Determinants of Decision Making in Novice and Elite Soccer Goalkeepers

Abstract

Eye tracking and EMG are novel measurement technologies that can be used to assess perceptual processes in sports under real-life conditions. The study was conducted on two groups of soccer goalkeepers (N = 60): Group A—expert goalkeepers (22.8 ± 2.15 years of age; training experience 12.77 ± 3.89 years); Group B—novice goalkeepers (15.70 ± 1.12 years of age; training experience 8.35 ± 2.68 years). Main findings: 1. The elite goalkeepers (Group A) focused most of their attention on only one main object (the foot of the opponent’s kicking leg) compared to novice youth goalkeepers (Group B), whose area of interest consisted of more elements: the knee, the lower leg, the foot of the attacking leg, and the ball. 2. The elite goalkeepers (Group A) showed a significantly shorter decision-making time (240–260 ms) than the novice goalkeepers (290–310 ms) in a two-on-one match situation. 3. The use of anticipatory perceptual skills resulted in more accurate anticipation and decision making in elite goalkeepers than in novice goalkeepers, whose perceptual patterns were more dispersed. 4. The anticipatory activity and structure of bioelectric tensions of the rectus femoris (RF) muscle provide useful information for the development of successful anticipatory actions. The ability to recognize signals is a prerequisite for combining movement sequences according to a predetermined pattern and allows for accurate decision making in the goalkeeper’s playing strategy.

1. Introduction

The study of decision making in various sports, including soccer, has taken on a new character in recent years. This has been brought about by technological improvements in eye trackers, which are used to assess perceptual processes in sports under real-life conditions. The new capabilities allow the recording of eye-tracking indices such as gaze fixation and saccades, which provide information about the areas and time of gaze fixation on a given object, as well as eye movements that reflect divided and alternating attention.

According to Schmidt’s motor control theory (control and regulation of human motor activities) [1], the learning and acquisition of motor patterns is conditioned by the speed of information processing. Three phases are critical: identification, choice, and programming of sensorimotor responses. The efficiency of cognitive processes plays a key role here because, according to Zbigniew Czajkowski [2], seeing is not the same as perceiving. In this sense, the basis for teaching sport movements and techniques should be the use of multifaceted perceptual training [3,4]. The present study combines two topics: decision making in sports competition, e.g., reactions of soccer or handball goalkeepers, and the acquisition of motor skills and habits in the long-term training process of players.

Effective perception must have significant anticipatory qualities (peripheral vision) that contribute to the development of athletes’ anticipatory abilities already in the pre-competitive training phase of given motor activities, e.g., soccer goalkeepers’ effective defense of free kicks or penalty kicks, or the anticipation of other technical-tactical actions in sports competition. With respect to peripheral vision, the determinants of goalkeeper decision making include training and competition experience, age, focused or distracted attention, current emotional state, knowledge, and ability to deal with various, often emergency, situations on the field [5,6,7,8]. It is very important to make decisions in such a way that they require the least amount of time—through so-called choice compromises. These compromises often make it difficult to draw conclusions based solely on the current situation on the field and eye movements. The information the goalkeeper receives from peripheral vision is also important. In addition to peripheral vision, the goal of the task and its end results are crucial [9,10]. The empirical determinant should be the analysis of indices expressed by fixations and saccades recorded by an eye tracker [11]. A comprehensive research approach should consider all of the above components.

In the present study, the projections of the visual observations are “heat maps” obtained through the application of eye trackers, showing the locations of the “focus” of the gaze of the tested goalkeepers.

An important innovation in the present study was the parallel combination of two research tools: eye tracking and surface electromyography (EMG). This approach makes it possible to record muscle tension and the rate of activation of specific muscles in response to anticipatory perceptual processes. Both eye-tracking and EMG indices provide answers to key questions about intergroup differences and interactions between the factors studied.

The following research hypotheses were adopted:

-

Heat mapping analysis determines the effectiveness of programmed training units focused on key technical-tactical solutions in two-on-one game situations. It is assumed that elite goalkeepers (Group A) focus most of their attention on the opponent’s foot of the attacking limb, unlike novice goalkeepers (Group B), whose area of interest includes a greater number of elements (knee, lower leg, foot of the attacking leg, ball), one second before and during the attacker’s shot on goal in an offensive shot-defense situation, i.e., the young goalkeepers’ gaze is more distracted.

