoring CB’s. I see no problem running a 3 man backline, however the concern I have is with regards to Tuchel’s interpretation of that formation – most specifically the WB’s he’s using. Tuchel spreads out his CB’s incredibly wide in the back in an effort to provide more room for our midfield to operate – dragging the opposition wingers out further and opening up the middle of the pitch. The problem is that this is almost suicidal when you have Pulisic as one wingback, and Schmelzer as the other. When Schmelzer plays as a wingback – he is typically further up the pitch than the midfielders and almost in line with Auba. This is one issue where the defense is structured to support a weaker midfield where we lack an 8 – and I could see this issue solving itself with the introduction of someone like Dahoud.

I'll be posting part 2 tomorrow - focusing on offense, ExpG, set pieces, possession, heat maps, and shot selection /chance creation.

 

Offense hardly looks to be much of an issue this year given the number of goals we’ve scored – but there are a few interesting/concerning stats below the surface. This isn’t surprising given the multitude of new players and the amount of injuries – but it’ something we should hopefully see some resolution to in the Ruckrunde (especially if we can add an 8).

INTERESTING FACT
You can tell we are a very immature team where momentum plays a key role. In every game where we have scored first (6 games) we have won the game, whereas in games in which we have conceded first (9 games), we have only won once (1 win, 3 loses, and 5 draws)

CHANCE CREATION
Overall chance creation is still relatively good at just under 11 per game – which is 2nd behind Bayern, who are in a different stratosphere than anyone else in this area (as noted on the chart below). One reason for Bayern’s dominance in this area is the incredible play of Thiago this season – who has now become the anchor of that creative midfield for the Bavarians. He along with Muller and Costa have created a combined 68 chances!

BVB has a strong concentration of created chances amongst two players – Dembele and Castro who have 25 and 21 respectively. Gotze has only created 14, and Weigl 4 (which is understandable for Weigl given his deep lying role). Gotze’s stats are somewhat concerning here – however I’ll add some detail to hat later.

Bayern has a lot of players who are secondary contributors with key passes/assists, which is something that BVB is missing. They have 6 players who have created 19 chances or more (Thiago, Muller, Alaba, Ribery, Alonso, and Costa), whereas Dortmund only has 2 – with Gotze third at 14.




CHANCE CREATION – ZONES
Under Tuchel Dortmund’s chance creation comes primarily through the middle of the pitch – zones 14 & 17. That hasn’t changed thus far this year, and we continue to focus on developing chances through the spine of the pitch (which is typical of TT, Pep, and other more possession oriented teams). Tuchel uses his FB/WB’s and typically one of the wingers to spread out the opposition and create passing lanes through the middle. Even when you look at someone like Dembele, the vast majority of the chances he creates (72%), have come through the middle.

Interestingly Bayern have really (materially) changed in this aspect under Ancelotti – as they now create a lot more chances down the wings than they did under Pep.

I should note that various sources categorize the zones slightly differently – meaning that some sources may have slightly different numbers.




SHOTS
Outlined below is a table of our shots (and below that a visual diagram of them). You’ll notice there is a relatively strong correlation between TSR and that teams relative place on the Buli table (albeit for us you need to find a way to account for just how horrendous Weidenfeller has been between the sticks). TSR is far from a perfect indicator, but it does provide a decent baseline.






CHANCE CONVERSION
Unlike the 2014/2015 campaign – chance conversion is likely BVB’s strongest asset this year. Our conversion both from general shots and shots on target has been tops in the league, and much of this comes down to the position of those shots (all our goals have been from within the 18 yard box). Most of our chances are “good” chances that are being developed through the middle of the pitch. (G/S = Goals per Shot & G/SoT = Goals per shot on target).




GOALS SCORED
Although our chance conversion has been very good – right now the squad is relatively one dimensional with regards to how we create and execute on these chances. When looking at the top 20 player combinations in the league (assist and goal scorer), we have the highest ranked combination – that is Dembele assisting Auba resulting in 7 goals. However, beyond that pairing we don’t have a single other combination within the top 20 in the league – whereas Bayern has 3.

Add to that the fact that we are terrible when it comes to set pieces or conversion or aerial crosses. We have scored only two goals on crosses (both Auba), and 3 goals on corners (Pisz and Auba). Comparatively speaking Bayern has scored on 2 set pieces, 4 crosses, and 2 corners. There admittedly isn’t a huge difference in this area of play (RB is perhaps even worse than we are), but given the talent we have and the sheer volume of opportunities we have had – it’s notable.

