Grouping playing styles in 2. Bundesliga: a tactical analysis using Kognia data

The challenge: Applying Kognia tactical data to get a clear understanding of team similarity in Germany’s second division.

The data: 35 tactical metrics from Kognia

We grouped 35 tactical metrics grouped into three blocks reflecting different phases of positional play:

• Attack (17 metrics): Organised attacking actions, box arrivals and finishing. Includes metrics such as shots, goals, key passes, crosses, runs into the box, receptions inside the box, dribbles into the final third, passes behind the defensive line and exploitation of defensive imbalances, among others.

• Defence (10 metrics): Defensive organisation, pressing and ball recovery. Includes metrics such as defending runs behind the line, interceptions, blocking finishing opportunities, marking inside the box, neutralising the opponent's defensive line imbalances, recovered balls, pressing during organised pressure and identifying passing lines under pressure.

• Passing (8 metrics): Progression and combinative play. Includes metrics such as filtering passes between lines, third-man play, overcoming opponents with vertical passes, vertical and horizontal overcoming support, and emergency support.

  
  

Further information on our approach can be found in the appendix.

The results: 4 tactical families in 2. Bundesliga

More proactive and Combinative play: Hannover 96, Paderborn and Hertha BSC

This quadrant groups teams that combine high progression and elaborate build-up capacity with active pressing when out of possession. These teams hunt the ball when they lose it and circulate it with purpose when they have it.

• Hannover 96 are the most extreme team in the quadrant and one of the most extreme in the entire league on the horizontal axis with the highest percentiles in 2. Bundesliga for runs into the box and third-man play. Their style combines verticality with elaborate build-up from the back, positioning them as the most tactically ambitious team in the league.

• Paderborn are the most balanced team in the quadrant. With high percentiles across all three categories they have the most complete profile in the competition. Their position on the chart places them close to Hannover, and they form the most similar pair in the league with a similarity index of 62%. 

• Hertha BSC occupy the more moderate zone of the quadrant. Their clearest strength lies in 1v1 situations and ball-carrying (the league's highest percentile in dribbling into the final third) and in play behind the defensive line, although their finishing and box occupation metrics are more modest.

More reactive and Combinative play: Magdeburg, Elversberg and Dynamo Dresden

Perhaps the most surprising quadrant. All three teams combine high offensive and progression activity with low pressing activity. These are teams that wait, defend in a low block and when they recover the ball have the technical and tactical ability to progress with purpose and arrive in the opposition box with danger.

• Magdeburg are the most extreme team in the offensive block across the entire analysis, with five metrics at the 100th percentile (shots, dribbles, key passes, receptions after runs into the box and exploitation of defensive imbalances). They are the most vertical and direct-to-goal team in the league, which combined with their low-block profile draws a very defined style: deep defensive shape and fast transition with high individual quality in the final third.

• Elversberg have the inverse profile within the quadrant: their strength lies not in shooting but in filtering passes between lines (100th percentile) and in receiving between lines. 

• Dynamo Dresden are the most balanced of the three, with a broad profile that includes strengths in crossing, runs into the box and marking, but without standing out in any single metric in an extreme way.

The most similar pairs in this quadrant show considerable distances, reflecting that the three teams, while sharing the same tactical family, have internally very differentiated styles.

More proactive and Direct play: Greuther Fürth, Preußen Münster, Holstein Kiel, Fortuna Düsseldorf and Karlsruher

This quadrant groups teams that press high or with intensity, but whose build-up and progression tends to be more direct.

• Greuther Fürth are the most extreme team in defensive pressing across the entire league. However, their offensive production is the lowest in the competition.

• Preußen Münster are in the high position on the vertical axis. That does not mean they press high in the conventional sense. The vertical axis also captures build-up elaboration metrics, which inflates their vertical position -further information about our approach in the appendix-. Münster are the clearest proof that tactical sophistication in possession and offensive output are not the same thing. Their defensive profile confirms this pressing identity: Münster rank 94th percentile in ‘moving forward during organized pressure’ and 82nd in ‘recovered ball’; two of the clearest indicators of a team that actively hunts the ball high up the pitch.

• Fortuna Düsseldorf are the most active crossing team in the entire league, combined with very high defensive pressing. They are a team that presses high, recovers and attacks quickly with balls into the box.

