AI, BlazePod Sensors, and Head Vests Implemented in Assessments on Reaction Time and Gaze Training Program in U10 Football Game

Abstract

In the context of the development of technologies, every sports club tends to improve its training methods to obtain the best possible results in sports training. The goal of the research is to develop a specialized training program designed to enhance ball-control skills so that children can play soccer with increased confidence, therefore reinforcing their need for constant visual contact with the ball during possession. The study participants are children between the ages of 8 and 10, who have acquired at least one year of consistent and well-structured football practice, divided into two groups, experimental group I and control group II. The T-Blaze test training, the Adams test, and the registration of the degree of head tilt using artificial intelligence and visual recognition were implemented. During the training, the authors used the BlazePod sensors to measure participants’ times more precisely, thus avoiding the inaccuracy of using a classic timer. At the same time, the authors used the Vesta HeadUp to block the child’s view of the ball when he has possession of the ball or is very close to it. The recording of time spent playing head-up and head-down revealed statistically significant differences between the three test sessions in favor of the experimental group. Considering the statistically substantial influence obtained, the authors can conclude that our intervention program based on specific means and using HeadUp vests was a decisive factor in achieving improved performance.

1. Introduction

Presently, the development of motor potential and specificity in motor qualities in physical education and sports are very intensely discussed themes. In today’s football, it is no longer enough for the player to possess a brilliant technique if it cannot be applied under the new conditions imposed by the football game (speed, force, permanent movement, long distances, increased aggression, etc.) [1].

From this perspective, literature exploration was the starting point in our initial research. Learning can overlap spontaneous and native reactivity, becoming, through repetition or consolidation, an important acquisition in sports games. Moreover, the integration of psychomotor elements in the learning process and the assessment stage, while paying special attention to young ages, provides young football players with decisive guidance and a proper context to cultivate specific skills. These skills will form the basis for effective performance during training and competition [2].

The authors identify a gap in the literature in the fact that the focus is on physical, theoretical, and tactical training, but a strategic element is to have the entire perspective of the field. To ensure an exhaustive presentation of the issue under discussion, the authors carried out research to identify the most significant and updated bibliographic sources to reflect the current status of the topic addressed. The authors highlight as elements of originality in the study the introduction of HeadUp vests for training, the implementation of advanced visual recognition software and artificial intelligence technology, and the use of advanced technology offered by BlazePod and sensors.

The fundamental concept behind this study’s creation stems from the underutilization of the extraordinary potential that 8 to 10-year-old youngsters possess when given the right instruction. The authors felt it was imperative to approach the learning process through sports workouts because of this age range. Thus, our study highlights the scientific status of teaching methodology and educational standards for motor quality and agility. The study aims to establish and assess the presence of Wellness Culture in primary education by applying an innovative methodology for developing motor quality agility and increasing motor capacity. The methodology includes an author’s set of motor exercises related to the BlazePod Trainer, as other authors have done before [1].

Our purpose is to develop a specialized training program designed to improve ball-control skills so that children can play football with increased confidence, thus reducing their need for constant visual contact with the ball during possession. The authors must mention from the beginning that this is preliminary research as the authors have not found our method applied to a large-scale population of 8–10-year-olds.

It is hypothesized that by decreasing the degree of inclination of the head during the football game, an increase in the number of successful passes and a decrease in the number of wrong passes during the soccer game can be achieved.

2. Literature Review

Young people learn cognitively, acquire behavioral, intrapersonal, and interpersonal skills through sports, and put themselves in an excellent position to be better athletes [3]. In this regard, the development of athletes, and therefore the development of future elite athletes, requires a continuous learning process. It can be said that all the structures available to a living system at a given time and which are not innate represent the result of learning. Learning is also closely related to both the level of individual attention and the individual ability to solve different tasks [4]. Concentration disruption is related to undesirable variables for athletes in general; for young athletes, it is related to unpleasant emotions such as depression, anxiety, and intrusive thoughts [5]. Thus, concentration disruption could be considered a maladaptive cognitive ability with a negative influence on the development of an athlete. In each sports discipline, whether it is a team or individual one, young athletes face various mental and emotional challenges due to the structure and the way of playing the sport concerned. Mental and social skills are particularly important to the development of young athletes. Mental qualities have a significant impact on the athlete’s progress, and sports psychology researchers have analyzed and highlighted some of the most important mental skills and traits that young athletes can cultivate, including concentration [6]. Research on the relationship between cognitive functions and motor skills specific to football players has reported that attention, perception, and the ability to track several objects at once are positively associated with sprinting, ball control, dribbling, and changes in direction [7]. Educating the components of coordination skills at the right time is decisive for the development of the future football player, considering the influence of these skills on performance factors [8]. There is a relationship between coordination and the ability to kick the ball, between balance and the ability to kick the ball [9,10,11,12,13,14].

Comprehension of the game, which includes identifying available areas, managing movements on the field, and identifying areas where a player can advance with the ball, can only be achieved by training children to keep their gaze up and improve their attention during play concerning the improvement of their ability to analyze different situations arising during the game, under conditions of space and time crisis. Keeping one’s gaze up requires well-developed coordination ability and automatically a good level of individual technique or ball sense. It might be imperative to create coordination tests that could be related to sports performance in general, and very specific coordination tests for each sport [15].

The demand for video analysis has rapidly increased over the past decade. Video analysis has started playing an essential role in sports, being also present during a game through the collection and use of real-time data. This information is helpful to the coach, referees, players, spectators, and physicians, but most of it aims to detect and track players and the ball [16]. Competencies associated with visual attention can provide athletes with the ability to monitor multiple stimuli simultaneously, to focus attention on significant elements of a given situation, or to ignore certain stimuli in favor of others. In the context of football, for example, visual attention can be used to accurately observe the movements of teammates and opponents in attacking and defending situations, thus facilitating appropriate decision-making [17].

