A Look at Borussia Dortmund’s Advanced Stats on FBref

9 min readApr 19, 2020

Last week, FBref (the world football statistical reference site from the Sports Reference crew) released a huge update. You can read all about it on their blog, or you can check out this video I recorded right after the launch.

Full disclosure: I worked closely with the FBref team on this release. I’ve been consulting with them for over five years and have specifically helped with FBref since its launch. But today I’m talking about it as a user and a Borussia Dortmund fan.

There are so many new statistics available on the site now. Here’s a look at some key ones and what they tell us about our beloved black and yellows.

Post-Shot Expected Goals

This has been a rough season for Roman Bürki from a statistical perspective. Despite the club’s second place standing in the league, his more traditional stats have taken a hit. His .522 save percentage is the lowest of his career (by far) and his 1.47 goals allowed per 90 minutes are his highest since he was 20 years old with Grasshopper Club Zürich. While save percentage and goals against average are fairly easy to understand, they aren’t accurate reflections of a goalkeeper’s skill. Not all goals allowed are the fault of the goalkeeper and not all shots faced are of equal difficulty.

FBref extensively uses expected goals (xG) data supplied by Statsbomb. While outfield players have xG, goalkeepers have xG against. FBref uses PSxG (Post-Shot xG) to measure goal likelihood of the shots on target a goalkeeper faced.

Post-Shot Expected Goals
PSxG is expected goals based on the fact that the shot is on target. xG totals include penalty kicks, but do not include penalty shootouts (unless otherwise noted).

PSxG also eliminates own goals (which is important because Roman Bürki has been the victim of three) since own goals do not have an xG value. Bürki has allowed 32 goals — 29 after removing the own goals. His PSxG against is 23.8, meaning he has surrendered 5.2 more goals than expected based on the shots on target he faced. This is… not good. Of the 18 qualifying (played at last half their team’s minutes) Bundesliga keepers, he ranks 15th. This is a huge drop from last season where his PSxG was basically in line with his goals against.

But Bürki’s PSxG issues are somehow not carrying over to Europe. Bürki’s PSxG+/– per 90 minutes is –0.24. In the Champions League, his per 90 rate is +0.24.

Marwin Hitz has played only 289 minutes in the Bundesliga, but he has fared well. He has surrendered one goal with a PSxG of 2.7. His PSxG+/– per 90 minutes of +0.51 ranks fourth in the league when you remove minute minimums. Just like Bürki has fared quite differently in the Bundesliga vs. the Champions League, it’s a safe bet that if FBref had xG data for the DFB-Pokal we’d see a similar trend with Hitz. In the Cup, he’s allowing an extra goal per 90 minutes and saving 20% fewer shots.

Progressive Passing Distance

FBref is now a treasure trove of passing data. In a previous release, they included completion percentage broken down by short, medium, and long passes. But now we have total distance for all passes and (even better) progressive distance.

Progressive Distance
Total distance, in yards, that completed passes have traveled towards the opponent’s goal. Note: Passes away from opponent’s goal are counted as zero progressive yards.

Achraf Hakimi has attempted (1,771) and completed (1,487) the most passes on the team. But the leader in total pass distance is Mats Hummels (33,756 yards on 1,451 completed passes). Of Hummels’ 33,756 yards, 10,223 have been progressively towards the goal (which also leads the team).

I took a look at the percentage of a player’s yards that were progressive and the average length of all of their passes (these are not on the site, but they’re easy to get from a CSV export). Interestingly, Hummels, Dan-Axel Zagadou, and Manuel Akanji all average 21.0 to 23.3 yards per pass and 30.3% to 31.0% of their passing yardage progressively toward the goal. The Dortmund center backs definitely have a style. The center mids in front of them are very different. Weigl had a similar average pass distance to the CBs while Axel Witsel and Emre Can are shorter. All three have a lower percentage of their passes moving toward goal.

