Introduction
Fencing is one of the oldest sports, as it was created in the 16th century . The traditional rules were gradually transformed to create modern fencing in the late 19th century, becoming one of the first Olympic sports. The International Fencing Federation (Fédération Internationale d'Escrime, FIE) prescribes very strict standards for the organization of fencing competition.
Fencing practice essentially involves touching the opponent in the fastest and most precise manner while avoiding being touched in return. So that, precision and speed of the movements and ability to react to the actions of the opponent are crucial for improving the performance in this sport, both for the amateur athlete and the high-performance athlete .
Sports vision involves a comprehensive visual analysis, including a complete assessment of the athlete's visual skills, not only visual acuity but also other skills relevant to the sport. After the initial evaluation, sports optometrists can assist fencers by prescribing optical compensation, whether glasses or contact lenses, to reach the visual skill up to the maximum levels. Additionally, they can provide fencers with vision training exercises and other techniques for improving their visual skills and enhance their performance on the track .
This visual training involves the use of many techniques and exercises specifically designed to improve visual skills as hand-eye coordination, visual reaction time, and dynamic visual acuity . Currently, several digital devices are used for training the visual skills on athletes, including virtual reality goggles, stroboscopic eyewear , and devices with lights .
Fencing is very fast and visually demanding sport that requires quick reflexes, excellent hand-eye coordination, and the ability to anticipate and react to an opponent's movements . Therefore, not only a good vision is essential for success in fencing, but also a good use and application of the visual skills on the sport practice.
In recent studies, reaction time light systems having been used for evaluating and training the visual skills of athletes. They consist of a several LED illuminated devices controlled by a software that have been installed in a tablet or personal computer and allows the athlete to be shown many visual stimuli in a random manner , . These devices are used as targets for the athlete, who must activate or deactivate them according to the program's requirements. These systems are designed to collect performance data related to visual, cognitive, and dynamic reactions. They are primarily used as a measurement instrument, but they can also serve as a training tool to improve performance, primarily reaction time, movement time, and other sensorimotor-cognitive abilities. They are portable and easy to set up and use systems; characteristics that makes them a very useful tool for both training and testing purposes in individual and collective sports.
In this manuscript, we aim to showcase the possibilities of applying digital devices in a programmed visual training, in order to assist coaches to improve the sport performance. The main objective of this study is to demonstrate that working the reaction time with lighting systems allow coaches and athletes to know, measure, and improve their reaction times. The secondary objectives include designing exercises that athletes can transfer to real combat and measuring reaction times as a starting point in a training program to enhance shooters' performance.
Materials and Methods
Sample
The sample consisted of 48 fencers (22 males and 17 females; age: 21 ± 13 years) from several fencing clubs in the region of Andalusia, competing at amateur (10 athletes), regional (17 athletes), national (7 athletes), and international (5 athletes) levels in the three official fencing weapons: Epee (E), Saber (S), and Foil (F). The study was conducted with the voluntary participation of all athletes after signing an informed consent. The sample included 22 (45% of the sample) fencers younger than 18 years, in those cases, the parental consent was requested.
All the participants completed a previous anamnesis including questions about health status and basic visual evaluation including, among others, visual acuity and ocular motility. Subjects who reported having health problems in the last week, had monocular (right or left) visual acuity lower than 0.7, binocular visual acuity lower than 0.8, and altered ocular motility were excluded from the sample.
Materials
To carry out the data collection, we deployed two sets of four Queling Sport lights (Queling China). The 8 lighting devices were synchronized by Bluetooth version 5.0 with the ReactionX application (Queling, China), designed for Android devices. The diameter of the lighting devices is 9 cm and they were attached with Velcro to vertical panel of 1.6 x 1.0 meters with an image of a fencer (Figure 1). The lighting devices have a maximum operational range of 35 to 40 meters and a battery life of 3 hours. It is also worth noting the responsiveness of the devices to diverse stimuli, including vibrational cues, tactile interactions, close-range proximity (8 to 15 cm), or distal proximity (30-40 cm). We selected the “tactile interaction” mode for this testing, so that the device was activated only when any part of its surface was touched.
