The aim of this systematic review is to provide a base of knowledge from studies that have dealt with the description of collective behaviour in young footballers according to the level of competence associated to that age group, taking representative tasks from positional data as our starting point. Following the PRISMA statement a systematic revision was carried out on three meta-search engines (PubMed, Web of Science and SportDiscus). The following key words were used in the search: football, tactical behaviour, positional data and age-group, together with their equivalents. Of the 423 articles identified, 11 fulfilled the inclusion requirements. The main results suggest that: the variables made up of the joining of two points with a line (Width, Length and distance between dyads) and the collective area covered increase with age; however, the individual area tends to reduce. The increase in level of competence appears to require a greater functional variability in order to generate uncertainty and to counteract that of the opposing team. These results could allow trainers to identify on which tactical behaviour to focus intervention with the aim of fostering optimal development according the age.

The aim of this study was to describe the effects on collective behaviour and the physical response of young elite football players (Under 15) in the same 11 versus 11 task in three different pitch lengths (100, 75 and 50 m, LSG100, LSG75 and LSG50, respectively), keeping the width constant (60 m). The intra-team variables were: convex hull (CH), stretch index (SI), width (W), length (L) and length/width (L/W). The inter-team variables were calculated by using the 20 outfield players: CH2, W2, L2, L/W2 and DC (distance between centroids). The physical variables were: maximum speed (Vmax), number of accelerations (Acc) and decelerations (Dec) and total distance (TD). The results showed small (LSG100 > LSG75) and moderate–long differences (LSG100 > LSG50) for CH, SI and L. There was a moderate increase in DC and W2 (LSG50 > LSG75) and a moderate decrease in SI2, CH2, SI, L and CH (LSG75 > LSG50). The entropy was higher in SSG50 > LSG75 > LSG100. On a conditional level, the values showed differences between the three formats studied, showing more similarities between LSG50 and LSG100 than with LSG75. The main conclusion of the study was that the variation in the dimensions of the pitch caused a non-linear variation in the behavioural and conditional response in teams and players.