-

Elite goalkeepers (Group A) show significantly shorter decision-making times than novice goalkeepers (Group B) in a typical two-on-one match situation.

-

EMG recordings show significant differences in the activity of postural muscles (RF—rectus femoris) between elite goalkeepers (Group A) and novice goalkeepers (Group B) in a match situation from the moment the attacker takes the shot until the final phase of the goalkeeper’s defense.

-

The use of integrated research tools (eye tracking and EMG) that simultaneously measure the investigated parameters is an innovative approach to improve training processes as a compilation of different views on technical and tactical actions in defensive game situations.

The present study is a multifaceted analysis of the visual-motor activity of soccer goalkeepers recorded under real-life conditions using two coordinated research tools in a typical two-on-one match situation. The results make it possible to assess the goalkeepers’ main areas of interest by analyzing the heat maps.

The study also aims to demonstrate the range of eye movements, including attention and decision making (decision time in a typical defensive situation). It also aims to determine the goalkeepers’ movement patterns through bioelectrical muscle tension (EMG) in the game situation, from the moment the attacker takes the shot to the final phase of the goalkeeper’s defense.

The main objective of the study was to evaluate eye-tracking and EMG indices as determinants of goalkeeper decision making according to their level (expert vs. novice).

2. Methodology

2.1. Participants

The study included 60 goalkeepers from Polish provincial teams (two groups of 30 players each: senior Group A and youth Group B). The main characteristics of the participants were: Group A: age 22.8 ± 2.15 years; training experience 12.77 ± 3.89 years, and Group B: age 15.70 ± 1.12 years; training experience 8.35 ± 2.68 years.

2.2. Procedures and Tools

The study was conducted after obtaining approval from the Bioethics Committee of the Medical Chamber (Resolution No. 346 of 23 June 2022, Bioethics Committee at the Opole Medical Chamber, ul. Grunwaldzka 23, Opole, Poland).

The main research tool was the Tobii Glasses 2 eye tracker worn on the eye (sampling frequency 50 Hz; H.264 scene camera video format, 1920 × 1080 pixels at 25 frames per second). iMotions 9.1 software was used to process the raw data. Based on the reference image from the eye-tracking device, on which the footballers focused their gaze (from the preparatory position, through focusing their attention on a specific object (AOI) or element during the waiting period, to the actual task performance), data on visual perception (information about the object) were obtained in relation to the areas of interest (heat maps) selected by the players.

The second research tool used a 16-channel EMG system (Noraxon, DTS, Desktop Direct Transmission System, Scottsdale, AZ, USA) to analyze the bioelectrical activity of the muscles under examination. The EMG system had a sampling frequency of 16 bits at 1500 Hz. Dedicated software was used to analyze the data (MyoResearch XP Master Edition for DTS Noraxon). A wireless transmitter-recorder (3-axis wireless DTS 3D acceleration sensor with a nominal output range of +/−6 g, sensitivity of +/−0.67 V/g and bandwidth of 5 Hz–1.8 kHz) was used.

The tests were conducted in accordance with the SENIAM project guidelines [12]. The analysis used the best of three attempts in a series of a given activity on the right and left sides of the football pitch.

2.3. Methods

The experimental procedures were adapted from the methodology described by Piechota et al. [13]. Before the start of the study, a typical individual football warm-up was performed. Eye-tracking glasses were placed on the head and electrodes on the legs. Photos of a senior goalkeeper are shown in Figure 1a,b. Adult goalkeepers (group A) and parents of teenage goalkeepers (group B) completed informed consent forms for participation in the study, as well as image use consent forms. Before each series of tests (3 trials on the left and 3 trials on the right side of the pitch), the eye-tracking glasses were calibrated using a single gaze point.

After putting on the eye tracker, each goalkeeper had electrodes placed on ten selected muscles of the lower limbs: the front part of the thigh muscles (RF RT and LT; VL RT and LT), the back part of the thigh muscles (BF RT and LT), and the gastrocnemius muscle (GAS MED. RT and LT; GAS LAT. RT and LT).