It makes sense — the more space you have behind you the more you can use it. For the most part, the further back you play, the higher percentage of your yards are progressive. But the center mids are the exception. Witsel, Can, and Thomas Delaney all have a lower progressive percentage than even Erling Haaland (while Weigl and Dahoud are slightly above Haaland).

Pass Types

Additionally, there is a separate table on FBref for pass types. This breaks down a player’s attempts by several different criteria. For example:

Overall attempts (as noted above, Achraf Hakimi leads with 1,771)
Live ball passes (Hummels leads with 1,602)
Dead ball passes (Hakimi leads with 212 thanks to his high number of throw ins)
Throw ins (Hakimi has 190)
Free kick passes (Hummels again with 45)
Switches, or passes that travel at least 40 yards across the pitch (Hummels leads again with 60)
Crosses (Thorgan Hazard leads the way with 94)
Corner Kicks (Hazard with 68)
Passes under pressure (Hakimi had 277)

Since Hakimi had the most passes, it makes sense that he had the most passes under pressure. On a percentage basis, Haaland leads the way (with 30.5% of passes made under pressure) but somewhat surprisingly Thomas Delaney is second with 23.8%. Reus (23.3%), Hazard (22.6%), Sancho (21.2% ), and Brandt (20.1%) are next.

There’s also a breakdown by pass height (where you can see Hakimi has the most passes on the ground with 1,290 but Axel Witsel has more on a percentage basis with 86.2%). Passes are broken down by body part where you can see that Raphaël Guerreiro is 95.1% left-footed and Lukasz Piszczek is 93.7% right footed. The most two-footed player on the team is Mats Hummels (32.5% left, 67.5% right). Hummels has also made the most passes with his head (86).

Passes are also broken down by the outcomes, such as completed, offside, out-of-bounds, intercepted, and blocked. I took a look at those last two in particular.

Achraf Hakimi had the most passes intercepted (40) and blocked (64), but he also attempted the most passes. Who has had the highest percentage of their attempted passes intercepted? That would be Paco Alcácer — and not by a little. An astounding 7.1% of his attempted passes were intercepted. Next is Marco Reus with 3.8%. In a small sample, Jacob Bruun Larsen saw 7.4% of his passes blocked. Next is Erling Haaland with 5.5%.

Goal and Shot Creation

FBref has introduced an entirely new concept with this release— shot and goal creating actions (SCA and GCA).

Shot-Creating Actions
The two offensive actions directly leading to a shot, such as passes, dribbles and drawing fouls. Note: A single player can receive credit for multiple actions and the shot-taker can also receive credit.

This is an attempt to use more than the standard assist (and even more than the hockey assists that I’ve been tracking) to see who is actually contributing to goal creation.

First, here are the leaders for the different times of shot-creating actions:

Live ball passes: Sancho, 79
Dead ball passes: Hazard, 20
Dribbles: Hazard and Sancho, 7
Shots: Reus, 6 (these are shots that were rebounded and led to other shots)
Fouled: Hakimi, 3
Total SCA: Sancho (98), Hazard (83), Hakimi (62), Brandt (56), Reus (53)
SCA per 90 minutes: Reyna (5.21), Sancho (4.88), Hazard (4.46), Dahoud (3.66), Brandt (3.36)

Here are the same leaders, but this time for goal-creating actions:

Live ball passes: Sancho, 22
Dead ball passes: Sancho, 2
Dribbles: Hakimi, 3
Shots: Sancho and Guerreiro, 2
Fouled: Haaland, 1
Total GCA: Sancho (28), Hakimi (17), Hazard (17), Brandt (11), Reus (8), Witsel (8)
GCA per 90 minutes: Sancho (1.39), Bruun Larsen (1.38), Hazard (0.91), Haaland (0.88), Hakimi (0.75)

Like the hockey assists I’ve been tracking, these numbers really hammer home how dominant Jadon Sancho has been and how many contributions Julian Brandt is making that are not reflected in traditional stats.