Decision-making is the human brain’s ability to select important contextual data from the visual field and is essential for high-level athletic performance [18]. A study that considered, in addition to coordination skills, the cognitive side, specifically attention and planning, demonstrated that the football group showed significantly greater post-test gains than the sedentary group in terms of agility, visuospatial working memory, attention, planning, and inhibition. Thus, the authors can say that the results highlight the need to think about the organization of planned sports activities as a natural and satisfying way to enhance cognitive skills [19]. A study related to the part of our assessment confirms that including cognitive training in the game of football leads to an improvement in players’ ability to anticipate, inform, and evaluate existing game situations [20]. Primary school students who played sports had higher levels of attention compared to those who did not play sports. Girls scored higher than boys on psychomotor speed and selective attention. Therefore, it is suggested to involve all primary school students in sports activities to improve their attention levels [21]. Focusing attention allows players to use complex means and techniques adapted to the specific game situations that arise every second [22].

The study of attention and coordination aspects in sports activity needs an interdisciplinary approach [23,24,25,26] to highlight relevant and determining factors for different age groups [27,28,29,30] regarding sports experience [31,32,33,34] and age characteristics [34,35,36,37,38]. The optimization of sports training in children’s football depends on the constant improvement of players’ psychological and physical characteristics. Several studies have revealed the major impact of implementing exercises that require a good ability to focus attention concerning increasing the coordination ability of players [39]. Our study is in line with previous studies and contributes to expanding knowledge about the importance of specialized training in the game of football for children, with an emphasis on improving attention and coordination.

Distributed consistency is a crucial aspect of distributed systems, but it can cause undesirable performance if not used judiciously. Blazes, a cross-platform program analysis framework, identifies program locations requiring coordination and automatically synthesizes application-specific coordination code. It outperforms general-purpose techniques and presents two case studies using annotated programs in the Twitter Storm system and the Bloom declarative language [3]. Blaze is a reliable test method for athletes, especially in soccer, suitable for individuals under 18 and those over. It offers simple, selective, and discriminating reaction times for performance monitoring. The BlazePod system, a wireless lighting system with eight LEDs and a central PDA controller, was used to record hand and leg movement response times in complex tasks. The FitLight^TM and BlazePod XLiGHT light sport training devices are moisture and dust-protected, with rechargeable batteries and standby mode capabilities [40,41]. The Flash Reflex (FRX) Training methodology utilizes BlazePod’s Pods and app to enhance users’ reactive intelligence, aiming to improve cognitive and physical abilities. Reactive Intelligence is a cognitive agility training method that focuses on enhancing physical and cognitive abilities such as reaction time, decision-making, and active thinking. The FRX Training methodology uses reflexes, which are the fastest physical movements in the human body, to speed up brain processing and react as quickly as a reflex. The Pods and app are essential tools for training with the FRX methodology, as they help enhance reaction time, the time between recognition and reaction. By improving reaction time, individuals can enhance their performance and control every scenario [3].

As they are also interested in technological development as presented in the literature, the authors turned our attention to different methods of recording the head tilt angle during play. To perform game analysis, the authors used the OpenPifPaf v0.13.11 project to collect data and the Python^TM programming language to process these data and calculate the gaze angle. A machine-learning artificial intelligence model specialized in analyzing and collecting data on a person’s posture in different poses was also used. Following a review of the last 50 years, there has been considerable development of sports motion analysis systems in terms of computer-motion modeling. Understanding the mechanics of sports techniques has become a profound challenge in many sports, with the musculoskeletal models being increasingly detailed. If it was quite difficult in the past to provide data acquisition equipment, presently, this factor is less limiting, but the fundamental understanding of how data are processed is still imperative [10]. Using the OpenPifPaf pose estimation technique, the authors performed key-point-based movement analysis by measuring reference and capacitive ECG synced with a video recording [42,43].

OpenPifPaf is based on deep neural network architecture and uses supervised learning techniques to learn to locate and track human joints. The neural network uses multi-class learning techniques to classify and regress joint positions in an image in parallel. An interesting aspect of OpenPifPaf is that it can detect and analyze human joints, even in the case of partially visible objects or in complex joint overlapping situations. This makes it suitable for applications such as real-time posture tracking, gesture analysis, and human activity recognition. OpenPifPaf is available as open source and provides detailed documentation, usage examples, and source code to help researchers and developers use it in their projects. It is an active project that continues to be developed and improved by its community [12].

Having in mind this context the research’s main goal is to show how well a modified training program, which includes an exercise that combines reaction speed work with a decreased visibility job to enhance passing, can increase passing. During the training program, the authors decided to exercise with Heads Vests to train the athletes to play more time with the gaze upwards. To evaluate the impact of the training based on the Cover’s method the authors used T-BlazePod (BlazePod Israel Ltd., Tel Aviv, Israel) to avoid the inaccuracy of the timer and to add technology to the classic test (t-test). The authors also choose to implement the Adam test for coordinative performance. For data collection and game performance analysis, the authors implemented OpenPifPaf facilities.

Given the research objective, the main hypothesis of the study, H1, arises: Training the children to play football with their heads up for a longer time will improve the number of passes that are successful and decrease the number of passes that are incorrect during a soccer match.

H1. 

There are significant statistical differences between the Experimental group/Control group regarding results in the Adams test.

H2. 

There are significant statistical differences between the Experimental group/Control group regarding Time spent playing with the head up.

H3. 

There are significant statistical differences between the Experimental group/Control group regarding Time spent playing with the head down.

