A comparison of match-physical demands between different tactical systems: 1-4-5-1 vs 1-3-5-2

The team tactical system and distribution of the football players on the pitch is considered fundamental in team performance. The present study used time-motion analysis and triaxial-accelerometers to obtain new insights about the impact of different tactical systems (1-4-5-1 and 1-3-5-2) on physical performance, across different playing positions, in a professional football team. Player performance data in fifteen official home matches was collected for analysis. The sample included twenty-two players from five playing positions (centre backs: n = 4; full-back/wide midfielder/ wing-back: n = 9; centre midfielder: n = 6 and centre forward: n = 3), making a total of 108 match observations. A novel finding was that general match physical demands do not differ considerably between these tactical formations, probably because match-to-match variability (variation of players’ running profile from match-to-match) might be higher than the differences in physical performance between tactical systems. However, change of formation had a different impact across playing positions, with centre backs playing in 1-4-5-1 performing significant more HIRcounts than in 1-3-5-2 (p = 0.031). Furthermore, a medium effect size (r = 0.33) was observed in HIRdist, with wide players covering higher distances when playing in 1-3-5-2 than in 1-4-5-1. These findings may help coaches to develop individualised training programs to meet the demands of each playing position according to the tactical system adopted.


Introduction
To better understand the constraints correlated with sporting success, match analysis has become an important tool in team sports. Nowadays it is well accepted among coaches and sport scientists that the match performance of a football team is, basically, based on four factors: physical, technical, tactical and mental [1]. Even though, the majority of research has been executed within the physical and technical performance domain, previous studies have a1111111111 a1111111111 a1111111111 a1111111111 a1111111111 The sample included 22 players (25.2 ± 4.4 years of age; 76.2 ± 6.4 kg of body mass; and, 181.6 ± 5.6 cm of height) across four different playing positions: centre back, CB (n = 4, observations[obs] = 37), full-back/wide midfielder/ wing-back, FB/WM/WB (n = 9, obs = 31), centre midfielder, CM (n = 6, obs = 26), and centre forward, CF (n = 3, obs = 14), making a total of 139 match observations (Table 1). Playing-positions were chosen according to the two tactical formations used by the team and previous research [9,21,22]. Team tactical systems and playing positions were determined by two UEFA-qualified coaches (one from the coaching staff of the team analysed) after visualizing video recordings of the sampled matches [17,20]. These observers subjectively determined the tactical systems used at the beginning of the match and verified if the formations were consistent throughout the matches [17]. Furthermore, 1-4-5-1 and 1-4-3-3 formations were combined, as well as 1-3-5-2 and 1-5-3-2. This procedure was applied due to difficulties in establishing specific differences between similar playing formations when in attacking and defending. When analysing the 1-3-5-2 formation the observers realized that the team often played in 1-5-3-2 formation when not in ball possession (defending) and in 1-3-5-2 with ball possession (attacking). On the other hand, when observing the 1-4-5-1 formation, the observers concluded that the team played in 1-4-5-1 when defending and in 1-4-3-3 when attacking [11,17]. No other changes in formations throughout the matches were noticed by the observers, therefor no matches were excluded from the analysis.
To ensure players confidentiality, all data was anonymized before analyses.

Procedures
A stationary radio wave-based Local Positioning Measurement (LPM) tracking system (ZXY Sport Tracking System, Trondheim, Norway), with a default resolution of 20Hz, was used to characterize match activity profiles within the team. Each player wore a specially designed belt, wrapped tightly around the waist, with an electronic sensor system at the player's lumbar spine, as reported previously [10]. At the stadium, where the matches occurred, there are 6 RadioEyes for optimal coverage, resulting in practically zero packet loss for transponders on the field. If packet loss occurred, the data was linearly interpolated. The accuracy and reliability of the system in measuring player movements in elite soccer competitions have been described in more detail in previous studies [23][24][25].
According to the ZXY Sport Tracking system accelerations were quantified through numerical derivation from positional data with a sampling frequency of 20Hz [25]. Furthermore, accelerations are defined by four event markers: (a) the start of the acceleration event is marked by the acceleration reaching the minimum limit of 1 m�s −2 , (b) the acceleration reaches the acceleration limit of 2 m�s −2 , (c) the acceleration remains above the 2 m�s −2 for at least 0.5 seconds and (d) the duration of the acceleration ends when it decreases below the minimum acceleration limit (1 m�s −2 ).
Turns were counted only if the player performed a continuous and significant body rotation of more than 90˚in one direction (derived from gyroscope and compass data). The end of a turn and the start of another occurs when a rotation in the opposite direction is measured. The angle threshold used by ZXY Sport Tracking system allowed us to analyse only angles �90˚.