Figure 2 shows the positioning of the players on the pitch and the situation in the game from the serving dribble (S) to the receiving and shooting of the ball by the attacker (A). A series of three attempts was carried out on each side of the pitch. Figure 2 shows the situation on the left side, and then the same action (S-A) was carried out three times on the right side of the pitch. The best of the three attempts from each side was used for the video analysis of shooting accuracy.

3. Results

Figure 3a,b and Figure 4a,b show on which objects the goalkeepers fixated their gaze most frequently (heat maps). The fixations indicate the most frequent objects of interest. Thanks to the use of eye trackers, the heat maps obtained provided information about where their eyes were focused at any given moment. In the case of the elite soccer goalkeepers (Group A), their gaze was most often directed at the foot of the attacker’s leg that kicks the ball towards the goal. By comparison, novice goalkeepers (Group B) most often focused their gaze on the knee, lower leg, and foot (alternating between different areas).

The analysis of muscle bioelectric activity covered the sequence from the moment the attacker took the shot, after a previous pass from the right or left side of the field by the server, to the end of the goalkeeper’s defensive action. The best of three trials (from each side of the field) was selected for further analysis.

The one-way (independent) ANOVA test showed that the extensor muscles (RF RT) played a key role in the investigated two-on-one game situation (

Table 1. Results of one-way ANOVA between elite goalkeepers (Group A) and novice goalkeepers (Group B) for the EMG (mV) analysis activity of selected leg muscles.

Group			Descriptive Statistics			Test Statistics				Effect Size
Outcome	Predictor	Group	n	Mean	Standard Deviation	dfM	dfE	f	p	ήp2
RF RT (uV)	GROUP	A	30	4.567	3.839	1	58	4.776	0.033 *	0.08
		B	30	2.6	3.092
RF LT (uV)	GROUP	A	30	0.03	0.042	1	58	5.672	0.021 *	0.09
		B	30	0.01	0.021
VL RT (uV)	GROUP	A	30	4.867	3.721	1	58	0.765	0.385	0.01
		B	30	5.7	3.659
VL LT (uV)	GROUP	A	30	0.024	0.041	1	58	1.212	0.276	0.02
		B	30	0.036	0.045
BF RT (uV)	GROUP	A	30	3.967	2.906	1	58	3.361	0.072	0.05
		B	30	5.467	3.411
BF LT (uV)	GROUP	A	30	0.021	0.048	1	58	1.529	0.221	0.03
		B	30	0.037	0.051
GAS.LAT. RT (uV)	GROUP	A	30	5.3	2.996	1	58	0.124	0.726	<0.01
		B	30	5.033	2.859
GAS.LAT. LT (uV)	GROUP	A	30	0.027	0.044	1	58	0.926	0.34	0.02
		B	30	0.048	0.109
GAS.MED. RT (uV)	GROUP	A	30	4.933	2.49	1	58	3.429	0.069	0.06
		B	30	6.133	2.529
GAS.MED. LT (uV)	GROUP	A	30	0.02	0.039	1	58	1.544	0.219	0.03
		B	30	0.033	0.044

*: p ≤ 0.05.

). The mean value for RF RT (uV) was higher in Group A (m = 4.57, SD = 3.84) than in Group B (m = 2.6, SD = 3.09). A one-way analysis of variance (ANOVA) showed that this difference was statistically significant (F(1.58) = 4.78, p = 0.033, ή²_p = 0.08). A series of post hoc Tukey’s HSD tests were then performed on each individual pair of results. A significant difference was found between Group A and Group B (p = 0.033) (Table 1).

The RF RT muscles played a key role in the first phase of the goalkeeper’s response to the attacker shooting the ball. The goalkeeper’s position on his toes, with his knees bent, lowering his center of gravity, prepared his body to react and intervene immediately. At the moment of the shot on goal, there was very little time for effective intervention. Therefore, the activity of the goalkeeper’s thigh and calf extensors was very important. This observation confirmed the importance of the preparatory phase in the process of visualizing the structure of the motor program to be executed.

Another factor that was analyzed was the goalkeepers’ decision-making time (Groups A and B) in a typical two-on-one game situation.

