Figure 1:
Flow chart of methodology used for current research.
Table. 1
Competitions Used in Data Sample
| League | Territory | Seasons |
|---|---|---|
| 1. Bundesliga | Germany | 2020/21; 2021/22; 2022/23 |
| Bundesliga | Austria | 2020/21; 2021/22; 2022/23 |
| Champions League | Europe | 2020/21; 2021/22; 2022/23 |
| Championship | England | 2020/21; 2021/22; 2022/23 |
| Eredivisie | Netherlands | 2020/21; 2021/22; 2022/23 |
| Jupiler Pro League | Belgium | 2020/21; 2021/22; 2022/23 |
| La Liga | Spain | 2020/21; 2021/22; 2022/23 |
| Liga Nos | Portugal | 2020/21; 2021/22; 2022/23 |
| Liga Profesional | Argentina | 2021; 2022 |
| Ligue 1 | France | 2020/21; 2021/22; 2022/23 |
| Premier League | England | 2020/21; 2021/22; 2022/23 |
| Serie A | Italy | 2020/21; 2021/22; 2022/23 |
| Série A | Brazil | 2021; 2022 |
| Super League | Switzerland | 2020/21; 2021/22; 2022/23 |
| Superliga | Denmark | 2020/21; 2021/22; 2022/23 |
| UEFA Europa Conference League | Europe | 2020/21; 2021/22; 2022/23 |
| UEFA Europa League | Europe | 2020/21; 2021/22; 2022/23 |
Table 2:
Positional Identifiers
| Statsbomb Position Label | Positional Identifier |
|---|---|
| Centre Attacking Midfielder | Central / Attacking Midfielders |
| Left Centre Midfielder | Central / Attacking Midfielders |
| Right Centre Midfielder | Central / Attacking Midfielders |
| Centre Back | Centre Back |
| Left Centre Back | Centre Back |
| Right Centre Back | Centre Back |
| Centre Defensive Midfielder | Defensive Midfield |
| Left Defensive Midfielder | Defensive Midfield |
| Right Defensive Midfielder | Defensive Midfield |
| Centre Forward | Forward |
| Left Centre Forward | Forward |
| Right Centre Forward | Forward |
| Left Back | Fullback / Wing Back |
| Left Wing Back | Fullback / Wing Back |
| Right Back | Fullback / Wing Back |
| Right Wing Back | Fullback / Wing Back |
| Goalkeeper | Goalkeeper |
| Left Attacking Midfielder | Winger |
| Left Midfielder | Winger |
| Left Wing | Winger |
| Right Attacking Midfielder | Winger |
| Right Midfielder | Winger |
| Right Wing | Winger |
Figure 2.
Bar Chart with SHAP values for the most important variables for broad position classification.
Figure 3.
Positional Classification following Uniform Manifold Approximation and Projection (UMAP).
Figure 4.
Positional Classification following Uniform Manifold Approximation and Projection (UMAP) and subsequent Gaussian Mixture Model (GMM) Clustering.
Figure 5.
SHAP Value Plot for Winger Classified Group to Identify Optimal Pressing KPIs
Figure 6
(a) Ridge plot displaying distribution of optimal variables within each cluster. (b) UMAP projection onto 2-dimensional space followed by GMM of optimal pressing variables.
Figure 7:
UMAP projection onto 2-dimensional space of players with closest similarity to Eberechi Eze from the 2022/2023 Season.
Table 3:
Table of Players Similar to Eberechi Eze following Similarity Search
| Player | Competition | Team | % Similarity | Distance |
|---|---|---|---|---|
| Wilfried Zaha | Premier League | Crystal Palace | 94.82851 | 0.36 |
| Simon Zoller | 1. Bundesliga | Bochum | 93.78860 | 0.43 |
| Nicola Sansone | Serie A | Bologna | 92.53857 | 0.51 |
| Amath Ndiaye | La Liga | Mallorca | 91.36516 | 0.59 |
| Hwang Hee-Chan | Premier League | Wolverhampton Wanderers | 90.60578 | 0.64 |
Figure 8:
(a) Dimensionality reduction & (b) lollipop chart displaying the evolving pressing role for Eberechi Eze over a 3-year season span.
Figure 9:
Boxplot showing the significant difference between pressing teams and aggressive actions.