Statistical analysis
The results are presented as mean and 95% confidence interval, unless otherwise stated. A linear mixed-effects model with restricted maximum likelihood estimations was used to examine differences in Local Positioning Measurement-derived variables and match duration between 1-3-5-2 and 1-4-5-1 formations. Mixed models can account for unbalanced repeats per player and thus used to model the data. Tactical formation, playing position and their interaction was modelled as fixed effects (effects describing the association between the dependent variable and covariates), while 'athlete ID' was included as a random effect (effects generally representing random deviations from the relationships of the fixed part of the model). An α-level of 0.05 was used as level of significance for statistical comparisons. Furthermore, multiple comparisons were adjusted using the Tukey method. The t statistics from the mixed models were converted to effect size correlations [27]. Effect sizes were interpreted as <0.1, trivial; 0.1-0.3, small; 0.3-0.5, moderate; 0.5-0.7, large; 0.7-0.9, very large; 0.9-0.99, almost perfect; 1.0, perfect [28]. All statistical analyses were conducted using the lme4, lsmeans and psychometric packages in R statistical software (version 3.4.1, R Foundation for Statistical Computing, Vienna, Austria).

Centre forwards
No significant differences were found regarding any parameter analysed. However, higher values, though with a trivial effect size, in HIR dist and sprint dist can be observed (Table 5) when playing in 1-3-5-2.

Tactical system
Significant differences were found in various parameters when comparing the physical performance of the whole team when playing with different tactical systems (Table 6). Significant higher values were observed in HIR counts (r = 0.25) and sprint counts (r = 0.22) when playing in Table 2. Mean and 95% confidence interval estimates of different physical parameters from centre backs, analysed according to the tactical system used, and respective p-value and effect size of differences observed (n = 4; observations = 37).

Variables
CB p-value Effect Size (r)  Table 3. Mean and 95% confidence interval estimates of different physical parameters from full-backs, wide midfielders and wing-backs analysed according to the tactical system used, and respective p-value and effect size of differences observed (n = 9; observations = 31).

Context
The present study provides new insights into the physical demands of two common tactical formations, in elite football players across different playing positions. The context of this study appeared with the change of the head-coach, and consequently, the tactical formation and style of play used of the professional football team analysed. Since this replacement happened in the middle of the season, both tactical formations analysed were composed by an almost equal number of matches (7 and 8 home matches each). It is also important to refer that the change of head-coach led not only to a simple switch of the tactical structure used, but also to a change to a more complex style of play. A more possession and position-oriented style of play Table 4. Mean and 95% confidence interval estimates of different physical parameters from centre midfielders, analysed according to the tactical system used, and respective p-value and effect size of differences observed (n = 6; observations = 26).

1-4-5-1 1-3-5-2
TotDist 12009.0 (218.5) 11820. 8 Table 5. Mean and 95% confidence interval estimates of different physical parameters from centre forwards, analysed according to the tactical system used, and respective p-value and effect size of differences observed (n = 3; observations = 14). were adopted (1-3-5-2) instead of the more direct play and counter-attack strategy used in the first half of the season (1-4-5-1). However, even with all these changes, the context remained the same (same players with similar physical capacities).

Comparison according to playing position
The results suggest that general match physical demands do not differ considerably between these two tactical formations when compared by playing position. Independent of formation and with few exceptions, players presented similar profiles in all the physical parameters analysed. The most relevant exceptions were the higher HIR counts in CB (1-4-5-1) and longer HIR dist in FB/WM/WB (1-3-5-2), with a medium and small effect size, respectively. CB playing in 1-4-5-1 performed more HIR counts , probably due to the larger area they needed to cover when compared to the area covered by the three CBs when playing in 1-3-5-2. When in defensive organisation (without ball possession), the defensive line of three CBs became most of the time a defensive line composed by 5 players (three CBs and two WBs). The increased number of players playing in the defensive line leads to less m 2 per player to cover.
Players in wide positions covered more HIR dist when playing in 1-3-5-2 most likely because in this formation the team played with only two wide players (WB), and they needed to cover all the flank, while with 1-4-5-1 formation, those flanks were covered by a total of four players (two on each side).
It has been speculated that match physical demands are higher for CF when playing "alone" in the offensive line (e.g. 1-4-5-1; 1-5-4-1), as they are very often isolated and marked by several opponents [29]. However, the results of the present study are slightly different, since higher, though not significant, values were found in HIR dist and sprint dist for CF, when playing with two attackers (1-3-5-2) compared with playing with only one (1-4-5-1).
Furthermore, no differences in playing time (substitutions) were observed in any playing position between the two tactical systems analysed.