For carrying out the test in all the conditions, plastic weapons of the same length as those used in actual combat were employed. These weapons featured different grips, allowing athletes to choose the one closer to their own combat weapon.
In addition, it was used a tablet running on the Android operating system with the ReactionX application installed to control and synchronize activation of the light systems and recording the reaction times of the fencers in all the evaluated tasks.
Procedures
In collaboration with coaches, four exercises were designed and conducted to systematically train reaction times and decision-making, simulating the controlled scenarios and actions that fencers face during a bout. (See figure 1).
All measurements were taken during a fencing training session. The duration of all the tested conditions (exercise 1 to 4) was around 10 minutes. All the exercises were explained and practiced for 3 minutes first with the hand, and later with the plastic weapon.
For all the exercises, the data provided by the system included the results for all trials (times and correctness), the fastest, the slowest, and the average of all trials. The quantity and color of the lights for each trial were also indicated.
To test reaction abilities, four exercises were conducted, with different levels of uncertainty:
Exercise 1 for Simple Reaction Time (SRT):
The ReactionX application was configured on the device in "Order" mode, with lights positioned according to Figure 1. Only one light, following a specific sequence known to the athlete, was illuminated in a single color (green) and the fencer had to switch it off with a contact of the weapon. The interval between the extinguish of the light and the activation of the next light was randomized with times between 0.5 and 1.5 seconds.
The fencer was positioned at a distance allowing them to reach the lights with their weapon (typically around 80-90 cm, depending on the athlete's height) and he/she started the test in a fencing guard position with the weapon pointed towards the lights and the target was to switch off the green lights as soon as possible when they appear.
Given the athlete's foreknowledge of the upcoming light, anticipatory responses occasionally resulted. Responses faster than 100 ms were considered as anticipations and they were discarded. Additionally, responses longer than 1000 ms were considered as blockages of the fencers and they were also discarded.
The average number of trials (lights exhibited) in the minute of testing was 33 ± 10, and the total number of discarded trials for anticipation or blockage for all tested subjects was 33 (2.08% of the trials).
Exercise 2 for Elective Reaction Time (ERT):
The ReactionX application was set to "Standard" mode, and the reaction time lights were arranged as per the previous exercise (Figure 1). Here, only one light out of the four illuminated randomly in a single color, while all other configurations of the app were the same that the preceding exercise. Athletes were requested to identify, locate, and react for touching only the illuminated light as fast as possible. Elective Reaction Time (ERT), was defined as the time an individual takes to make a decision and respond to the visual stimulus which, in this case, had four possible options.
Responses faster than 100 ms were considered as anticipations and they were discarded. Additionally, responses longer than 1000 ms were considered as blockages of the fencers and they were also discarded.
The average of trials (lights exhibited) in the minute of testing were 27.69 ± 3.31 and the total number of discarded trials for anticipation of blockage for all the tested subjects were 22 (1.61% of the trials).
Exercise 3 Go/Non-Go Exercise (G/NG):
One of the main decisions a fencer must make is whether to enter in striking distance, to increase the possibility of touching during a bout. To assess this decision-making skill, the Reaction X application was set in "True-False" mode. In each trial a single green light (or none) was lightened with a 0.5-second delay between the extinguishing of the previous trial. The rest of the lights (or the four) were illuminated in red or yellow. In each trial, the lights remained illuminated for a maximum of 2 seconds.
The fencer was asked to extinguish only the green lights and disregarding lights of other colors. The sequence of the lights was always randomized by the system. The fencer positioned themselves at the same distance as in previous exercises, commencing from the fencing guard position. In this exercise, we measured time included apart from the elective reaction time, the go/non-go response time. The fencer had to react to the colors of the lights to execute a touch or to stay in the guard position. The application provided acoustic feedback on incorrectly extinguished lights, as well as lights that were missed.