H4. 

There are significant statistical differences between the Experimental group/Control group regarding the Number of successful passes.

3. Materials and Methods

3.1. Study Design

The study is longitudinal research that took place for a long period, between May 2021 and May 2023 to be able to evaluate the sportive improvements. The study involved the use of BlazePod technology and sensors in standardized tests to assess coordination skills, meanwhile, the Adams test assesses coordinative abilities [43]. BlazePod™ is a Bluetooth low-energy technology to assess response times in simple and choice tasks. The device, which uses LED lights, has been proven reliable in single-leg striking and agility tasks. The noise interference was active construction site noise, with noise-cancellation headphones provided. Body weight acute exercises were used to induce localized muscular workload [18,44,45]. In addition, in the context of this investigation, the authors introduced the assessment of the degree of head tilt during the game after applying the Adams test™, using technology based on artificial intelligence and visual recognition. The number of successful passes was analyzed during the game by video analysis. The order of performing the tests is the one previously specified.

The statistical analysis was based on the Levene test to evaluate the homogeneity of our variable. Then, the authors applied inference tests to evaluate the Adams results, like Student (S) for homogeneous variables and Mann–Whitney U (MW) for non-homogeneous variables.

3.2. Participants

In the proposed study, the group of participants consists of members of Chelsea Football Club Bucharest, currently called AFC Blue Lions. The informed consent was obtained from their parents, and data confidentiality was ensured. The research has been carried out in accordance with the Declaration of Helsinki of the World Medical Association, revised in 2013 for experiments involving humans, as well as in accordance with the EU Directive 2010/63/EU for animal experiments. The ethics commission of the National University of Physical Education and Sport gave us consent to develop this study. This research group consists of 42 boys aged between 8 and 10 years. The criterion for inclusion in the studio was the acquisition of at least one year of consistent and well-structured experience in training and the official football championship. These children actively participated in the matches of the Municipal Championship and participated in various tournaments played throughout the country. The experimental group I included 21 boys, and the control group II included 21 boys. The inclusion of athletes in different groups was undertaken randomly by drawing lots for the name of the group (experimental/control). The group performed the 100 and HeadUp training program, and the control group performed the classic training. The control group has the training between 17:00 and 18:30, and the experimental group between 18:30 and 20:00.

3.3. Experimental Program

T-Blaze Training

Description: Four BlazePods are placed in the shape of the letter T—A, B, C, D. The participant starts from BlazePod A. At the command of the software, the athlete sprints to BlazePod B and touches its base with his right hand. Then, he runs sideways to the left with added steps and moves to BlazePod C, touching its base with his left hand. The athlete returns using the same run to BlazePod B and touches its base again with his right hand. Then, he runs sideways to the right with added steps and touches BlazePod D, returns and touches BlazePod B with his left hand. Finally, the athlete runs backward to BlazePod A, the route ending when he touches it.

The training program includes 100 exercises divided into 6 categories, but also a training tool called the HeadUp vest, which is designed to help children play the game of football with their gaze upwards.

The HeadUp vest, a wearable device, is being combined with AI technology to enhance player performance and safety in football. The vest monitors the physiological and biomechanical parameters of players in real time, providing actionable insights. AI algorithms can detect patterns, predict injuries, and offer personalized training recommendations. Real-time data analysis allows coaches to make immediate decisions during training sessions, potentially preventing injuries. AI can create tailored training programs based on individual needs and performance metrics, ensuring effectiveness and relevance. It can also predict injury risk factors and provide rehabilitation support for injured players. AI can also enhance tactical and strategic decisions by analyzing team performance and opponents’ strengths and weaknesses. The holistic approach to player wellbeing includes considering psychological and emotional factors and managing load. This innovative approach has the potential to revolutionize football training, competition, and recovery, making the sport safer and more efficient. These issues confirm that this method addresses the shortcomings of previous studies, which emphasizes the importance of this research even more.

The HeadUp vest is used during training to block the child’s gaze toward the ball when in possession or very close to the ball. The sense of the ball is an important component in the process of starting to use the HeadUp vest, which requires the child to have a developed ball sense and optimal intersegmental coordination ability. The vest is risk-free for the child who wears it or those who participate in the game/exercises; on the contrary, being made of sponge, it protects them against accidents, is comfortable, and is very light (0.33 kg).

The shape of the vest forces the player to lift their gaze from the ground and adapt to the new conditions. After wearing the vests during training, the appearance of the reflex to look up from the ground, the emergence of clairvoyance during gameplay, and the development of peripheral vision are pursued.

The authors present the periodization of the means used throughout the implementation of the action system. The sequencing of the means is shown in

Table A1. Scheduling of the 100 exercises included in the action system.