Comparison according to team workload
When playing position was not taken into consideration and the work-load of the whole team was analysed, the physical workload in some variables was significantly different between tactical systems used. Small significant differences were observed in HIR counts and sprint counts , with the team performing more runs (>19,8 km/h) when playing in 1-4-5-1. The number of Acc Table 6. Mean and 95% confidence interval estimates of different physical parameters from the whole team, analysed according to the tactical system used, and respective p-value and effect size of differences observed.

Variables
Tactical system p-value Effect Size (r) and Dec was also higher when the 1-4-5-1 system was used. In general, almost all variables analysed presented higher values during the first period of the season (1-4-5-1) than in the second (1-3-5-2). Previous research [30,31] has suggested that teams who are winning the match tend to relax and decrease their work-rate. Alternatively, although teams who are losing the match may increase their work-rate during a specified period [32,33], they may quickly lose the motivation to keep the elevated work rate, which may be especially evident when the goal difference increases negatively (conceding more goals) [34]. In fact, the differences observed between these two tactical systems might be, in part, justified by the significant discrepancy between the score line and match final results achieved during the first and second part of the season. While playing in 1-4-5-1 the team achieved one victory, four draws and three defeats in the eight home matches played. On the other hand, while playing in 1-3-5-2, the team had better results, with five victories, one draw and one defeat in the last seven home matches played. The match results (considerably more draws) and the differences in style of play may therefore, partly justify the higher work-rate of the 1-4-5-1 tactical system.

Limitations
Our initial hypothesis was that, despite playing in their specific positions, players would accumulate different external workload in matches, depending on the preferred tactical formation. However, the results presented in this study do not fully support the hypothesis, probably because the match-to-match variability might be larger than the differences in physical performance between tactical systems. Like most of the measures in team sports performance, the physical variables used in this study are not stable and are subject to a high variation between successive matches [35]. Furthermore, it has been proved that within-subject (player) and between-match variation in physical performance across the season might be experienced due to changes in the physical condition of the player [36,37] and environmental conditions [38]. Previous studies have shown that match-to-match variability in performance characteristics of elite soccer players is high [35,39,40] and that future research based in match performance requires large sample sizes to identify true systematic changes in workload. In fact, the sample size (22 players/108 observations) might be of such small numbers that true differences can be masked due to a statistical type 2 error, and such a consequence cannot be conclusively ruled out. Previous similar studies have analysed more matches [17] or used considerably larger sample sizes [11] than in the present study. However, they have not compared the physical demands of different tactical systems within the same players in the same context (same team and season) and to do so, a larger sample size than the one used in the present study becomes a difficult task to fulfil.
Even though, the methodology used to determine the team formations is in line with previous studies [11,14,17,20,41,42], the process of defining team formations and controlling their consistency throughout the matches was based on the subjective assessment of observers. Further research is needed to attempt to define objectively team formations and to identify when changes occur [17].
Goalkeepers were not included in the present study, however their match activity profiles might be useful and interesting to analyse in different tactical systems and styles of play in future research. All these limitations should be taken into consideration when designing future studies.

Perspectives and practical application
Since previous research has shown that the players' physical demands in matches are highly dependent on their positional role in the team [43,44], analytics, in general, have become a crucial component of team organization and content of training, to meet the position-specific requirements of physical conditioning [45]. This study goes beyond the individualization of training demands according to playing position, also suggesting that the change of tactical system might influence, specific variables of the team's overall match activity profile, and those differences should be taken into consideration when designing training programs. On the other hand, differences are not notable in all playing positions and these findings should be interpreted with caution, as differences might be team dependent since other teams using the same tactical systems, probably appear with different styles of play.
Change of formation had a different impact on different playing positions, with CB and wide positions presenting more substantial differences than CM and CF. As previously mentioned, the present study and its findings may provide useful and novel insights for coaches on physical performance demands in different tactical formations across playing positions. The information provided should be taken into consideration when designing and implementing training program cycles, according to players' playing position, the team's tactical formation and style of play. The individualization and specialization of the training should, therefore, be a matter of reflection and analysis from practitioners.