Table 2. Results of Mann-Whitney U test for decision-making time between elite goalkeepers (Group A) and novice goalkeepers (Group B).

Group			Descriptive Statistics		Test Statistics
Outcome	Predictor	Group	n	Median	U	z	p
Decision-making time	GROUP	A	30	250	18.5	−6.417	<0.001
		B	30	300

and Figure 5 show significant differences in decision-making time from the moment the ball was struck until the goalkeeper completed the defensive sequence (goalkeeper catches the ball, ball hits the post or crossbar, or attacker scores).

When the ball crossed the goalkeeper’s body, the time in the elite group of goalkeepers ranged from 240 to 260 ms (mean 250 ms), while in the novice group, it ranged from 290 to 310 ms (mean 300 ms). Group B had higher scores for defense time (300) than Group A (250). The Mann-Whitney U test showed that this difference was statistically significant (U = 28.5, p < 0.001, r = 0.81).

4. Discussion

For goalkeepers, decision making is the basis of competition in soccer, given their unique position on the field. The high level of cognitive processing that should characterize goalkeepers determines the effectiveness of defensive actions in soccer teams. In these complex actions, anticipation of actions of individual players as well as offensive formations of the opposing team is crucial. To achieve a high level of perception, specific perceptual training is required. Visual perception should be considered the first stage in the decision-making process of soccer players and goalkeepers. The ability to focus and divide attention is a key element in several technical actions in response to challenges (shots on goal) from opponents [14,15]. Perceptual-cognitive skills are also related to motivational factors, which are provided by guidance (feedback) from the coaching staff [14]. The technical-tactical demands of one’s own and the opponent’s game must also be taken into account, as they comprehensively influence the final decision [16]. In this sense, peripheral perception as well as ideomotor (imaginal) factors promote decision-making effectiveness.

The main premise of the present study was to integrate informational and neurophysiological processes with the speed of neuropsychological processes. The results of the study contribute to a novel and holistic understanding of the determinants of decision making in soccer. The following conclusions can be drawn from the study:

• The heat mapping analysis showed that the ability to perceive objects in a defensive situation affects the performance of programmed training units aimed at key technical-tactical solutions in two-on-one match situations. The initial hypothesis was confirmed that elite goalkeepers (Group A) focused most of their attention on only one main object (the foot of the attacking leg), compared to young novice goalkeepers (Group B), whose area of interest consisted of a greater number of elements: the knee, the lower leg, the foot of the attacking leg and the ball, before and during the attacker’s shot on goal.
• The elite goalkeepers (Group A) showed significantly shorter decision times (240–260 ms) than the novice goalkeepers (Group B) (290–310 ms) in a typical two-on-one match situation.

In elite goalkeepers, anticipatory perception led to more accurate prediction and decision making than in novice goalkeepers, whose perceptions were more dispersed. The better results obtained by the experienced goalkeepers may indicate differences in motor sequence-specific search behavior and cognitive processes between the two groups studied [17]. Thus, early prediction and identification of the opponent’s intentions allows goalkeepers more time to prepare and make a final decision [18,19]. The ability to anticipate more than the timing of decision making differentiates the sport level of soccer players [13,20]. According to Witkowski et al. [21], anticipating the opponent’s actions already in the preparation phase is based on the “visual image” of the given situation, including the opponent’s likely targets and movements, in motor memory [22]. With accurate pre-imagination, players can program and continuously adapt their motor activities [23] to the current and changing situation on the field.

3.

The presented structure of muscle bioelectric tensions (EMG) in elite goalkeepers confirmed the hypothesis of significant differences in postural muscles (rectus femoris). Postural muscles also include thigh flexors and gastrocnemius muscles. The rectus femoris muscle is the first to be activated (a specific movement pattern structure) in a typical defensive situation. The calf muscles (gastrocnemius) have also been mentioned in previous studies. These muscle groups play a key role in every defensive situation, from assuming the correct goalkeeper position to reaching (diving for) the ball.