Responses faster than 100 ms were considered as anticipations and they were discarded. Additionally, responses longer than 1000 ms were considered as blockages of the fencers and they were also dicharded.
The average of trials (lights exhibited) in the minute of testing were 17.27 ± 3.61 and the total number of discarded trials for anticipation of blockage for all the tested subjects were 3 (0.24% of the trials).
Exercise 4 Decision-Making Exercise (DM)
This is an evolution of the previous exercise where the fencer had to perform a specific type of exercise based on prior instructions, depending on the color of the light observed. The Reaction X application was configured on the device with the "Combined" mode. Only one lamp out of the four would light up randomly with one of the three predetermined colors (i.e. blue, red and green), each associated with a specific movement.
When the lamp illuminated with green light, the subject had to perform a lunge to touch the lights as quickly as possible (DML). If the illuminated light was red, the subject executed a marching movement to touch (DMM), and when the illuminated light was blue, they had to perform a breaking movement before executing a march or lunge to touch the activated device (DMB).
With the reaction time lights located in the same position as the others exercise (Figure 1), the fencers stood in the guard position at a distance from the lights that allowed them to touch the lights by making a typical fencing displacement with the weapon (usually about 100-120 cm from the target). In each module of the exercise (i.e. DML, DMM and DMB), the times taken to turn off the lights for each module were displayed.
Responses faster than 100 ms were considered as anticipations and they were discarded. Additionally, responses longer than 2500 ms were considered as blockages of the fencers and they were also discarded.
The average of trials (lights exhibited) in the minute of testing were 9.85 ± 4.24 and the total number of discarded trials for anticipation of blockage for all the tested subjects were 45 (3.17% of the trials).
It must be noted that, both fencers and coaches, were more interested in the fourth exercise as it closely resembled real combat.
Statistical Analysis
For the descriptive statistics, the average of all the attempts of each exercise was obtained, after discarding the anticipated responses and blockings.
The normality of all the considered variables was tested by the Shapiro-Wilk test. The normality was assumed only for the results of the DM4 exercises, but not the ERT, SRT and G/NG exercises. Given the limited sample size, in the results section of the manuscript both parametric and non-parametric test have been shown in the tables, according to the sample distribution. For all the statistical test the level of significance was set at = 0.05)
Kruskal-Wallis test were done for searching differences on the response times in the ERT, SRT and G/NG exercises by weapon, level of the fencer and gender, applying standard post hoc test for defining significant differences among groups.
ANOVA for independent samples were done for searching differences on the response times by weapon and level of the fencer, applying Tukey post hoc test for defining significant differences among groups. Additionally, the η² was determined to define the effect size, considering values <0.01 as small effect, values >0.01 and <0.14 as medium effect and values >0.14 as high effect size.
Finally, Student’s t test for independent samples was run for evaluating differences by gender in the results of the fourth exercise (DM4), and Cohen d was indicated to express the effect size of the differences by gender <0.2 as small effect, values >0.2 and <0.8 as medium effect and values >0.8 as high effect size.
1.Results
General results
Table 1 summarizes the mean values for each exercise for the entire sample. We note that the simple reaction time (SRT1) was (388 ± 146 ms), and the elective reaction time (ERT2) was (543 ± 115 ms). The mean value for the go/no-go exercise time (570 ± 79 ms) is quite similar to the elective reaction time. In the decision-making exercises, the fastest movement was the lunge (DML4) (1083± 185 ms), followed by the marching movement (DMM4) (1325 ± 234 ms), and the slowest was the breaking (DMB4) (16271 ± 268 ms).
