Abstract

Background:

Stroboscopic training uses intermittent visual occlusion to disrupt continuous visual information flow, inducing adaptive improvements in visuomotor processing. Although its efficacy has been demonstrated across several fast-response sports, its application in boxing — particularly among female athletes — remains largely unexplored. This study investigated the effects of a six-week stroboscopic training program on visual performance skills and punching accuracy in female amateur (Olympic-style) boxers, and examined the durability of training effects at a four-week follow-up.

Methods:

Twenty-six female amateur boxers (mean age 24.69 ± 5.48 years; mean boxing experience 7.19 ± 2.51 years) were randomly assigned to a stroboscopic training group (n = 13) or a non-stroboscopic control group (n = 13). Visual performance was assessed across 10 metrics using the Senaptec Sensory Station at pre-test, post-test, and retention. Punching accuracy (%Hit) was assessed from video recordings of official bouts and formal intra-team sparring sessions. The six-week intervention consisted of three 35-minute sessions per week combining general visual reaction drills and boxing-specific reactive tasks, performed under stroboscopic or normal visual conditions.

Results:

Significant group × assessment phase interactions were observed for five visual performance metrics: Eye-Hand Coordination (EHC; F(2,48) = 6.105, p = 0.004, ηp² = 0.203), Reaction Time (RT; F(2,48) = 3.230, p = 0.048, ηp² = 0.119), Go/No-Go (GNG; F(2,48) = 4.807, p = 0.013, ηp² = 0.167), Perception Span (PS; F(2,48) = 6.005, p = 0.005, ηp² = 0.200), and Multiple Object Tracking (MOT; F(2,48) = 6.039, p = 0.005, ηp² = 0.201). Punching accuracy also improved significantly (F(2,48) = 4.626, p = 0.015, ηp² = 0.162), with post-test %Hit (35.02 ± 7.27%) significantly higher than pre-test (28.85 ± 7.45%). Training effects on EHC, GNG, PS, and MOT were partially retained at the four-week follow-up; RT and %Hit improvements were not retained. No significant changes were observed for basic visual functions (VC, CS, DP, TC, NFQ) in either group.

Conclusions:

Six weeks of stroboscopic training significantly improved key visuomotor skills and punching accuracy in female amateur boxers, with partial retention of effects four weeks post-intervention. These findings support the integration of stroboscopic training into evidence-based boxing preparation programs and highlight the need for periodic booster sessions to sustain long-term gains.

1 Introduction1.1 Visual performance skills in boxing

Visual performance skills represent a multifaceted integration of sensory perception, neural processing, and motor response coordination—capabilities that are not merely supplementary but foundational to elite performance in combat sports, especially boxing (Müller et al., 2024; Tarasi et al., 2024). Unlike static visual tasks (e.g., reading or identifying stationary objects) that rely on sustained focus on unchanging stimuli, boxing imposes extreme dynamic visual demands: athletes must simultaneously track an opponent’s limb movements, facial cues, and body posture; anticipate the trajectory and force of incoming punches; calibrate spatial distance between themselves and their opponent; and initiate precise, timed motor responses—all within a time window as short as 100–300 milliseconds for defensive reactions and 402–405 milliseconds for offensive punch delivery (Stanley et al., 2018). This high-stakes, time-constrained environment elevates visual performance to a primary differentiator between successful and unsuccessful boxers, as even a 50-millisecond delay in visual processing can result in a missed defensive opportunity or an inaccurate punch.

Key dimensions of visual performance that are particularly salient to boxing include eye-hand coordination (the ability to synchronize visual input with manual motor actions), rapid reaction time (the latency between visual stimulus detection and motor initiation), selective attention (the capacity to focus on relevant cues—e.g., an opponent’s jab—while filtering out distractions), multiple object tracking (monitoring multiple moving targets, such as an opponent’s hands and feet), and near-far quickness (rapidly shifting focus between close-range targets, like a focus mitt, and distant targets, like an opponent’s torso) (Buscemi et al., 2024; Lee et al., 2022b; Millard et al., 2020);. Epidemiological and sports science research has consistently confirmed the primacy of vision in athletic decision-making: approximately 80% of the sensory information athletes use to navigate competitive environments and execute tactical choices is derived from the visual system (Ripoll et al., 1995). In boxing, this statistic is amplified, as the sport’s confrontational nature leaves no margin for error in interpreting visual cues—every misjudgment of distance, timing, or opponent intent can lead to a competitive disadvantage or injury.