Types of Exercises	A	B	C	D	E	F
	Standing, without a Ball	Running, without a Ball	Standing, with a Ball	Running, with a Ball, without Adversity	Running, with a Ball, with a Semi-Active Opponent	Game Situations
No. of Selected Exercises	9 Exercises	24 Exercises	24 Exercises	24 Exercises	9 Exercises	10 Exercises
MeC. 2—MiC. 7	1, 2	1, 2, 3	1, 2, 3	1, 2, 3	1	1
MeC. 2—MiC. 8	1, 2	1, 2, 3	1, 2, 3	1, 2, 3	1	1
MeC. 3—MiC. 9	2, 3	4, 5, 6	3, 4, 5	4, 5, 6	1, 2	1, 2
MeC. 3—MiC. 10	2, 3	4, 5, 6	3, 4, 5	4, 5, 6	1, 2	1, 2
MeC. 3—MiC. 11	3, 4	5, 6	5, 6, 7	7, 8	6, 7	5
MeC. 3—MiC. 12	3, 4	5, 6	5, 6, 7	7, 8	6, 7	5
MiR—December 2021	1, 2, 3	1, 2	1, 2	24	-	1
MiA—January 2022	1, 2, 3, 4	5, 6	1, 2, 3	1, 2, 3	1	2
MeB. 2—MiB. 1	1, 2, 4, 5	7, 8, 9, 10	8, 9, 10	22, 23, 24	6, 7	1, 2, 5
MeB. 2—MiB. 2	1, 2, 4, 5	7, 8, 9, 10	8, 9, 10	22, 23, 24	6, 7	1, 2, 5
MeB. 2—MiB. 3	3, 4, 5	9, 10	10, 11, 12	21, 23	5, 6	3, 4, 5
MeB. 2—MiB. 4	3, 4, 5	9, 10	10, 11, 12	21, 23	5, 6	3, 4, 5
MePC. 2—MiS. 2	1, 2, 5, 6, 7	11, 12, 13, 14	10, 11, 12, 13, 14	7, 8, 9	3, 4	9, 10
MePC. 2—MiS. 3	1, 2, 5, 6, 7	11, 12, 13, 14	10, 11, 12, 13, 14	7, 8, 9	3, 4	9, 10
MePC. 2—MiS. 4	6, 7, 8	12, 13, 14	14, 15, 16	11, 12, 14, 15	8	5, 10
MeC. 4—MiC. 1	8, 9	14, 15, 16	16, 17, 18, 19	11, 12, 14, 15	6, 8, 9	6, 7, 10
MeC. 4—MiC. 2	7, 8, 9	14, 15, 16	18, 19, 20, 21	14, 15, 16, 17	8, 9	8, 9, 10
MeC. 4—MiC. 3	3, 4, 6, 7, 8	17, 18, 19, 20	19, 20, 21	18, 19, 20	8, 9	5, 8, 9
MeC. 4—MiC. 4	1, 2, 5, 9	17, 18, 19, 20	22, 23, 24	17, 18, 19, 20	6, 7, 8, 9	5, 6, 7
MeC. 5—MiC. 5	4, 7, 8, 9	15, 18, 20	19, 21, 23, 24	15, 18, 20, 21	6, 7, 8, 9	5, 10
MeC. 5—MiC. 6	4, 7, 8, 9	15, 18, 20	19, 21, 23, 24	15, 18, 20, 21	6, 7, 8, 9	5, 10
MeC. 5—MiC. 7	3, 5, 6	9, 10, 19, 20	15, 17, 20, 23, 24	20, 21, 22, 23, 24	6, 7, 8, 9	1, 2, 3, 7
MeC. 5—MiC. 8	3, 5, 6	9, 10, 19, 20	15, 17, 20, 23, 24	20, 21, 22, 23, 24	6, 7, 8, 9	1, 2, 3, 7
MeC. 6—MiC. 9	6, 7, 8, 9	21, 22, 23, 24	10, 11, 12, 13, 22, 23, 24	20, 21, 22, 23, 24	6, 7, 8, 9	4, 5, 6, 7
MeC. 6—MiC. 10	6, 7, 8, 9	21, 22, 23, 24	10, 11, 12, 13, 22, 23, 24	20, 21, 22, 23, 24	6, 7, 8, 9	4, 5, 6, 7
MeC. 6—MiC. 11	1, 2, 8, 9	21, 22, 23, 24	5, 6, 7, 8, 9, 10	1, 2, 3, 4, 5, 6, 7	1, 2, 3, 4	1, 2, 3, 8, 9
MeC. 6—MiC. 12	1, 2, 8, 9	21, 22, 23, 24	5, 6, 7, 8, 9, 10	1, 2, 3, 4, 5, 6, 7	1, 2, 3, 4	1, 2, 3, 8, 9
MiR.—June 2022	8, 9	1, 2, 3	22, 23, 24	7, 8, 9, 10	3	10
MiR.—June 2022	9	9, 10, 11	1, 2, 3, 16, 17	17, 18, 19	4	10
MiR.—July 2022	9	13, 14, 15	13, 14,	22, 23	2	10

Legend: MeC. = Competitive Mesocycle; MiC. = Competitive Microcycle; MiR. = Recovery Microcycle; MiA. = Adaptation Microcycle; MeB. = Basic Mesocycle; MiB. = Basic Microcycle; MePC. = Pre-competitive Mesocycle; MiS. = Specific Microcycle.

. The exercises were divided into several categories as follows:

• Type A exercises (9 exercises)—standing coordination exercises—without a ball;
• Type B exercises (24 exercises)—running coordination exercises—without a ball;
• Type C exercises (24 exercises)—standing coordination exercises—with a ball;
• Type D exercises (24 exercises)—running coordination exercises—without adversity;
• Type E exercises (9 exercises)—running exercises with a ball—with a semi-active opponent;
• Type F exercises (10 exercises)—themed games (1 vs. 1/1 vs. 2/2 vs. 2, etc.).

All this information is presented in Table A1 from Appendix A.

These exercises are specific to Coerver Coaching. The Coerver method involves progress in the football player’s technique in a structured, pyramidal way, starting from the basics of ball control to a tactically driven group attack. Wiel Coerver developed a training technique called the Coerver method by analyzing the videos of great players, including Pele, managing to develop a new concept in football that claims that the skill specific to the game of football could be transmitted in a complete academic way.

The Coerver Pyramid consists of 6 levels, which are closely related to each other, namely:

• The foundation, control/touch, and trust that affects every other part of the pyramid.
• Reception and passing.
• Individual skills, to keep possession and create space and time for dribbling, passing, or finishing.
• Mental and physical speed, with and without the ball.
• Synchronization, concentration, and responsibility skills.
• Combined play, small group defense, fast play, teamwork.