| All (n=48) | Males (n=25) | Females (n=23) | K-W / t | P | Cohen d | |
|---|---|---|---|---|---|---|
| SRT1 * | 388 ± 146 | 429 ± 164 | 360 ± 121 | 0.843 | 0.358 | |
| ERT2 * | 543 ± 115 | 541 ± 146 | 546 ± 68 | 0.396 | 0.529 | |
| G/NG3 * | 570 ± 79 | 556 ± 85 | 576 ± 73 | 0.038 | 0.845 | |
| DML4 | 1083 ± 185 | 1104 ± 159 | 1016 ± 211 | 0.841 | 0.405 | 0.243 |
| DMM4 | 1325 ± 234 | 1331 ± 220 | 1275 ± 251 | 0.827 | 0.412 | 0.291 |
| DMB4 | 1627 ± 268 | 1612 ± 256 | 1675 ± 281 | -0.809 | 0.423 | -0.234 |
| DM complete | 1336 ± 319 | 1338 ± 291 | 1323 ± 348 | 0.241 | 0.810 | 0.040 |
Results by sex
The table 1 also displays the results separated by gender (males and females). As can be observed the Kruskal-Wallis test or the Student’s T test (depending in the character of the sample) do not show significant differences between genders in all the exercises) (p>0.05), and the results are quite similar without a clear trend on the differences by gender.
Results by weapon
Table 2 shows the averaged values by the fencer’s weapon. The ANOVA analysis pointed to significant differences only in the decision-making exercises, in the DML4 (F=4.741; p<0.05) and DMM4 (F=6.196; p<0.05), but not in the DMB4 (F=1.073; p=0.350). When the average of the three decision-making exercises were considered (DM4), the results showed also significant differences by fencers weapon (F=4.688; p<0.05). In all the cases, the Tukey post-hoc analysis indicated that the significant differences were between Sabre and Foil groups (p<0.05), with higher decision-making values in the Sabre group and lower values in the Foil group. In all the cases, the Eppe group had intermediate results without significant differences among the other two groups.
| Epee (a) (n=17) | Sabre (b) (n=18) | Foil (c) (n=13) | K-W / F | P | ||
|---|---|---|---|---|---|---|
| SRT1 * | 397 ± 113 | 473 ± 178 | 366 ± 115 | 3.358 | 0.187 | |
| ERT2 * | 561 ± 141 | 581 ± 117 | 544 ± 68 | 0.942 | 0.625 | |
| G/NG3 * | 564 ± 66 | 602 ± 99 | 568 ± 59 | 1.198 | 0.549 | |
| DML4 | 1110 ± 186 | 1192 ± 188 c | 999 ± 121 b | 4.741 | 0.014 | 0.174 |
| DMM4 | 1274 ± 215 | 1432 ± 211 c | 1166 ± 210 b | 6.196 | 0.004 | 0.216 |
| DMB4 | 1611 ± 282 | 1714 ± 308 | 1584 ± 165 | 1.073 | 0.350 | 0.046 |
| DM4 | 1332 ± 309 | 1446 ± 320 c | 1250 ± 299 b | 4.688 | 0.011 | 0.062 |
Results by level of the fencer
In Table 3 is can be seen the average results by the level of the fencer. The general trend of the results indicates the worst performance (longer times) in amateur fencers, followed by regional and national level fencers, with the best performance (shorter times) seen in international level fencers. However, the one-way ANOVA or Kruskal-Wallis analysis only exhibits significant differences (p<0.05) for the ERT2 (K=8,649; p<0.05) and the DMM4 (F=4.539; p<0.05) exercises, and for average of the three decision-making exercises (DM4) (F=4.688; p<0.05). In all the cases, the Tukey post-hoc analysis indicated significant differences were between the amateur and international level fencers (p<0.05). Additionally, for the decision-making exercises (DMM4 and DM4) they were also found significant differences between amateur and national level fencers (p<0.05).