Elite boxers exhibit measurable superiority in visual performance metrics compared to sub-elite athletes and non-athletes, underscoring the role of visual skill in athletic development. For example, a study by Laby et al. (2011) demonstrated that Olympic-level boxers exhibited significantly superior stereoacuity (a critical component of depth perception) compared to other athletes, enabling them to more accurately estimate punching range during both close-quarters clinches and long-range striking exchanges. This precision is particularly valuable in modern boxing, where scoring criteria prioritize technical accuracy over brute force. Thomson et al., (2013) adoption of the “10-point must” scoring system marked a paradigm shift in bout evaluation, reorienting judges’ focus toward four core criteria: effective punching, defensive proficiency, ring generalship, and clean punching (Thomson et al., 2013). Among these, effective punching—defined as the number of legal punches landed on valid target areas (head, torso above the waist, and shoulders)—carries the highest weighting, accounting for approximately 40% of a judge’s scoring decision (Davis et al., 2013).

This scoring reform has disproportionately elevated the importance of punching accuracy (calculated as the percentage of successful punches relative to total punches thrown) in women’s boxing. Unlike men’s boxing, where knockout power is often a decisive factor, women’s boxing typically emphasizes technical precision, tactical awareness, and consistent punch placement—traits that are directly linked to visual performance. Dunn et al. (2017) further validated this relationship in a longitudinal study of 52 amateur female boxers, finding that athletes who landed 15% more scoring punches than their opponents had a 78% higher win rate in regional and national competitions. As a result, female boxers and their coaching staff have increasingly prioritized evidence-based visual training interventions as a means of improving punching accuracy, with targeted visual skill development emerging as a critical gap in current training paradigms.

1.2 Visual training in combat sports

Although targeted visual training has been proven to have significant efficacy, research on visual training in boxing has developed relatively slowly, and its overall effectiveness remains insufficient. Conventional visual training approaches typically include static eye exercises (e.g., eye rolls, focus on stationary objects), simple light reaction drills (e.g., responding to a single flashing light), and shadowboxing with static focus mitts. While these methods can produce modest improvements in basic visual skills (e.g., static focus), they lack the realism and complexity of actual competitive conditions. For instance, static eye exercises do not replicate the dynamic, fast-paced visual stimuli athletes encounter in the ring, and simple light drills fail to simulate the multi-cue environment of a boxing match (e.g., simultaneous movement of an opponent’s hands, feet, and body) (Appelbaum and Erickson, 2018). This disconnect limits the transferability of training effects to in-ring performance, as athletes are not challenged to apply their visual skills in contextually relevant scenarios.

A growing body of empirical research has established robust correlations between visual performance skills and striking performance in boxing, confirming that visual skill is not an innate trait but a malleable ability that can be enhanced through targeted training (Wu et al., 2024; Appelbaum et al., 2016). Wu et al. (2024), in a landmark study of 38 male amateur boxers, found that eye-hand coordination (r=0.62, p<0.001), reaction time (r=-0.58, p<0.001), and visual tracking ability (r=0.55, p<0.001) were strongly correlated with punching accuracy, indicating that improvements in these visual metrics could directly translate to better in-ring performance. Importantly, these correlations were not limited to male athletes: a follow-up study by the same research team, focusing on 29 female amateur boxers, reported similar effect sizes (r=0.59 for eye-hand coordination, r=-0.54 for reaction time), confirming the generalizability of the relationship between visual skill and punching accuracy across genders.

