Wouter Frencken

Oscillations of centroid position and surface area of soccer teams in small-sided games

Abstract There is a need for a collective variable that captures the dynamics of team sports like soccer at match level. The centroid positions and surface areas of two soccer teams potentially describe the coordinated flow of attacking and defending in small-sided soccer games at team level. The aim of the present study was to identify an overall game pattern by establishing whether the proposed variables were linearly related between teams over the course of the game. In addition, we tried to identify patterns in the build-up of goals. A positive linear relation and a negative linear relation were hypothesized for the centroid positions and surface areas respectively. Finally, we hypothesized that deviations from these patterns are present in the build-up of goals. Ten young male elite soccer players (mean age 17.3, s=0.7) played three small-sided soccer games (4-a-side) of 8 minutes as part of their regular training routine. An innovative player tracking system, local position measurement (LPM), was used for obtaining player positions at 45 Hz per player. Pearson correlation coefficients were calculated to investigate the proposed linear relation of the key variables. Correlation coefficients indicate a strong positive linear relation during a whole game for the centroid position in all three games, with the strongest relation for the forward-backward direction (r>0.94). For 10 out of 19 goals a crossing of the centroids in this direction can be seen. No negative linear relation was found for surface area (−0.01 < r<0.07). From this study, we concluded that over the course of a whole small-sided game, the forward-backward motion of the centroids is most strongly linearly related. Furthermore, goals show a specific pattern in the forward-backward motion of the centroid. Therefore, surface area and particularly centroid position may provide a sound basis for a collective variable that captures the dynamics of attacking and defending in soccer at team level. Future research should develop these ideas further.

Length, width and centroid distance as measures of teams tactical performance in youth football

Abstract Small-sided games are commonly used in training and teaching contexts of football. However, few studies have focused on the tactical implications of this type of drills. The aim of this study is to identify how tactical collective behaviour varies with age in different small-sided game formats. We investigated the in-game field position in three different age groups of youth football players [under-9 (n=10; age = 8.5±0.53), under-11 (n=10; age = 10.4±0.52) and under-13 (n=10; age = 12.7±0.48)], participating in two different small-sided game conditions (GK + 3×3 + GK and GK + 4×4 + GK). A team variable was created based on the players’ length per width ratio (lpwratio), and a match variable was calculated as the distance between the centroid of the two teams. Results show that team variable values were influenced by the age of the players, as younger teams tend to present a higher value of lpwratio in their dispersion on the pitch. The variability of this variable also showed a decrease for teams with older players, suggesting a more consistent application of the width (stretching and creating space) and concentration (compressing into a confined area) principles of play and reflecting a higher level of collective tactical behaviour. Match variable showed a larger centroid distance for the older age groups in comparison with the younger players in the GK + 3×3 + GK, while all age groups demonstrated similar large centroid distances in the GK + 4×4 + GK game format. These results suggest that length and width ratio and centroid distance are useful measures of tactical performance in small-sided games in youth football.

Soccer-specific accuracy and validity of the local position measurement (LPM) system

Limited data is available on accuracy and validity of video-based, GPS and electronic tracking systems, particularly with reference to curved courses and short high intensity running activities. The main goal of this study was to assess soccer-specific accuracy and validity of the radio-frequency based local position measurement (LPM) system (1000Hz) for measuring distance and speed during walking and sprinting. Three males walked and sprinted 4 soccer-specific courses 10 times each. Distance and speed recorded by LPM were compared to actual distance and speed measured by measuring tape and timing gates. In addition, accuracy was assessed. The static accuracy (SD of the mean) is 1cm for devices put on the pitch and 2-3 cm when worn by participants. LPM underestimates actual distance (mean difference at most -1.6%). Coefficient of variation becomes larger at higher speed and increased turning angle. With regard to speed, validity correlations are high (range: 0.71-0.97). The LPM speed is significantly and systematically lower, although absolute and relative differences are small, between -0.1 km h⁻¹ (-1.3%) and -0.6 km h⁻¹ (-3.9%). The typical error of the estimate increases with increased speed, but does not increase with increased turning angle. Because the reported differences are small, we conclude that the LPM-system produces highly accurate position and speed data in static and dynamic conditions and is a valid tool for player tracking in soccer and ball team sports in general.

Size matters: Pitch dimensions constrain interactive team behaviour in soccer

Pitch size varies in official soccer matches and differently sized pitches are adopted for tactical purposes in small-sided training games. Since interactive team behaviour emerges under constraints, the authors evaluate the effect of pitch size (task) manipulations on interactive team behaviour in small-sided soccer games. Four 4-a-side (plus goalkeepers) small-sided games were played: a reference game (30×20 m), length manipulation (24×20 m), width manipulation (30×16 m), and a combination (24×16 m). Using position data (100Hz), three measures quantifying the teams’ interaction were calculated: longitudinal inter-team distance, lateral inter-team distance, and surface area difference. Means and standard deviations, correlations and coupling values were calculated. Running correlations were calculated over a 3-s window to evaluate interaction patterns. As expected, a shorter pitch results in smaller longitudinal inter-team distance, lateral inter-team distance decreased for narrow pitches, and smaller total playing area resulted in decreased surface area. Unanticipated, a crossover effect was present; length and width manipulations also triggered changes in lateral and longitudinal direction respectively. Inter-team distances and surface area difference differed significantly across conditions. Interaction patterns differed across conditions for all measures. So, highly tactically relevant, soccer teams seem to adapt their interactive behaviour according to pitch size in small-sided games.