• Karlsruher and Holstein Kiel both combine an active defensive profile with moderate build-up and limited offensive production.

More reactive and Direct play: Braunschweig, Schalke 04, Kaiserslautern, Arminia Bielefeld, Bochum and Nürnberg

The most populated quadrant in the league with six teams. It groups teams that defend in a low or medium-low block and whose attack tends to be more direct, with less elaboration in the build-up phase. It is the most tactically conservative quadrant.

Braunschweig are the most extreme team in the quadrant and the most extreme at the defensive end of the entire league, and practically all offensive production at very low percentiles.

• Schalke 04 and Kaiserslautern share the lower end of the vertical axis, indicating that both have a profile particularly marked by direct play into the box and crossing, with barely any build-up elaboration.

• Arminia Bielefeld are also positioned in that zone: they are the team with the greatest box space occupation in the entire league, but with very low build-up metrics.

• Bochum and Nürnberg are the most moderate in the quadrant, close to the centre of the chart, with more balanced profiles albeit still tending towards a low block.

Conclusion

Kognia's tactical metrics applied here capture real and consistent playing patterns that allow the identification of tactical families, similar pairs of teams and extreme stylistic contrasts.

For analysts, coaches and directors, this type of analysis has direct applications in searching for tactical replacements, opposition analysis, or in identifying playing model references.

Bonus content

To conclude, the chart below allows us to visualise the impact of the three managerial changes. The most striking case is Preußen Münster: under Schwartz, the team leaves its position as the league's most elaborate build-up side and shifts towards a more direct and low-block profile; the most pronounced tactical transformation of the three.

Holstein Kiel under Walter shows a more moderate and barely perceptible movement on the map, reflecting a gradual adaptation without a break from the team's prior identity.

Eintracht Braunschweig under Kornetka starts from the league's most defensive extreme and shows early signs of opening up, though eight matches are still insufficient to produce a significant displacement.

Appendix

Some notes on the approach taken:

With 35 metrics per team, the similarity between teams cannot be represented in a conventional chart. The solution is Principal Component Analysis (PCA), a statistical technique that transforms a set of correlated variables into a smaller number of independent components, ordered by the amount of variance they explain.

The process was as follows:

1. Data standardisation. Before applying PCA, we normalised all variables to have zero mean and unit standard deviation. This prevents metrics with very different numerical ranges from distorting the result.

2. PCA application. We reduced the 35 dimensions to two principal components. The first component (horizontal axis) captures mainly the difference between teams with greater offensive and progression activity and teams with greater defensive line activity. We interpreted this as the Direct play ↔ Combinative play axis, where teams further to the right combine more to progress and those on the left tend towards a more vertical and direct approach towards the box.

   The second component (vertical axis) captures mainly the difference between teams with high third-man play, passes between lines and elaborate support -at the top- versus teams with more shots, box occupation and crosses -at the bottom-. We interpreted this as the More reactive ↔ More proactive axis, where teams at the top press higher and build more elaborately from recovery, while those at the bottom tend to sit deeper and attack more directly into the box.

3. Categorisation into four quadrants. The intersection of the two axes naturally generates four tactical quadrants, each with a distinct playing identity.

4. Similarity calculation between pairs. To identify the most similar teams within each quadrant, we calculated the Euclidean distance between all pairs of teams in the space of the 35 original metrics and converted it into a similarity index from 0 to 100%.

5. Explained variance conclusions. The PCA applied to the 35 tactical metrics explains approximately 50% of the total variance between teams across the two represented axes. This means the chart captures half of the tactical information available in the data, specifically the two dimensions that most differentiate teams from one another. Far from being a limitation, this result is consistent with the nature of football: playing style is multidimensional and no pair of axes can fully summarise it. The chart should therefore be read as a highly useful tactical approximation, not as an absolute truth, that allows tendencies, team families and stylistic contrasts to be identified on a solid statistical basis.

An important note on interpretation: PCA is a dimensionality reduction tool, not an absolute truth. The chart captures the two dimensions of greatest tactical variation between teams, but football is more complex than two axes. The quadrants are tendencies, not rigid categories, and some teams on the boundary between quadrants may have mixed profiles that the chart does not fully reflect.