| X | Amateur (a) (n=12) | Regional (b) (n=19) | National (c) (n=10) | International (d) (n=7) | K-W / F | P | |
|---|---|---|---|---|---|---|---|
| SRT1 * | 480 ± 194 | 435 ± 134 | 369± 101 | 329 ± 79 | 5,347 | 0,148 | |
| ERT2 * | 620 ± 127d | 572 ± 128 | 543 ± 479 | 479 ± 71a | 8,649 | 0,034 | |
| G/NG3 * | 620 ± 96 | 577 ± 72 | 580 ± 63 | 547± 89 | 2,471 | 0,481 | |
| DML4 | 1178 ± 240 | 1125 ± 160 | 1058 ± 190 | 1033 ± 100 | 1,277 | 0,294 | 0,080 |
| DMM4 | 1486 ± 156 c,d | 1287 ± 188 | 1180 ± 291 a | 1215 ± 226 a | 4,539 | 0,007 | 0,236 |
| DMB4 | 1763 ± 389 | 1666 ± 205 | 1595 ± 161 | 1439 ± 191 | 2,655 | 0,060 | 0,153 |
| DM | 1476 ± 363 c,d | 1360 ± 292 | 1277 ± 317 a | 1229 ± 241 a | 4,688 | 0,011 | 0,062 |
Proposal of reference values
With all the registered data, we have created a proposal of evaluation scale of the different exercises considered in the study (Table 4), which can be used for estimating the score of any fencer on each exercise. For better understanding and interpretation of the evaluation scale, we have applied it to obtain the points of the average values on each exercise for the four competition levels of fencers (shown in Table 3).
2. Discussion
Reaction time lights are employed to enhance visual skills across various sports. The analysis of reaction time and visual anticipation should be integral to the design and implementation of a visual training program aimed at improving athletes' visual skills to enhance their sports performance . They have demonstrated their validity in specific training sessions for fencers of varying levels to enhance the physical performance of fencers ,, .
Based on this premise and with the assistance of coaches, we formulated specific exercises for evaluating the simple and elective reaction time and decision-making in fencers. We will discuss the obtained results, first in general and, later, according the gender, the weapon and the level of the fencer and we will propose a draft of a reference scale to establish the level of the fencers on those visual skills.
General results
The general results indicated that the reaction and decision-making times are directly related to the difficulty of the task. The fastest times corresponds to the SRT1 (388 ± 146 ms) because it was used a single light and there is a lack of cognitive processing in the task. The fencer just touched the light when it was switched on.
The second exercise in term of cognitive complexity was the ERT2 (543± 115 ms), in which the fencers had four lights and they had to touch only the light that it was switch on.
If we consider the SRT2 and ERT2 results, we can observe the differences of the fencers’ reaction times between planned and unplanned actions; when the fencers knew which light would be illuminate (SRT1), the times were shorter than when they were unaware (ERT1), these results are slightly different in absolute values but similar with the findings from other researchers with other sport and protocols , .
The results for third exercise G/NG3 were quite similar to the ERT (570 ± 79 ms), similar to those found by other authors 0.523 ± 0.074 ms. At this point, it must be considered that, as it was mentioned in the data collection protocol, the number of mistakes or blockages in the ERT2 exercise was higher (22 discarded; 1,61% of the whole) than in the G/NG3 (3 discarded; 0.24% of the whole). That point is very important to interpreter the results because in the ERT2 exercise, the fencers started the movement when the four lights were activated, and the action could be optimized during the movement of the weapon. However, in the G/NG3 the action could not be corrected once the movement was started, and it could not be modified the trajectory of the weapon; so that, the decision-making had to be completely and perfectly done before starting the action. Tasks such as Go/No-Go, addressing choice and decision speed during exercise, and they can identify defects in visuomotor control and balance that may persist after recovery .