The thigh muscles, which are considered postural muscles, are fundamental in the case of goalkeepers due to the incredible dynamics of their technical-tactical actions during a match. The short reaction time (decision making) of goalkeepers requires very strong and at the same time powerful leg muscles to perform quick changes of direction to run to the ball, vertical jumps to intercept the ball, or sudden changes of decision. Leg strength improves goalkeepers’ decision making time [24,25].

The EMG analysis showed that bioelectrical muscle tension plays an important role in goalkeepers’ match performance. Hypothetically, the higher this tension, the better the coordination and the body’s ability to respond to dynamic changes (sudden changes in decision making). According to Zachry et al. [26], increased EMG activity may be indicative of “noises” in the motor system. These noises, in turn, may result from intrinsic excessive attentional focus (focus on the activation of specific motor units in a given movement sequence), which hinders movement precision and control. The noise therefore makes performance less precise and reliable and extends the relevance of the task, i.e., decision making. This is due to the basic fact that athletes should focus on the execution of the movement itself, i.e., use marginally short decision time, while maintaining the quality of the movement in accordance with the accepted movement pattern. This determines the quality and efficiency of the postural muscles, which increase the accuracy of movement execution. On the other hand, Vance et al. [27] confirmed that decreased neuromuscular activity is associated with increased movement accuracy. Reduced EMG activity and possibly reduced motor unit recruitment are indicative of improved movement accuracy and fluidity. In a defensive game situation, experienced goalkeepers generate significantly lower muscle bioelectric tensions, also in the postural muscles.

The recording of the anticipatory structure of bioelectric muscle tension, mainly in the rectus femoris muscle, is a novelty since the recorded activity is a source of information for effective anticipatory actions. A great advantage of the elite goalkeepers was also their faster and more rational information processing, which was expressed by their lower EMG results.

4.

The use of integrated research tools (eye tracking and EMG) with simultaneous measurement of the studied parameters is an innovative approach to improve the training process as a compilation of different views on technical-tactical actions in defensive situations. A synchronous analysis under real training conditions can provide answers to key questions about intergroup differences and interactions between the factors studied.

In summary, perceptual processes are closely related to motor learning because, in order to recognize the correct signals, soccer players must have already seen them and practiced the appropriate responses. These types of motor reactions and behaviors are a kind of prototype of motor skills and sports technique. The ability to recognize signals is a prerequisite for combining sequences of movements according to a predetermined order and allows for proper decision making in the coordination of upper and lower limb actions, which is so important in sports with open habits, including soccer goalkeeper technique and game strategy.

Perception of stimuli should be not only selective, but also active, including awareness of the goals of the sport competition. An important criterion is the maximum reduction of processing time, which affects the speed and quality of activation of the relevant muscles. Therefore, an important training task is to teach athletes to perceive stimuli selectively, to understand them correctly, and to subtly discriminate among them. This includes the development of the ability to ignore irrelevant or false signals, i.e., to distinguish between real and fake actions, e.g., strikers’ feints when executing shots or free and penalty kicks.

5. Conclusions

• The elite soccer goalkeepers (Group A) focused most of their attention on only one main object (the foot of the attacking leg) compared to novice soccer goalkeepers (Group B), whose area of interest included a greater number of elements: knee, lower leg, foot of the attacking leg, and ball.
• The elite goalkeepers (Group A) showed significantly shorter decision times (240–260 ms) than the novice goalkeepers (Group B) (290–310 ms) in a typical defensive two-on-one match situation.
• Anticipatory perception leads to more accurate anticipation and decision making by elite goalkeepers than by novice goalkeepers, whose perceptions were more dispersed.
• Anticipatory activity and the structure of bioelectrical tension of the rectus femoris muscle are valuable sources of information for the execution of effective anticipatory actions. Experienced goalkeepers also have the advantage of being able to process information quickly and make decisions in dynamically changing conditions.
• Synchronous analysis under actual training conditions can provide answers to key questions about intergroup differences and the relationships between the studied factors.

6. Limitations

This study has several limitations that should be taken into account in future research.

One of the key factors is the difference in training experience, which can have a significant impact on the final results.

Another difference may be the terrain on which the players train (indoor hall, grass pitch, artificial turf pitch).

Another important and frequently used situation, apart from 2:1, is a free kick in football.

Further research should be conducted using the above key factors and the results should be compared.