To these exercises, the authors have also added the HeadUp vests to achieve our goals.

In the training microcycles, these 100 exercises were introduced gradually, from easy to hard, from simple to complex, and from known to unknown, using the HeadUp vests.

For the experimental group, the training that included the 100 specific exercises plus the HeadUp vests was used. For the control group, these exercises or the HeadUp vests were not used. Thus, the authors were able to observe if our intervention program managed to help us achieve the proposed objectives/confirm the hypotheses/achieve the goal.

The authors have followed the development of the coordination capacity without the ball in place » with the ball in place without adversity » with the ball away from home with a semi-active opponent.

3.4. Assessment Instruments

To achieve the presented goal, the authors used several assessments, measurements, and training methods related to the game of football, the level of the research participants, and the available material conditions.

In a nutshell, the authors have done 3 types of tests: Test 1—Adams, Test 2—T-Blaze test training, Test 3—The test for the evaluation of participants with the help of artificial intelligence, measuring time spent playing with the head up/down. The authors also counted the number of successful passes. Each test was repeated three times.

For the motor assessment, i.e., the measurement of coordination skills, the authors chose a specific test: the Adams test.

Adams Test™

Description: Crossing lines with a length of 1 m are drawn, and the four spaces created are numbered as follows: 2 and 4 at the top and 1 and 3 at the bottom. The athlete starts from space 1 and jumps on two feet in order of the numbers to perform as many jumps as possible within 15 s. Touching the line or not touching the ground with both feet is considered a mistake.

To record the head tilt angle during gameplay, the authors used technology based on artificial intelligence and visual recognition and set the following benchmarks to define exactly how the authors would relate to the head tilt angle during the game:

• Gaze upwards = 80–100%;
• Gaze downwards ≤ 80%.

To perform game analysis, the authors used the OpenPifPaf™ project to collect data and the Python programming language to process these data and calculate the gaze angle. A machine-learning artificial intelligence model specialized in analyzing and collecting data on a person’s posture in different poses was also used.

In research aimed at explaining the methods used by OpenPifPaf™ and their effectiveness, it has been found that OpenPifPaf is more efficient in terms of processing speed and accuracy than any other currently existing method, such as CrowdPose or PoseTrack2018. Increased accuracy was mainly observed in the analysis of pictures/videos with larger numbers of people, which makes it suitable for use in the analysis of a football game. OpenPifPaf is an open-source framework for the detection and analysis of human joints in images and videos. The main goal of the project is to achieve accurate detection of joint positions in an image frame and thus reconstruct human posture [12].

The software is represented by a machine-learning artificial intelligence model that has been trained to detect various data about a person’s posture, using different types of skeletons with different key points in the human body to obtain data about that person’s position in a picture/video. In this position, the player on the left side has his head almost perfectly up (gaze upwards represents approximately 80–100%), the player in the middle keeps his gaze slightly downwards, and the player on the right side, who is in possession, keeps his gaze downwards (34%) looking at the ball. The model is not based on a video image, which is only created to render a video interface and broadly understand how it works. Instead, it relies on the information gathered about each key point and the relationships between key points (this information is collected for each player between 10 and 120 times per second, which would be impossible to display in a video unless it were put in slow motion), as the authors may see in Figure 1.

4. Results

4.1. Adams Test

Statistical analysis of the results obtained for test 1 within the Adams test (aimed at assessing coordination skills for the experimental group) allows us to make some clarifications regarding the mathematical and statistical indicators. Thus, for the experimental group, test 1 highlights an average score of 8.76, test 2 shows an average score of 3.62, and test 3 reveals an average score of 0.76. For the control group, test 1 highlights an average score of 10.48, test 2 shows an average score of 7.95, and test 3 reveals an average score of 5.76 (

Table A2. Adams test—Experimental group/Control group—Descriptive statistics.

Group	Variables	No.	Minimum	Maximum	Arithmetic Mean	Standard Deviation	Coefficient of Variation
Experimental	Adams test—test 1	21	2	15	8.76	3.00	34.22
	Adams test—test 2	21	1	6	3.62	1.66	45.8
	Adams test—test 3	21	0	2	0.76	0.70	91.92
Control	Adams test—test 1	21	4	16	10.48	2.69	25.72
	Adams test—test 2	21	3	13	7.95	2.40	30.15
	Adams test—test 3	21	2	10	5.76	2.28	39.54

). In the context of the Adams test, the authors observe a significant improvement in the average score achieved by the experimental group, in contrast to the control group, which shows a modest improvement. The obvious progress of the experimental group is validated by the significant results obtained from the statistical analysis. All this information is presented in Table A2 from Appendix A

The following variables are normally distributed for both the experimental group and the control group: Adams test—test 1 and Adams test—test 2. The authors applied the Levene test for homogeneity of variances. For all three cases, p > 0.05, so the condition of homogeneity of variances was satisfied (

Table A3. Levene, Student (S), and Mann–Whitney U (MW) test values for the Adams test.

Variables	Experimental Group—Control Group
	Levene		Test Applied	Student/ Mann–Whitney U		Effect Size
	F	p		t or Z	p	d or r
Adams test—test 1	0.095	0.760	S	t = −1.949	0.058	-
Adams test—test 2	1.303	0.260	S	t = −6.813	0.0005	d = 2.154
Adams test—test 3	-	-	MW	z = −5.584	0.0005	r = 0.862

F—Levene’s value, p—significance level, d/r—effect size.

). By applying the Student test for independent samples, significant differences were recorded only for the variable Adams test—test 2, t(40) = −6.813; p = 0.0005, in favor of the experimental group, the effect being very large (d = 2.154).