In recent years, technological advancements have paved the way for innovative visual training methods that address the limitations of traditional approaches, including prismatic adaptation (Giustino et al., 2024) and stroboscopic training. Among these, stroboscopic training has emerged as one of the most promising interventions, with a growing body of research demonstrating its efficacy in enhancing visual performance and sport-specific skills across a range of fast-response sports, including badminton, baseball, and volleyball (Appelbaum et al., 2016; Appelbaum and Erickson, 2018; Hülsdünker et al., 2019). Stroboscopic training differs from traditional visual training in its ability to simulate the dynamic, high-pressure visual environment of boxing by using intermittent visual occlusion (via specialized glasses) to disrupt the continuous flow of visual information. This “visual stress” induces adaptive changes in the nervous system, enhancing the brain’s ability to process visual cues quickly and accurately—skills that are directly applicable to boxing’s demands. Unlike traditional methods, stroboscopic training is inherently contextually relevant, as it requires athletes to apply visual skills in dynamic, movement-based tasks, thereby maximizing the transfer of training effects to in-ring performance.

1.3 Stroboscopic training: mechanisms and efficacy

The underlying mechanism of stroboscopic training is analogous to high-altitude training for endurance athletes (Appelbaum and Erickson, 2018). By creating a “stressful” visual environment (intermittent occlusion), the training induces adaptive changes in the nervous system that enhance performance under normal visual conditions. Specifically, stroboscopic stimulation has been shown to improve rapid information encoding, central and peripheral visual sensitivity, temporal anticipation, and functional connectivity between key brain regions involved in visual processing and motor control (Appelbaum et al., 2011, 2012; Lee et al., 2022a). Neuroimaging studies have further revealed that stroboscopic training enhances functional connectivity between the superior parietal lobule (SPL)—which may be important for spatial attention and visual-motor integration—and the premotor cortex (PMC)—which is involved in planning and executing motor responses (Ellison et al., 2020). This strengthened connectivity shortens the delay between visual input and motor initiation, a key advantage in fast-response sports like boxing.

A growing body of research has demonstrated the efficacy of stroboscopic training in improving sport-specific performance across various disciplines. Hülsdünker et al. (2019), for example, investigated the effects of a four-week stroboscopic training program in top-level badminton players, finding significant improvements in visual reaction time (reduced by 12 ms) and shuttlecock tracking accuracy (increased by 22%). Similarly, Liu et al. (2020) reported that a six-week dynamic vision training program (incorporating stroboscopic training) significantly improved launch angle and hit distance during batting practice among collegiate baseball players, further supporting the potential transfer effect of stroboscopic training to sport-specific batting performance (Liu et al., 2020). Other studies have shown positive effects in volleyball (improved reactive agility and saccadic eye movement speed (Zwierko et al., 2023)), soccer (enhanced dribbling accuracy and visual search efficiency (Palmer et al., 2022)), and softball (increased hitting accuracy and decision-making speed (Dendy et al., 2025)). These findings suggest that stroboscopic training is a versatile intervention that can be adapted to the specific visual demands of different sports.

Despite its proven efficacy in other fast-response sports, stroboscopic training has not been systematically investigated in boxing. This represents a significant gap in the literature, given the sport’s heavy reliance on visual performance skills. Hülsdünker et al. (2021) explicitly noted the need for validating stroboscopic training effects in sports beyond badminton, highlighting boxing as a compelling candidate due to its high visual-motor demands. Additionally, Wu et al. (2024) called for further research exploring how targeted visual training might enhance boxing performance, particularly in female athletes, who have been underrepresented in previous studies.

1.4 Rationale and research hypotheses

The primary rationale for this study is to address the gap in the literature by investigating the effects of stroboscopic training on visual performance skills and punching accuracy in female amateur boxers. Female boxers face unique physiological and competitive challenges (e.g., smaller body mass, different weight class distributions, and distinct scoring tendencies), making it inappropriate to generalize findings from male cohorts. Additionally, the study explores the retention of training effects four weeks post-intervention, a critical question for practical application (i.e., whether improvements are sustained beyond the training period). The present study therefore pursues three primary aims: (1) to determine whether a six-week stroboscopic training program improves key visual performance skills (EHC, RT, GNG, PS, MOT) in female amateur boxers compared to a control group; (2) to examine whether stroboscopic training enhances punching accuracy (%Hit) in official competitive settings; and (3) to assess the durability of any observed training effects at a four-week follow-up retention test.