The most difficult tasks from a cognitive point of view were those included in exercise 4 (DML4, DMM4, DMB, and DM4 as the average of the three tasks). Considering all the decision-making exercises, we obtained results closely related to the technical aspects of the movements required of the fencer, even when significant differences between the exercises were not found. (p>0.05). The fastest movement was the lunge (DML4) (1083± 185 ms), which is the fastest attack in fencing, followed by the marching movement (DMM4) (1325 ± 234 ms), and the slowest results were for the breaking action (DMB4) (16271 ± 268 ms), because this movement requires a previous step backwards before starting the attack action. These times were higher than those obtained by other authors (1083± 185 ms vs. 601 ± 82ms); these differences may be due not only to the different methodology of the studies, but also to the different level of the fencers, as they had only had elite fencer of aiming weapons: foil and epee, and in our sample we had also non-elite level athletes including saber fencers. The saber fencers normally touch with the edge of the weapon and they start from a different guard position; these conditions were different to the requested in the exercises. Other authors who measured accuracy and touch times in fencing with a protocol similar to ours obtained closer exercise times 1336 ± 319 ms vs. 1680 ± 250 ms , (1336 ± 319 ms vs. 1194 ± 86 ms)
Results by sex
The results separated by gender do not show significant differences between males and females in any of the exercises. (p>0.05), being the results quite similar in most of the exercises without a clear trend on the differences by gender. Although other authors also found no significant differences in attack speed between gender with a sample of 9 males and 13 female young fencers 13.4±0.85 years old (men 3.6 ± 0.90 m/s and women 2.90 ± 0.72 m/s) . Other studies on fencing did show differences in the attack times of men compared to those carried out by women with a sample of 13 men aged 25.9 ± 2.8 years and 13 women aged 25.8 ± 3.1 years foil fencer finalists from the FIE competitions.
Results by weapon
When we analyze the data by weapon, the average values of the foil fencers were the fastest, in all the exercises, being the differences statistically faster than the sabre fencers in the decision-making exercises DML4 (Foil= 999 ± 121 ms vs Sabre= 1192 ± 188 ms) (F=4.741; p<0.05) and DMM4 (Foil= 1166 ± 210 ms vs Sabre= 1432 ± 211 ms) (F=6.196; p<0.05), but not in the DMB4 (F=1.073; p=0.350). When the average of the three decision-making exercises were considered (DM4), the results showed also significant differences by fencers’ weapon (F=4.688; p<0.05), with higher decision-making values in the sabre group and lower values in the foil group (Foil= 1250 ± 299 ms vs Sabre= 1446 ± 320 ms). In all the cases, the Eppe group had intermediate results without significant differences among the other two groups.
These results were expected because, although the Sabre fight is faster, its touch requires less accuracy and the usual touch is with the edge of the weapon, which is less accurate and not the requested technique in the exercises. Besides, the sabre fencers start from a different guard position, and they were not used the guard position requested on the exercises. These times were higher than those obtained (1110 ± 186 ms vs. 601 ± 82ms ) for sword fencer, and (999± 121 ms and 1194 ± 86 ms) for foil fencers, because these authors only had elite foil and epee fencers and in our sample were included fencers of different practice levels including amateur fencers.
Results by level of the fencer
When the averages obtained in the four exercises were separated by the level of competition of the athletes, The general trend of the results indicates the worst performance (higher times) in amateur fencers, followed by regional and national level fencers, with the best performance (lower times) seen in international level fencers, and the best results (lower times) were for the international level fencers, like other authors found for different sports (Van de Water et al., 2017). This is logical because the faster and more accurate is an athlete, the better can qualify in the championships; however, Kruskal-Wallis analysis only exhibited significant differences (p<0.05) for the ERT2 between the amateur fencer group and the international fencer group, (International= 479 ± 71 ms vs Amateur= 620 ± 127 ms) (K=8.649; p<0.05).
When analyzed one way ANOVA, significant differences were shown in the DMM4 exercises between the amateur fencer group (1486 ± 156 ms) compared to the national (1180 ± 291 ms) and international (1215 ± 226 ms) fencer groups (F=4.539; p<0.05), and for average of the three decision-making exercises (DM4) with lower times for the amateur group 1476 ± 363 ms) compared with the International (1229 ± 241ms) and National (1277 ± 317 ms) groups (F=4.688; p<0.05). In all the cases, the Tukey post-hoc analysis indicated significant differences were between the amateur, and national, international level fencers (p<0.05).