The experimental group data for the variable Adams test—test 3 are not normally distributed. In test 3, the authors applied the Mann–Whitney U test for two independent samples. Significant differences were recorded in favor of the experimental group: z = −5.584; p = 0.0005; large effect, r = 0.862. All this information is presented in Table A3 from Annex 1.

4.2. Time Spent Playing with the Head Up (Gaze Directed Upwards)

The following variables are normally distributed for both the experimental group and the control group: Time spent playing with the head up—test 1 and Time spent playing with the head up—test 2. The authors applied the Levene test for homogeneity of variances. For the two tests, p > 0.05, so the condition of homogeneity of variances was satisfied (

Table A4. Levene, Student (S), and Mann–Whitney U (MW) test values for the variable Time spent playing with the head up.

Variables	Experimental Group—Control Group
	Levene		Test Applied	Student/ Mann–Whitney U		Effect Size
	F	p		t or z	p	d or r
Time spent playing with the head up—test 1	0.089	0.767	S	t = 1.54	0.131 p > 0.05	-
Time spent playing with the head up—test 2	0.377	0.542	S	t = 3.414	0.001	d = 1.08
Time spent playing with the head up—test 3	-	-	MW	z = −3.937	0.0005	r = 0.607

F—Levene’s value, p—significance level, d/r—effect size.

).

The authors applied the Student test for independent samples. Significant differences were recorded only for the variable Time spent playing with the head up-test 2, t(40) = 3.414; p = 0.0005, in favor of the experimental group, the effect being very large (d = 1.08).

Data of the experimental group for the variable Time spent playing with the head up—test 3 are not normally distributed. In test 3, the authors applied the Mann–Whitney U test for two independent samples. Significant differences were recorded in favor of the experimental group: z = −3.937; p = 0.0005; large effect, r = 0.607. All this information is presented in Table A4 from Appendix A.

Figure 2 shows us a comparison of means for the three testing sessions.

4.3. Time Spent Playing with the Head Down (Gaze Directed Downwards)

The following variables are normally distributed for both the experimental group and the control group: Time spent playing with the head down—test 1, Time spent playing with the head down—test 2, and Time spent playing with the head down—test 3. The authors applied the Levene test for homogeneity of variances to test the normally distributed variables. For all three cases, p > 0.05, so the condition of homogeneity of variances was satisfied (

Table A5. Levene and Student test values for the Time spent playing with the head down.

Variables	Experimental Group—Control Group
	Levene		Student
	F	p	t	p
Time spent playing with the head down—test 1	0.366	0.549	0.135	0.893
Time spent playing with the head down—test 2	0.641	0.428	−1.659	0.105
Time spent playing with the head down—test 3	0.907	0.347	−4.908	0.0005

F—Levene’s value, p—significance level.

). The authors applied the Student test for independent samples. No significant differences between groups were recorded in test 1 and test 2 (Table A5). In test 3, the difference between the experimental group and the control group was significant: t(40) = −4.908; p = 0.0005; effect size = 1.552, so it was very large. All this information is presented in Table A5 from Appendix A.

Figure 3 shows the comparison of means for the three testing sessions.

4.4. Successful Passes

The following variables are normally distributed for both the experimental group and the control group: Successful passes—test 1 and successful passes—test 2. Concerning the 3rd test, only for the control group the data are normally distributed.

The authors applied the Levene test for homogeneity of variances to check the normally distributed variables. For test 1 and test 2, p > 0.05, so the condition of homogeneity of variances was satisfied (

Table A6. Levene, Student (S), and Mann–Whitney U (MW) test values for the variable Successful passes.

Variables	Experimental Group—Control Group
	Levene		TEST Applied	Student/ Mann–Whitney U		Effect Size
	F	p		t or z	p	d or r
Successful passes—test 1	0.691	0.411	S	t = 0.76	0.451	-
Successful passes—test 2	0.025	0.876	S	t = 1.487	0.145	-
Successful passes—test 3	-	-	MW	z = −4.295	0.0005	r = 0.93

F—Levene’s value, p—significance level, d/r—effect size.

). The authors applied the Student test for independent samples. No significant differences between groups were recorded in test 1 and test 2 (p > 0.05) (Table A6).

The experimental group data for this variable at test 3 are not normally distributed. Therefore, the authors applied the Mann–Whitney U test for two independent samples. Significant differences were recorded in favor of the experimental group: z = −4.295; p = 0.0005; large effect, r = 0.93. All this information is presented in Table A6 from Appendix A.

Figure 4 presents the comparison of means for the three testing sessions.

4.5. Correlation

The authors notice that there is a positive, significant correlation between the time played with the head up and the number of successful passes r (17) = 0.696, p < 0.01 (df = n − 2), which means that the athletes of the experimental group who more the head up during the game makes a greater number of successful passes. If the authors refer to Cohen’s criteria, this relationship is a strong one. The coefficient of determination R² = 0.484, so 48.4% of the variation of the variable successful passes—test 3 is explained by the variable Time spent playing with the head up—test 3.

5. Discussion

5.1. Result Summary and Interpretation

The research aims to demonstrate the effectiveness of a modified training program that combines reaction speed work with decreased visibility work to improve passing. The training program includes using Heads Vests to train athletes to play more time with their gaze upwards. To evaluate the impact of the training, Cover’s method was used, and the Adam test was implemented for coordination performance. OpenPifPaf facilities were used for data collection and game performance analysis. The main hypothesis, H1, suggests that training children to play football with their heads up for longer periods will improve the number of successful passes and decrease the number of incorrect passes during a soccer match.