Based on the existing literature, we formulated three primary research hypotheses:

Hypothesis 1: A six-week stroboscopic training program will significantly improve key visual performance skills (eye-hand coordination, reaction time, Go/No-Go, Perception Span, and Multiple Object Tracking) in female amateur boxers relative to a control group receiving non-stroboscopic training.

Hypothesis 2: The stroboscopic training program will significantly improve punching accuracy (%Hit) in female amateur boxers during official competition relative to the control group.

Hypothesis 3: The improvements in visual performance skills and punching accuracy will be retained four weeks after the conclusion of the training program, albeit with a potential slight decline in effect size.

By addressing these aims, the study provides evidence-based guidance for coaches and sport scientists on integrating stroboscopic training into female boxing preparation programs.

2 Methods2.1 Experimental design

This study used a randomized controlled trial design with two between-subjects groups (experimental vs. control) and three within-subjects time points (pre-test, post-test, retention test). The study timeline spanned 12 weeks: 1 week of baseline assessment (pre-test), 6 weeks of training intervention, 1 week of post-test assessment, and a 4-week de-training follow-up period culminating in the retention test. The precise schedule was as follows: (1) Pre-test – Visual performance assessments were conducted at the Motion Vision Performance Laboratory, Dalian University, and boxing performance data were collected during an official team competition held in Xi’an, Shaanxi Province, with all bouts concluded by 28 April 2025. (2) Six-week training intervention – Early May to mid-June 2025, conducted entirely at the Beijing Boxing Team’s home training facility. (3) Post-test – Completed during the week of 9 June 2025, comprising laboratory visual assessments at the Motion Vision Performance Laboratory, Dalian University, and a formal intra-team sparring bout conducted under official competition rules at Dalian University. (4) Retention test – Conducted on 14 July 2025 (four weeks after cessation of training) at Dalian University, comprising both laboratory visual assessments and a formal intra-team sparring bout. All visual performance assessments were administered under standardized conditions (same time of day: 9:00–11:00 AM; same laboratory environment, Motion Vision Performance Laboratory, Dalian University; same assessor, blinded to group assignment) to minimize confounding variables.

2.2 Participants

We recruited 26 trained female amateur boxers. All participants were members of the Beijing Boxing Team and were recruited directly by the research team through the team’s coaching staff.Inclusion criteria were: (a) active participation in amateur (Olympic-style) boxing training for at least 3 years of competitive experience; (b) engagement in at least six days of boxing-specific training per week (≥2 hours per session); (c) participation in the 2025 China National Boxing Championship; (d) normal visual acuity (≥20/20 Snellen equivalent) and normal visual function (no history of ophthalmological disease, strabismus, or amblyopia)—these data were self-reported by participants via a standardized screening questionnaire administered at recruitment (it should be noted that self-reported visual status represents a methodological limitation, discussed further in the Study Limitations section); (e) no prior experience with stroboscopic training or other specialized visual training programs; (f) no acute injuries (e.g., fractures, sprains) or concussions within the past 6 months; (g) willingness to comply with the training protocol and assessment schedule. Exclusion criteria were: (a) use of corrective lenses (refractive correction was not permitted, as it could alter visual performance and interact with the stroboscopic intervention); (b) pregnancy; (c) chronic neurological conditions (e.g., epilepsy, migraine with aura) that could be exacerbated by stroboscopic stimulation.

The final sample consisted of 26 female amateur (Olympic-style) boxers (mean age 24.69 ± 5.48 years; range 18–35 years; mean boxing experience 7.19 ± 2.51 years; range 3–12 years; height 170.81 ± 6.66 cm; weight 67.07 ± 7.39 kg). All participants were members of the Beijing Boxing Team. They represented six official weight categories in accordance with International Boxing Association (IBA) regulations: flyweight (50 kg, n=3), bantamweight (54 kg, n=1), featherweight (57 kg, n=4), lightweight (60 kg, n=5), welterweight (66 kg, n=3), and light heavyweight (75 kg, n=9). All participants had competed in at least two regional-level boxing competitions across a minimum of two competitive seasons prior to the study, with 14 (53.8%) having previous national-level competition experience. Because all participants were drawn from a single team, both the experimental and control groups trained together at the same facility under the supervision of the same coaching staff; randomization was performed at the individual level to ensure balanced group composition with respect to weight class and competitive experience.