PASSING
Our passing has been relatively consistent with last year (albeit down from 622/game to 591/game). The composition also hasn't changed a ton, albeit we are actually passing backwards less than we did last year (which I wouldn't have guessed).

When you look at the metrics with regards to our passing structure, it really comes down to a slight reduction in almost all categories rather than any one issue. It’s not that things have fundamentally changed, but simply that we are not quite at the same level of execution as we were last year (which is to be expected given the personnel changes).

• Pass completion down from 85% to 84% (not overly material)
• Key passes down from 10.1/game to 8.9 (compared to Bayern at 13)
• Possession down from 59% last year (full year) to 58% this year

MARIO GOTZE
Now comes the contentious part – trying to analyze individual player stats. Obviously at an individual level player stats are highly subjective and need to be understood in context to their roles, however I’m just throwing my perspective on a few players out there. If certain individuals decide to turn this into a Kagawa v. Gotze thread, I’ll just ask @MS to delete the posts or delete this thread entirely.

When Mario made his move back to the Westfalenstadion, I think we all knew we were getting a player who lacked confidence & fitness, and would need some time to find his rhythm again. He has been a disappointment during the hinrunde – but I still believe he can be a good addition to this club. A few key things to note with Gotze thus far:

Shots: He still hasn’t found his space in this offense – and his shot totals speak to this. In his last season with Dortmund before leaving for Bavaria he was averaging almost 2 shots per game. Since his return he has been averaging 0.6 shots per game…. hardly what one would expect from a key cog in the offense. His role (base position) plays a big factor here, but I’ll touch on that later.

Dribbling: One of the key attributes that Gotze traditionally brought to a game was his ability to create in small spaces, and facilitate for others. A significant aspect of that was his dribbling and ability to challenge defenders. In his 12/13 season, Gotze was averaging nearly 3 successful dribbles per game and over 3 per 90min. This season he is averaging just under 1 per game compared to approximately 4 or 5 per game from Dembele (again it depends on the source). Now they are completely different players, as Dembele has otherworldly quickness (something Gotze has never had), however this is nonetheless a dramatic drop-off.

Key Passes and Assists: During his 12/13 prime, Gotze was assisting approximately once every 3 games, and averaging around 2 key passes per game (break-open passes). Since his return, the drop-off here is similar to the drop-off in other areas, as he’s only had 1 assist this season, and is averaging 1.4 key pass per match.

Positioning: Now here comes the contextual piece that is needed to make sense of all these stats – Mario’s role in the squad thus far. In his prime under Klopp, Gotze was the quintessential 10 – often having the ball at his feet in the attacking third, and having permission to roam the pitch to find pockets of space and create for Lewa, Reus, etc. Under Tuchel we have seen Mario’s role change considerably, which might in large part be due to necessity. To highlight this, I’ve outlined Mario’s heatmaps for the 4 matches he played 90min in thus far this season. His base position is clearly much deeper, and in some cases much wider than I believe he’s suited for. Interestingly, the only game he played more in line with a true 10 was early in the season against Freiburg, and even in that game he wasn’t moved forward until the 2nd half.




OUSMANE DEMBELE
While I knew he was an interesting prospect when we signed him, Dembele has exceeded every expectation I think anyone set out for him during the hinrunde. His decision making and maturity still need some work, but that has already shown improvement as the season has progressed. His athleticism and skill on the ball make him almost impossible to defend 1vs1, and from a statistical perspective he and Auba have been our 2 best players outside of defense.

To put some perspective on what he has accomplished, outlined below is his radar chart in comparison to Gotze in his prime (12/13) thanks to @footballradars. This really highlights 2 things; how talented Dembele is at 19, and just how good Gotze was.



The radar might be tough to read, but here are a few key stats for you (based on 90min played):




I’ll be adding some player stats for Pulisic, Bartra, Weigl, and the wonder that has been Rode shortly.

 

Just ignore him please. Please!

Sent from my SM-G900W8 using Tapatalk

 

Thanks for your note - appreciate it.

I absolutely agree with regards to the 90min stats, and that's one reason I never put much emphasis on the per 90min goal stats for Ramos. While Ramos has done some great things, I would expect him to have a high scoring rate when he's coming on in the 75th minute with the sole intention of getting into the box.