The study also introduced the Adams test™, which measures head tilt during a soccer game using artificial intelligence and visual recognition technology. The study included 42 boys aged between 8 and 10 years who participated in matches and tournaments. The experimental group consisted of 21 boys, while the control group consisted of 21 boys. The experimental program involved four BlazePods placed in the shape of a letter T, with participants sprinting to BlazePod B, touching its base, and returning to BlazePod B. The training program included 100 exercises, and a training tool called the HeadUp vest. The main hypothesis was that a drop in head angle during a soccer game improves the number of successful passes and decreases incorrect passes.

The study analyzed the Adams test results for tests 1 and 2 to assess coordination skills in the experimental group. The experimental group showed a significant improvement in the average score compared to the control group. The statistical analysis revealed that the experimental group had a larger effect on the Adams test, with a significant difference in the time spent playing with the head up (gaze directed upwards) and the time spent playing with the head down (gaze directed downwards). The Levene test for homogeneity of variances was satisfied for all three cases, and the Student test for independent samples showed significant differences in favor of the experimental group. The Mann–Whitney U test for independent samples also showed significant differences in favor of the experimental group. The study concluded that the experimental group demonstrated a significant improvement in coordination skills compared to the control group.

The study found no significant differences between groups in tests 1 and 2 but a significant difference in test 3 between the experimental and control groups. The data for successful passes were normally distributed for both groups, but the experimental group’s data were not normally distributed. The Levene test for homogeneity of variances was satisfied for tests 1 and 2, and the Mann–Whitney U test for two independent samples showed significant differences in favor of the experimental group. The study also found a positive, significant correlation between the time playing with the head up and the number of successful passes, with 48.4% of the variation in successful passes—test 3 explained by the variable time spent playing with the head up.

The experimental group showed a significant improvement in average scores on the Adams test compared to the control group. Statistical analysis showed significant differences between the three testing sessions, with athletes who played with their heads up having more successful passes. The intervention program and HeadUp vests were a decisive factor in achieving these results. Video analysis is becoming increasingly relevant in sports, especially for elite and senior athletes.

5.2. Comparison with Existing Research

Young athletes develop cognitive, behavioral, intrapersonal, and interpersonal skills through sports, preparing them for future elite careers. Continuous learning is crucial for their development, as it affects their attention and task-solving abilities. Concentration disruption, a maladaptive cognitive ability, can lead to negative emotions like depression and anxiety. Mental and social skills are crucial for young athletes’ progress, and sports psychology researchers have identified key mental skills like concentration. Research shows that attention, perception, and tracking of multiple objects positively impact sprinting, ball control, dribbling, and direction changes. Educating coordination skills at the right time is crucial for future football players’ performance [3,4,5,6,7,8,9].

The comprehension of a game [15], including identifying available areas, managing movements, and advancing with the ball, can be achieved by training children to maintain their gaze and improve their attention during play. This requires well-developed coordination ability and ball sense. Video analysis has become increasingly important in sports, collecting real-time data to help coaches, referees, players, spectators, and physicians [16]. Visual attention competencies can help athletes monitor multiple stimuli simultaneously, focus attention on significant elements, or ignore certain stimuli. In football, visual attention can be used to accurately observe teammates’ and opponents’ movements, facilitating appropriate decision-making [17,18]. Decision-making is essential for high-level athletic performance, and studies have shown that the football group showed greater post-test gains than the sedentary group in agility, visuospatial working memory, attention, planning, and inhibition. Involving all primary school students in sports activities can improve their attention levels and allow them to use complex means and techniques adapted to specific game situations [19].

In the current research it was calculated the average time spent playing with the head up on experimental group on three times momentum: T1 = 429.62 s (p = 0.13) → T2 = 513.9 s (p = 0.01) → T3 = 568.29 s (p-value = 0.005) and control group T1 = 388.86 s (p = 0.13) → T2 = 416.95 s (p = 0.01) → T3 = 429.86 s (p-value = 0.005). One may observe the difference between the experimental and control group T1 = 40.76 s → T2 = 96.95 s → T3 = 138.43 s. This huge difference is in favor of the experimental group that improved their understanding of a game, which includes seeing open spaces, controlling movements, and moving forward thanks to strong coordination and ball feel.

The study of attention and coordination aspects in sports activity needs an interdisciplinary approach to highlight relevant factors for different age groups and age characteristics [20]. The optimization of sports training in children’s football depends on the constant improvement of players’ psychological and physical characteristics. Implementing exercises that require good attention ability can increase coordination ability [21].

Distributed consistency is crucial in distributed systems, but improper use can lead to poor performance. Blazes, a cross-platform framework, helps identify coordination locations and automatically synthesizes code. The Flash Reflex (FRX) Training methodology uses BlazePod’s Pods and app to improve reactive intelligence and cognitive and physical abilities. This method uses reflexes to speed up brain processing and improve reaction time, enhancing personal performance and controlling scenarios [3].

Another study contributes to the knowledge about the importance of specialized training in the game of football for children, focusing on improving attention and coordination. One study explores technological advancements in recording head tilt angle during play using the OpenPifPaf project and Python programming. A machine-learning AI model analyzes posture data. Over the past 50 years, sports motion analysis systems have developed, making understanding sports mechanics more challenging. Although data acquisition equipment has improved, understanding data processing remains essential for sports [10]. Thus, the authors decided to use these technologies in our paper.

The HeadUp vest is a training tool designed to block a child’s gaze towards the ball during training. It requires a developed ball sense and optimal intersegmental coordination ability. The vest is made of sponge, safe, comfortable, and lightweight. It forces players to lift their gaze from the ground and adapt to new conditions. The vest also promotes reflexes, clairvoyance, and peripheral vision development.