Participants were randomly assigned to either the experimental group (stroboscopic training, n=13) or the control group (non-stroboscopic training, n=13) using a computer-generated random sequence (Random.org). Randomization was stratified by weight class to ensure balanced representation of weight categories in both groups. Baseline characteristics of the two groups are presented in Table 1, with no significant differences observed between groups in any anthropometric or demographic variable (all p ≥ 0.05). Because all participants belonged to the same team and trained together daily, both groups completed the same regular team training sessions; the stroboscopic glasses were worn exclusively during the designated 35-minute intervention blocks by the experimental group, while the control group wore identical transparent (non-stroboscopic) glasses during the same blocks to control for awareness of group assignment. Participants were instructed not to discuss their assigned training condition with teammates to minimize potential contamination effects. All participants completed three intervention sessions per week (Tuesdays, Thursdays, and Saturdays), ensuring a uniform intervention load across groups.

VariableControl group (n = 13)Stroboscopic group (n = 13)p-valueAge (years)24.54 ± 6.6324.85 ± 4.320.890Body height (cm)171.31 ± 6.97170.31 ± 6.590.710Body mass (kg)66.46 ± 7.5067.69 ± 7.530.680Boxing experience (years)7.85 ± 2.446.54 ± 2.500.190Effective training intervention duration (min/week)31.31 ± 2.3930.38 ± 2.600.356

Baseline descriptive characteristics of the experimental and control groups (mean ± SD).

No significant between-group differences were observed for any variable (all p ≥ 0.05, independent samples t-test). Effective training intervention duration refers to the average weekly minutes of stroboscopic or placebo intervention completed per participant.

2.3 Ethics approval

All procedures were conducted in accordance with the Declaration of Helsinki (World Medical Association, 2013) and were approved by the Ethics Committee of the China Institute of Sport Science (Approval No. CISS20240902). Prior to participation, all participants were provided with a detailed information sheet explaining the study purpose, procedures, potential risks (e.g., mild eye strain from stroboscopic stimulation), and benefits. Written informed consent was obtained from all participants. Participants were also informed that they could withdraw from the study at any time without penalty. Throughout the study, a sports medicine physician was available to address any adverse reactions to the training intervention. No adverse events were reported during the study period.

2.4 National boxing championship schedule clarification

All participants were selected for the 2025 China National Boxing Championship (qualifying tournament for the 15th National Games), organized by the Chinese Boxing Association. The championship comprised three stages: the preliminary qualification round (April 2025, held in Xi’an, Shaanxi Province), the national semi-finals (June 2025), and the national finals (November 2025). Boxing performance data collection was integrated into this competition schedule as follows. The pre-test boxing performance data were collected during the preliminary qualification round in Xi’an, with all bouts completed by 28 April 2025. The six-week training intervention was then carried out (early May to early June 2025). The post-test boxing performance data were collected during a formal intra-team sparring session conducted under official competition rules on 9 June 2025 at Dalian University. The retention test boxing performance data were collected in the same manner on 14 July 2025, four weeks after cessation of training. The use of a formal intra-team bout (rather than the national semi-finals) for post-test and retention test data collection was necessitated by the practical constraint that not all 26 participants advanced to the same subsequent stage of the championship, and by the need to align boxing performance assessments with the laboratory visual testing schedule.

It should be noted that the pre-test boxing performance was assessed in an external official championship context, while post-test and retention test boxing performance were assessed during formal intra-team sparring at the home facility. This difference in competitive context introduces a degree of heterogeneity in opponent strength, psychological pressure, and match intensity across assessment time points, and is acknowledged as a limitation of the study (see Study Limitations). All 26 participants successfully completed all three assessment time points with no dropouts or mandatory medical suspensions.