 

MATS HUMMELS
Of all the personnel losses we had during the summer, the stats highlight that Mats Hummels has likely been the most significant departure. One could very well argue that Gundogan was just as (or perhaps more) significant, but I’ll state my reasoning as follows:

1. We went from a team that had a solid aerial presence and won a good number of aerial battles on the defensive end of the pitch (Mats is best in the league in this area), to being one of the worst teams in this area during the hinrunde. Bartra has struggled mightily here. An example of the importance of this is RB Leipzig, who are not great defensively (statistically), with the exception of being dominant in the air.
2. We are also one of the worst teams in the league with regards to defensive interceptions. This again was an area we had some strength in with Hummels in the backline, as he was very good at reading and breaking up the opposition build-up play.
3. During this hinrunde we had the lowest number of longballs in the league – our build-up play is relatively predictable. Tuchels tactics are obviously the primary driver here, but this was a dynamic that Hummels provided that we have definitely lost.

GONZALO CASTRO
My personal opinion on Castro is that he is a good player who can provide solid coverage at the 8 against most Buli teams. However, he does not have the skillset required to breakdown top tier teams (Bayern, RB Leipzig, or Champions League competitors).

Chance Creation – Outside of Dembele, Castro has created the most number of chances on this squad with 21. He also has the 5th highest performance score on Squawka, behind Auba, Dembele, Weigl, and Sokratis, and the 2nd most assist behind Dembele at 4. This is undoubtedly impressive! What is very interesting however is that Castro’s numbers drop materially in this area against better competition – specifically in the Champions League or against teams like RB Leipzig where he was completely bossed and had his worst performance of the year. In easier matches he has nearly 2 key passes per game, yet in the CL and against RB Leipzig he was averaging barely half that, which personally I attribute to being afforded less time and space.

Dribbles – While Castro has solid passing stats and very good completion %’s, his dribbling leaves something to be desired. While I don’t expect him to be Dembele, he is averaging approximately 0.5 successful dribbles per game – which is less than any other attacking player on the team outside of Ramos. It’s tough to be a top tier 8 on this squad when you don’t have the ability to attack defenses with any degree of proficiency. Putting that into perspective – Naby Keita is averaging 2.9 this season. Side note – Gotze has also been quite terrible in this area thus far, but still materially better than Castro (approximately double).

Passing – Castro’s passing % is very solid. Outside of Weigl and Kagawa (who has a limited sample size), Castro has the best passing % of our midfield/attacking players at 86%. With that said, he does utilize a lot of back passes – especially against better competition. I’ve only looked at the passing maps for the matches against RB Leipzig (0-1 L) and Freiburg (3-1 W). In the match against RB Leipzig Castro had only 14 passes, only 2 of which were in a forward direction. This is an extreme example, but I intend to do some more digging on this.

Castro is a player where I actually want to spend some time breaking down game specific stats.....

 

Kicker also rated Hummels as the best defender of the hinrunde, and this is clearly supported by the statistics. He has dominated the air, done a solid job in 50-50 tackles, and made very few errors.

I will say that I agree with @bvbSlash, that the midfield that is in front of Hummels has a lot to do with this.

 

Going to spend some time on this tomorrow - but at face value the Toprak move makes no sense. We have weakness in the air and lost a lot of positional intelligence (interceptions).... and Toprak doesn't solve either of those. He is bad in the air, and relatively average in his break-up play.

My only thought is that Toprak is being brought in as a veteran on the bench (3rd or 4th on the depth chart). I still don't like that.....

 

For all the nerds out there.... here's an interesting article on calculating possession stats.

https://chanceanalytics.com/2017/01/15/are-we-calculating-possession-all-wrong/

 

Planning on reviving this thread for a year-end review. @MatthausSammer could you please rename the thread "Dortmund Analytics Thread".

Interesting Stat off the Season which describes the season for me:

After all the excitement surrounding the return of Gotzeus - the played a sum total of 97 minutes on the pitch together (not including extra time) this season in all competitions. They only played three games together, and never more than 41 minutes in any of those three games.

 

Interesting video (especially the 2nd half), highlighting BVB's tactical changes under Tuchel.

Interesting statistical look at Dortmund over the past three years from Opta. (h/t @TobiHerrmann8) https://t.co/vyIbcqW6zL

— Lewis (@LGAmbrose) June 19, 2017