Application of the reference values
After evaluating the fencers on any of the exercises considered in this manuscript, their results can be scored with the proposed reference scale (Table 4). The example shown in the Table 5 allows see in a glance that the level of performance is directly related with the competition level of the fencer. This reference scale could also be applied in processes of talent detection as an easy way to evaluate the specific perceptive ability of the fencer, or to check the effect of the perceptive trainings, just comparing the scores before and after the intervention program.
Conclusions
This manuscript proposes four specific exercises for evaluating the simple and elective reaction time and decision-making in fencers. The general results indicate that the reaction and decision-making times are directly related to the difficulty of the task. As other previous studies, we confirm that the fastest times corresponds to the SRT exercise followed by, at the same level by the ERT and G/NG3 exercise; however, the number of blockage and mistakes and the technique of the required action must be considered in these cases.
The most difficult tasks from the cognitive point of view were the decision-making exercises (DML, DMM and DMB) in which the results are close related to technique of the movements requested to the fencer. The fastest movement was the lunge (DML), followed by the march (DMM) and the slowest was the breaking action (DMB).
The results separated by gender did not show significant differences between males and females in all the exercises. Considering the results by weapon, the foil fencers were faster than the saber fencers in all exercises, but they were only significantly faster in the DML and DMM exercises. In all the cases, the Eppe group had intermediate results without significant differences among the other two groups. Finally, when results were analyzed by the level of competition, performance in the all exercises was directly related with the competition level of the fencer.
We have edited a reference scale to evaluate the perceptive ability of the fencers that could be applied in processes of talent detection and for evaluating the specific perceptive ability of the fencers.
Acknowledgments
We thank all the fencing clubs of Andalusia that have participated voluntarily for their collaboration, sala de esgrima princesa de armas (Sevilla), Sala de esgrima (Bahia de cadiz Chiclana, Cadiz), Sala de esgrima de Huelva (Huelva), Sala de esgrima Almedina de Córdoba (Córdoba), and especially to the essential help to the fencing master D Emilio Quintela, for his collaboration and ideas for the development of the exercises.
References
1.
1. Murgu AI. Fencing. Phys Med Rehabil Clin N Am. 2006; 17(3), 725-736. https://doi.org/10.1016/j.pmr.2006.05.008
2.
2. FIE. Rules. Available from: https://fie.org/fie/documents/rules.
3.
3. Roi GS, Bianchedi D. The science of fencing: implications for performance and injury prevention. Sports Med. 2008; 38, 465-481. https://doi.org/10.2165/00007256-200838060-00003.
4.
4. Tsolakis C, Tsekouras Y, Daviotis T. Koulouvaris P. Papaggelopoulos P. Neuromuscular Screening to predict young fencers’ performance. J Biol Exer. 2018: 14. Available from https://doi.org/10.4127/jbe.2018.0134. https://www.researchgate.net/publication/325017779_Neuromuscular_Screening_to_predict_young_fencers'_performance
5.
5. Khathutshelo PM. A review of assessment and skill training methods used in sports vision. Afr. J. Phys. Health Edu. Recreat. Dance. 2014; 20 (1): 204 - 213. https://www.researchgate.net/publication/273103846_A_Review_of_Assessment_and_Skill_Training_Methods_used_in_Sports_vision
6.
6. Jung-Ho K, Ho-Jun S, Seung-Hyun L, Soonchul K VR/AR Head-mounted Display System-based Measurement and Evaluation of Dynamic Visual Acuity. J. Eye Mov. Res. 2019; 12(8) https://doi.org/10.16910/JEMR.12.8.1
7.
8.
8. Katanić B, Ilić P, Stojmenović A, Vitasović M. The application of Fitlight trainer system in sports. Fizička kultura. 20724; (2), 115-126. https://doi.org/10.5937/fizkul74-27189
9.