2.5 Assessment tools and procedures2.5.1 Visual performance skills assessment

Visual performance skills were assessed using the Senaptec Sensory Station (Senaptec Inc., Beaverton, OR, USA), a validated, computerized neurocognitive assessment tool widely used in sports vision research (Millard et al., 2020; Poltavski and Biberdorf, 2015; Knöllner et al., 2022). The reliability of this tool has been confirmed in previous studies, with Erickson et al. (2011) demonstrating good to excellent test-retest reliability for key visual performance metrics, and Wang et al. (2015) validating its structure for measuring perceptual and visual-motor abilities in healthy young adults. The Senaptec Sensory Station consists of a 27-inch high-definition display, a response panel, and specialized software that administers 10 standardized sensorimotor assessments, each designed to evaluate a specific component of visual performance. Table 2 provides a detailed overview of each assessment metric, including its definition, task description, and scoring method.

Test IndicatorsDetailed MethodsEvaluation Criteria
Visual Clarity VC TaskThe participant holds a mobile device and stands 3 meters away from a tablet screen. They judge the direction of the gap in a C-shaped figure displayed on the tablet and swipe the corresponding direction on the mobile device. Monocular vision is tested first for the left and right eyes, followed by binocular vision.The smaller the figure size at which the participant can accurately judge the direction, the better. The test metric is expressed in logMAR units, where lower values indicate better performance. A 5-point scoring system is used, with 5 being the highest score.
Contrast Sensitivity CS TaskThe participant holds a mobile device and stands 3 meters from the tablet. Four black circles appear on the tablet screen; one circle, randomly oriented, contains concentric circles of varying shades. The participant must identify this circle and swipe in the corresponding direction on the mobile device.As accuracy improves, the contrast within the concentric circles becomes less distinct. The test metric is measured in logCS (log Contrast Sensitivity), with higher values indicating better sensitivity.
Depth perception DP taskStanding 3 meters from the tablet, the participant sees four black circles on the screen; one randomly displays a stereoscopic effect. The participant must find this circle and swipe in the corresponding direction on the mobile device. Binocular vision is tested first, followed by separate tests for the right and left eyes.As accuracy increases, the contrast and stereoscopic effect of the target circle become less pronounced. The test metric is in arcseconds (arcsec), where smaller values indicate better stereoscopic vision.
Near-far Fast N/FQ MissionThe participant stands 3 meters from the tablet, holding the mobile device so that its top is 40 cm below the bottom of the tablet screen. During the test, C-shaped figures alternate between the tablet (far) and mobile device (near). The athlete switches focus between near and far every 30 seconds to judge the gap direction, then quickly swipes in the corresponding direction on the mobile device.Faster judgment speed and higher directional accuracy are preferable. Test metrics include the number of swipes within 30 seconds (higher is better) and reaction times for near and far focus shifts measured in milliseconds (ms), where lower values indicate faster responses.
Target acquisition TC missionThe participant holds a mobile device and stands 3 meters from the screen, aligning the blue reference line in the center of the screen with their line of sight, focusing on the center point. C-shaped figures randomly appear in the four corners of the screen; the participant judges the gap direction and swipes the corresponding direction on the mobile device.Reaction times measured in milliseconds (ms); faster speeds correspond to better performance.
Perception span PS taskThe participant stands 60 cm from the tablet screen, with eyes level to the center of the screen. Radial circles emanate from the center, some of which briefly flash black dots at their centers. The athlete must identify and tap the circle containing the black dot.The number of circles and black dots increases continuously across a wider range. Scores are based on the cumulative number of correct identifications, with higher scores indicating better performance.
Multi-target tracking MOT missionThe participant stands 60 cm from the tablet screen, eyes level with the screen center. Several groups of spheres appear, each containing two black spheres. One sphere briefly changes to red then quickly returns to black before rotating randomly clockwise or counterclockwise. After rotation stops, the participant must identify the sphere that initially turned red in each group.Test metrics include the number of correctly tracked targets, tracking speed (degrees per second, °/s), percentage-based scores, and composite scores. Higher values indicate better performance.
Response Time RT TaskThe participant stands 60 cm from the tablet screen, eyes level with the screen center. Radial circles emanate from the center; some flash black dots briefly. The participant must identify and tap the circles containing the black dots.Similar to above, with increasing numbers and range of circles and black dots, scoring is based on cumulative correct judgments; higher scores indicate better performance.
Hand-eye coordination EHC taskThe athlete stands 60 cm from a large screen, raising the screen’s midline to align with or slightly above the arms to avoid obstructing peripheral vision. The screen displays 8 columns by 10 rows of hollow rings. One ring randomly changes to blue-green, and after the participant clicks it, another ring appears at a random position. The goal is to click as many as possible quickly and accurately within the allotted time.Metrics include total time, average reaction time, central region reaction time, and peripheral region reaction time all measured in milliseconds (ms), with lower values indicating better performance. 
Go/No Go(GNG) taskThe athlete stands 60 cm from the large screen, with the screen midline aligned at or slightly above arm level to prevent interference with peripheral target tapping speed. The screen shows 8 columns of circles identical to those in the eye-hand coordination test. Green and red dots appear randomly on the circles; green dots require a quick tap, while red dots must not be tapped.Metrics include overall score, number of correct taps (higher is better), and number of incorrect taps (lower is better).