Anthropometric Injury Risk Factors in Elite-standard Youth Soccer

The aim of the study was to investigate whether an increased risk of injury occurrence can be determined through frequent anthropometric measurements in elite-standard youth soccer players. Over the course of one season, we followed 101 male elite-standard youth soccer players between 11 and 19 years of age. Height and body mass were monitored at monthly measurement intervals and fat percentage was assessed every 3 months by use of the sum of skinfold method. Growth in height (cm), alternations in body mass index (kg/m(2)), fat percentage and fat-free mass index (kg/m(2)) were calculated. Injuries were recorded in accordance with the recommendations of the FIFA Consensus Model for Injury Registration. Odds ratio scores and 95% confidence intervals were calculated using binary logistic regression analyses. The following anthropometric injury risk factors were identified: ≥ 0.6 centimeter growth per month (p=0.03; OR=1.63; 95% CI: 1.06-2.52), ≥ 0.3 kg/m(2) increase of body mass index value per month (p=0.03; OR=1.61; 95% CI: 1.04-2.49) and low fat percentage; i. e., < 7% for players aged 11-16 and < 5% for players over 16 years (p=0.01; OR=1.81; 95% CI: 1.18-2.76). Individual monitoring of anthropometrics provides useful information to determine increased risk of injury occurrence in elite-standard youth soccer.

Examination of the external and internal load indicators' association with overuse injuries in professional soccer players

OBJECTIVES Research in professional soccer focusing on the relevance of external and internal load indicators for injury prevention is scarce. This study examined the relationship between load indicators and overuse injuries. DESIGN Prospective cohort study. METHODS Data were collected from 35 professional male soccer players over two seasons. Following load indicators were examined: total distance covered (TD), distance covered at high speed (THSR; >20kmh-1), number of accelerations (ACCeff; >1ms-2), number of decelerations (DECeff; <-1ms-2), and rating of perceived exertion (RPE) multiplied by duration. Cumulative 1-, 2-, 3-, 4-weekly loads and acute:chronic workload ratios (ACWR) were calculated and split into low, medium and high groups. Only overuse injuries were included in the analysis to focus on their specific relationship with the load indicators. Generalized estimating equations were applied to analyse the relationship between load indicators and overuse injuries in the subsequent week. RESULTS In total, 64 overuse injuries were registered. For cumulative loads, results indicated an increased injury risk for higher 2- to 4-weekly loads as indicated by TD, DECeff, and RPE multiplied by duration. For ACWR, a high ratio for THSR (>1.18) resulted in a higher injury risk. In contrast, a lower injury risk was found when comparing medium ratios for ACCeff (0.87-1.12), DECeff (0.86-1.12), and RPE x duration (0.85-1.12) to low ratios. CONCLUSIONS Findings demonstrate that mainly external load indicators are associated with increased or decreased injury risk. The monitoring of various load indicators is recommended for injury prevention in professional soccer.

Match-derived relative pitch area changes the physical and team tactical performance of elite soccer players in small-sided soccer games

ABSTRACT Small-sided games (SSGs) are used in training sessions to prepare for full-sized matches. For the same number of players, smaller pitch sizes result in decreased physical performance and shorter interpersonal distances. A relative pitch area derived from the full-sized match results in larger pitch sizes and this may increase the fit between SSGs and full-sized matches. This study aimed to investigate SSGs with a traditional small pitch and a match-derived relative pitch area in youth elite soccer players. Four age categories (under-13, under-15, under-17 and under-19) played 4 vs. 4 plus goalkeepers on a small (40x30m, 120m2 relative pitch area) and large pitch (68x47m, 320m2 relative pitch area). The number of games per age category ranged 15–30. Positional data (LPM-system) were collected to determine physical (total distance covered, high intensity distance and number of sprints) and team tactical (inter-team distance, LPW-ratio, surface area, stretch indices, goalkeeper-defender distance) performance measures and tactical variability. On a large pitch, physical performance significantly increased, inter-team and intra-team distances were significantly larger and tactical variability of intra-team distance measures significantly increased. The match-derived relative pitch area is an important training manipulation and leads to changes in physical and tactical performance 4 vs. 4 plus goalkeepers.

Using Deep Convolutional Neural Networks to Predict Goal-Scoring Opportunities in Soccer

Deep learning approaches have successfully been applied to several image recognition tasks, such as face, object, animal and plant classification. However, almost no research has examined on how to use the field of machine learning to predict goal-scoring opportunities in soccer from position data. In this paper, we propose the use of deep convolutional neural networks (DCNNs) for the above stated problem. This aim is actualized using the following steps: 1) development of novel algorithms for finding goal-scoring opportunities and ball possession which are used to obtain positive and negative examples. The dataset consists of position data from 29 matches played by a German Bundlesliga team. 2) These examples are used to create original and enhanced images (which contain object trails of soccer positions) with a resolution size of

Differences in game reading between selected and non-selected youth soccer players

256 \times 256

Comment on: "Talent Identification in Sport: A Systematic Review"

pixels. 3) Both the original and enhanced images are fed independently as input to two DCNN methods: instances of both GoogLeNet and a 3-layered CNN architecture. A K-nearest neighbor classifier was trained and evaluated on ball positions as a baseline experiment. The results show that the GoogLeNet architecture outperforms all other methods with an accuracy of 67.1%.