9. Hagemann N, Schorer J, Cañal-Bruland R, Lotz S, Strauss B. Visual perception in fencing: do the eye movements of fencers represent their information pickup?. Attent. percep. & psycho., 2010;72(8), 2204–2214. https://doi.org/10.3758/bf03196695
10.
10. Mackala K, Vodicar J, Zvan M, Kriza J, Stodolka J, Rauter S, Coh M. Evaluation of the Pre-Planned and Non-Planed Agility Performance: Comparison between Individual and Team Sports. Int J Environ Res Public Health. 2020; 17(3), 975. https://doi.org/10.3390/ijerph17030975.
11.
11. Shelly Z, Stewart E, Fonville T, Chander H, Strawderman L, May D, Bichey C. Helmet prototype response time assessment using NCAA Division 1 collegiate football athletes. Int. J. Kinesiol. Sports Sci. 2019; 7(4), 53-65. https://doi.org/10.7575/aiac.ijkss.v.7n.4p.53
12.
12. Örs BS, Cantas F, Gungor EO, Simsek D. Assessment and comparison of visual skills among athletes. Int J Perf Anal Spor. 2020; 1 https://doi.org/10.17155/omuspd.522342
13.
13. Kuan Y, Zuhairi N, Manan F, Knight V, Omar R. Visual reaction time and visual anticipation time between athletes and non-athletes. J. Public Health Med. 2018, 135-141. https://www.researchgate.net/publication/324441797_Visual_reaction_time_and_visual_anticipation_time_between_athletes_and_non-athletes.
14.
14. Vargas PC, Jiménez JM. Reducing Training Volume during Tapering Improves Performance in Taekwondo Athletes. J. Phys. Educ. Sport. 2018; 18 (4), 2221 - 2229. https://doi.org/10.7752/jpes.2018.04334
15.
15. Liu, Y., See, L., Chang, S., Lee, J., Shieh, L., Ning, Y., Lim, A.Y., & Chen, W. (2019). Simple and choice response time among elite and novice karate athletes and nonathletes. 2019. https://api.semanticscholar.org/CorpusID:201044382
16.
16. Rauter S, Coh M, Vodicar J, Zvan M, Krizaj J, Simenko J, Mackala K. . Analysis of reactive agility and change-of-direction speed between soccer players and physical education students. Hum. Mov., (2018) 19 (2), 68-74. https://doi.org/10.5114/hm.2018.74061
17.
17. Gutiérrez-Dávila M, Zingsem C, Gutiérrez-Cruz C, Giles FJ, Rojas FJ. Effect of uncertainty during the lunge in fencing. J Sports Sci Med. 2014; Jan 20;13(1): 66-72. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3918569/
18.
18. Mitchell KM, Cinelli ME. Balance control in youth hockey players with and without a history of concussions during a lower limb reaching task. Clin Biomech. 2019: 67, 142-147. https://doi.org/10.1016/j.clinbiomech.2019.05.006
19.
19. Gutiérrez-Davila M, Rojas FJ., Gutiérrez-Cruz C, García C, Navarro E. Time required to initiate a defensive reaction to direct and feint attacks in fencing. J. Appl. Biomech.. 2016; 32(6), 548-552. https://doi.org/10.1123/jab.2015-0333
20.
20. Witkowski M, Bojkowski Ł, Karpowicz K, Konieczny M, Bronikowski M, Tomczak M. Effectiveness and Durability of Transfer Training in Fencing. Int J Environ Res Public Health 2020; 17(3), 849. http://dx.doi.org/10.3390/ijerph17030849
21.
21. Gutiérrez-Cruz C, Rojas FJ, Gutiérrez-Davila M. . Effect of defense response time during lunge in foil fencing. J. Sports Sci. 2016; 34(7),651–657. https://doi.org/10.1080/02640414.2015.1068434.
22.
22. Kontochristopoulos N. Tsolakis C. Offensive and defensive efficacy among male and female elite foil fencers. J Hum Sport Exer. 2020, 15(2): 294-302. https://doi.org/10.14198/jhse.2020.152.05