Defensive Actions

Here, we can see a player’s tackles, pressures, and more. Mats Hummels leads the team with 43 tackles attempted, just ahead of Hakimi (41). On a rate basis, Emre Can is attempting more than a tackle and a half more per 90 minutes than Hummels (3.62 vs. 2.03). Thorgan Hazard has won the most tackles, ensuring BVB won possession of the ball on 30 of his 34 attempts. Julian Brandt also has won a high percentage of his tackles, taking 26 of 33.

Hakimi leads with 27 attempts in the defensive zone while Witsel has the most in the middle third (19) and Marco Reus has the most up top (9).

FBref also specifically has tackles against opposing dribblers. The top two players in tackles vs. dribblers are Lukasz Piszczek and Achraf Hakimi. But look at these percentages:

Piszczek: 18 tackles in 27 attempts, 66.7% success rate
Hakimi: 17 tackles in 52 attempts, 32.7% success rate

The only BVB player with a better percentage than Piszczek is Emre Can with 9 wins in 13 (69.2%).

Thorgan Hazard has applied the most pressures with 463. Marco Reus is second with 442.

Pressures
Number of times applying pressure to opposing player who is receiving, carrying or releasing the ball

While the pressing forwards apply the most pressures, they don’t have the highest success rate. A successful pressure is one where BVB regains the ball within five seconds of the pressure. On a rate basis, the defenders’ success rate is much higher with Dan-Axel Zagadou successful 46.6% of the time.

Hummels has made the most blocks on the team with 40 (12 shots and 28 passes). Hakimi is second with 34 (2 shots, 32 passes) and Piszczek is third with 32 (12 shots, 20 passes). Hummels also leads the way in interceptions (43) and clearances (143, more than twice as many as any other player).

Possession

Lastly, we have possession. This table contains touches, but also carries and dribbles.

Unsurprisingly, Achraf Hakimi leads the team in total touches with 2,070. On a per 90 basis, the team leader is Lukasz Piszczek with 97.0 touches (when you remove the small samples of Leonardo Balerdi and Mahmoud Dahoud). FBref has a breakdown of where on the pitch these touches occur (each third of the pitch as well as both penalty areas).

It’s important to distinguish between dribbles and carries. Dribbles are take-ons while carries occur when a player simply moves with the ball at their feet.

Jadon Sancho has attempted the most dribbles in the entire Bundesliga with 129. His number of successful dribbles (71) and number of players dribbled past (77) also lead the league. Achraf Hakimi isn’t far behind with 57 successful dribbles in 97 attempts (62 players dribbled past).

In terms of dibbling success rate, there are some players excelling in small samples, like Dan-Axel Zagadou (6-for-6), Mahmoud Dahoud (5-for-6), and Gio Reyna (9-for-11). Axel Witsel tops the list among the regulars at 16-for-21 (76.2%).

BVB is among the naughtiest team in the Bundesliga when it comes to nutmegs. Three of the league’s top seven players are in black and yellow. Sancho is tied for second with 8 and Guerreiro and Hakimi are right behind him with 6.

I’m really excited about the new carry data on FBref. Similar to the passing updates, we now have total distance carried and progressive distance carried. With this type of progressive yardage, we now can look at things through an NFL-style “passing and rushing yardage” lens.

I created this table combining the passing and carrying yards (and calculating progressive yards per 90 minutes).

Hummels’ progressive passing distance leads the team and is fifth among outfield players in the Bundesliga. He’s also first on the team on a per-game basis (if you leave out Balerdi’s 12 minutes). Hummels is sixth in the Bundesliga in progressive carry distance, but is third on the team (Hakimi leads the Bundesliga and Sancho is fifth).

Add it all up, and Hummels is progressing the ball more than 700 yards toward the goal each game. The only outfield players who have played more than half the season with a better per-game total are David Alaba (842) and Dayot Upamecamo (782).

Lastly, the posession table includes targets, or how often each player is passed to. I recommend you read a recent article on Saturdays on Couch about target share based on FBref data. There’s a great chart for each team.

I hope you’ll check out all the advanced stats on FBref and let me know what you think. You can always find me on Twitter at @fussballtwit.