Tier-specific contextualised high-intensity running profiles in the English Premier League: more on-ball movement at the top

Abstract

The present study aimed to determine the physical-tactical profiles of elite football teams and individual players according to final league rankings. A total of 50 English Premier League matches (n = 100 match and 583 player observations) were analysed by coding the player's physical-tactical actions through synchronising tracking data and video. Final league rankings were categorised into Tiers: (A) 1^st-5^th ranking (n = 25), (B) 6^th-10^th ranking (n = 26), (C) 11^th-15^th ranking (n = 26), and (D) 16^th-20^th ranking (n = 23). One-way analyses of variance were used to compare match performances between different Tiers, and effect size (ES) was determined for the meaningfulness of the difference. Tier A teams covered 39-51% more high-intensity distance for 'Move to Receive/Exploit Space' (ES: 1.3-1.6, P < 0.01) and 'Run with Ball' (ES: 0.9-1.0, P < 0.05) than Tier C and D, and 23-94% more distance for 'Over/Underlap' (ES: 1.0, P < 0.01), 'Run in Behind/Penetrate' (ES: 0.7, P < 0.05), and 'Break into Box' (ES: 0.9, P < 0.05) compared to Tier C. Central and Wide Defensive Players in Tier A covered 65-551% more high-intensity 'Move to Receive/Exploit Space' distance compared to other Tiers (ES: 0.6-1.0, P < 0.01). Moreover, the additional options within the physical-tactical actions and zonal differences unveiled more meaningful insights into 'HOW' the top Tier teams physically and tactically perform. Thus, the amalgamated physical-tactical data help improve our understanding of a team's playing style relative to their competitive standard.

Keywords: Football; Match performance; Physical-tactical data; Ranking; Standard.

FIG. 1

Contextualised high-intensity distances covered by various positions in different Tiers. *Greater distance covered for ‘Move to Receive/Exploit Space’ than Tier A, B, and C (P < 0.01). #Greater distance covered for ‘Push up Pitch’ than Tier C (P < 0.05). ◇Greater distance covered for ‘Over/Underlap’ than Tier C (P < 0.05). ◆Greater distance covered for ‘Recovery Run’ than Tier B (P < 0.05). ΔGreater distance covered for ‘Run with Ball’ than Tier B and C (P < 0.05). ●Greater distance covered for ‘Run in Behind/Penetrate’ than Tier B and C (P < 0.01). ○Greater distance covered for ‘Break into Box’ than Tier C (P < 0.01). ☆Greater distance covered for ‘Covering’ than Tier A (P < 0.05). ★Greater distance covered for ‘Covering’ than Tier A and B (P < 0.05). The volume of ‘Interception’ and ‘Push up Pitch’ distances was relatively small; thus, they are invisible on the figure.

FIG. 2a

Comparison of additional options for in- (A, B and C) and out-of-possession (D and E) variables between Tiers. Symbols denote differences (P < 0.05). ΔMore actions performed than all other Tiers. *More actions performed than Tier D. #More actions performed than Tier C and D. ●More actions performed than Tier C. ◆More actions performed than Tier B and C.

FIG. 2b

Comparison of additional options for in- (A, B and C) and out-of-possession (D and E) variables between Tiers. Symbols denote differences (P < 0.05). ΔMore actions performed than all other Tiers. *More actions performed than Tier D. #More actions performed than Tier C and D. ●More actions performed than Tier C. ◆More actions performed than Tier B and C.

FIG. 2c

Comparison of additional options for in- (A, B and C) and out-of-possession (D and E) variables between Tiers. Symbols denote differences (P < 0.05). ΔMore actions performed than all other Tiers. *More actions performed than Tier D. #More actions performed than Tier C and D. ●More actions performed than Tier C. ◆More actions performed than Tier B and C.

FIG. 2d

Comparison of additional options for in- (A, B and C) and out-of-possession (D and E) variables between Tiers. Symbols denote differences (P < 0.05). ΔMore actions performed than all other Tiers. *More actions performed than Tier D. #More actions performed than Tier C and D. ●More actions performed than Tier C. ◆More actions performed than Tier B and C.

FIG. 2e

Comparison of additional options for in- (A, B and C) and out-of-possession (D and E) variables between Tiers. Symbols denote differences (P < 0.05). ΔMore actions performed than all other Tiers. *More actions performed than Tier D. #More actions performed than Tier C and D. ●More actions performed than Tier C. ◆More actions performed than Tier B and C.

FIG. 3a

Frequency of high-intensity running in relation to (A) in-possession and (B) out-of-possession categories with special reference to different zones (defensive, middle, and final third). Symbols denote differences (P < 0.05). ΔMore actions performed than Tiers. #More actions performed than Tier D. *More actions performed than Tier A. ●More actions performed than Tier C and D. ◆More actions performed than Tier C. ■More actions performed than Tier B and C.

FIG. 3b

Frequency of high-intensity running in relation to (A) in-possession and (B) out-of-possession categories with special reference to different zones (defensive, middle, and final third). Symbols denote differences (P < 0.05). ΔMore actions performed than Tiers. #More actions performed than Tier D. *More actions performed than Tier A. ●More actions performed than Tier C and D. ◆More actions performed than Tier C. ■More actions performed than Tier B and C.

FIG. 4

Correlation between physical-tactical (x axis) and technical (y axis) variables. Dotted lines indicate 95% confidence intervals.