The action system includes various types of exercises, such as standing coordination exercises, running coordination exercises, and themed games. Assessment instruments include the Adams test, T-Blaze test training, and time spent playing with the head up and down. The Adams test measures coordination skills by dividing a line into four spaces and aiming to perform as many jumps as possible within 15 s.

In the current research, it was also calculated the average time spent playing with the head down in experimental group T1= 342.19 s (p = 0.89) → T2 = 305.05 s (p = 0.105) → T3 = 191.05 s (p-value = 0.005) and control group T1 = 339.43 s (p = 0.89) → T2 = 341.14 s (p = 0.105), T3 = 259.81 s (p-value = 0.005). One may observe the difference between experimental and control T1 = 2.76 s → T2 = −36.09 s →T3 = −68.81 s. The consistently large decrease in this difference highlights the effectiveness of the Blaze technology application of the Adam test and the effectiveness of the training program, which is based on the cognitive, behavioral, attention, perception, and tracking of multiple objects. These factors have a positive impact on running, ball control, dribbling, direction changes, coordination skills, and task-solving abilities. They also have a positive impact on lowering negative emotions like anxiety and depression.

Game analysis is performed using the OpenPifPaf project, a machine-learning artificial intelligence model, and Python programming language. OpenPifPaf is more efficient in processing speed and accuracy than other methods, such as CrowdPose or PoseTrack2018. It is suitable for analyzing pictures and videos with larger numbers of people, making it suitable for football game analysis.

In the current research, it was also calculated the number of passes, based on OpenPifPaf technology, on experimental group T1 = 16.43 passes (p = 0.45) → T2 = 18.76 passes (p = 0.14) → T3 = 26.67 passes (p-value = 0.005) and control group T1 = 15.43 passes (p = 0.45) → T2 = 17 passes (p = 0.14) → T3 = 18.86 passes (p-value = 0.005). One may observe that the difference between the experimental and control regarding the number of passes increased in favor of the experimental group: T1 = 1 passes → T2 = 1.76 passes → T3 = 5.81 passes. Once again, our research proved that OpenPifPaf is very accurate in data processing and helps trainers and athletes understand sports mechanics.

5.3. Implications for Using HeadUp Vest

The main implications, the mechanism, and the theoretical explanation of using a vest to maintain heads up are:

• Physiological Monitoring and Performance Optimization
• Injury Prevention and Management
• Enhanced Tactical and Strategic Decision-Making
• Psychological and Emotional Wellbeing
• Personalized Training and Rehabilitation Programs

The HeadUp vest, paired with AI technology, could significantly impact football by monitoring physiological parameters and biomechanical data. AI algorithms analyze these data to provide insights into a player’s physical condition and performance metrics. This could help maintain homeostasis and adapt to changes in training loads, enhancing performance and reducing injury risk. The technology also aligns with the Supercompensation Theory, ensuring optimal training stress and recovery.

AI can predict injury patterns and risk factors through historical and real-time data analysis. Real-time alerts allow for adjustments in activity. The HeadUp vest uses cumulative load theory and kinetic chain theory to prevent injuries from repetitive stress and imbalances in body segments during movement.

The HeadUp vest uses AI to provide in-game data on team and individual performance, enabling coaches to make strategic adjustments. It also provides detailed information on ecological dynamics and situational awareness, enhancing team performance and enhancing responsiveness during matches.

The HeadUp vest monitors psychological indicators like stress and emotional states, providing personalized recommendations. It uses AI to analyze data, supporting the biopsychosocial model and cognitive load theory. This holistic approach helps maintain players’ mental resilience and focus, promoting optimal performance.

The HeadUp vest uses AI to create personalized training and rehabilitation programs based on individual needs and progress. It uses the Individual Differences Principle and Periodization Theory to optimize training outcomes and adjust programs dynamically based on real-time data.

5.4. The Practical Implications of the Study

The HeadUp vest, with its integration of AI technology, offers significant advancements in football by leveraging theoretical principles from physiology, biomechanics, psychology, and training science. Its ability to provide real-time, personalized insights enhances performance, prevents injuries, optimizes tactical decisions, and supports overall player wellbeing. These mechanisms and theoretical explanations highlight the profound impact of this innovation on the sport.

The present research found a significant positive correlation between head-up time and successful passes among experimental group athletes, indicating a higher number of successful passes reflected in a higher performance of athletes. Decreasing head-down time gives athletes a better understanding of their situation/playground and enables them to adopt the trainer’s recommended strategy or create their own strategies in real time. Overall, the training program and the using the HeadUp vest proved to be the winning approach in the football training and game.

Limitations of the study: application of the study only to U10 children playing football; U10 female samples were not included in the study. Strengths: the large number of tests aimed at assessing the participants’ attention and coordination; the specialized experimental program adapted to the study topic, age characteristics, and training level; and the relatively long implementation period of the experimental program.

6. Conclusions

The authors observe, in the context of the Adams test, a significant improvement in the average score achieved by the experimental group, in contrast to the control group, which shows a modest improvement. The obvious progress of the experimental group is validated by the significant results obtained from the statistical analysis. Recording the time spent playing with the head up but also with the head down reveals statistically significant differences between the three testing sessions in favor of the experimental group. Also, the athletes of the experimental group who had more head-ups during the soccer game had a greater number of successful passes, and thus our hypothesis is confirmed.

In light of the statistically significant influence obtained, the authors can conclude that our intervention program based on specific means and the use of HeadUp vests was a decisive factor in achieving these results. Given the continuous progress of video analysis, this method is becoming increasingly relevant in sports, especially for elite and senior athletes. The results of the current research suggest that video analysis can be successfully implemented for children as well, provided that the necessary resources and facilities are available to support this process.