Detailed description of motor visual ability test.

Prior to each assessment session (pre-test, post-test, retention test), participants completed a 5-minute standardized warm-up consisting of eye-movement exercises (saccades, smooth pursuits) and light stretching to prepare the visual and neuromuscular systems. Assessments were administered in a fixed order to ensure consistency across time points: (1) Visual Clarity (VC), (2) Contrast Sensitivity (CS), (3) Depth Perception (DP), (4) Target Capture (TC), (5) Near-Far Quickness (NFQ), (6) Reaction Time (RT), (7) Eye-Hand Coordination (EHC), (8) Go/No-Go (GNG), (9) Perception Span (PS), (10) Multiple Object Tracking (MOT). Each task began with an animated demonstration (2 minutes) followed by three practice trials to ensure participants understood the task requirements. Formal testing consisted of 10–15 trials per task (depending on the metric), with a 30-second rest period between tasks to minimize fatigue. Total assessment time was approximately 25 minutes per participant.

2.5.2 Boxing performance analysis

Boxing performance was assessed using video recordings of participants’ official bouts at the 2025 National Boxing Championship. Video footage was captured using two Sony FDR-AX700 cameras (1080p resolution, 60 frames per second) positioned at ring-side (45° angle relative to the ring, 5 meters from the ring apron) to ensure full visibility of both participants and the ring area. Cameras were synchronized to capture simultaneous front and side views of the bouts, enabling precise analysis of punch trajectory and landing location.

Video analysis was conducted using Dartfish ProSuite 12.0 software (Dartfish SA, Fribourg, Switzerland), a specialized sports performance analysis tool widely used in boxing research (Thomson et al., 2013; Davis et al., 2013, 2015). All videos were analyzed by a single trained boxing performance analyst (with 5 years of experience in boxing video analysis) who was blinded to group assignment. The analyst reviewed each bout frame-by-frame (0.1-second increments) to count three key variables: (1) total punches thrown (all legal punches attempted, including jabs, crosses, hooks, and uppercuts), (2) successful (scoring) punches (legal punches that landed on valid target areas: head, torso above the waist, and shoulders), (3) missed punches (legal punches that did not land on the opponent or landed on invalid target areas: back, neck, legs, or arms). Punch accuracy was calculated as: %Hit = (number of successful punches/total number of punches thrown) × 100.

To ensure reliability, intra-rater and inter-rater reliability analyses were conducted. For intra-rater reliability, the analyst re-analyzed 5 randomly selected bouts (19.2% of total bouts) 2 weeks after the initial analysis. For inter-rater reliability, a second experienced boxing analyst (with 7 years of experience) independently analyzed the same 5 bouts. Intra-class correlation coefficients (ICC) for punch accuracy were 0.92 (95% CI: 0.87–0.96) for intra-rater reliability and 0.89 (95% CI: 0.83–0.94) for inter-rater reliability, indicating excellent reliability for the measurement protocol.

2.6 Training protocol

Both the experimental and control groups completed the same 6-week training program, consisting of three weekly sessions (Tuesdays, Thursdays, Saturdays) with each session lasting 35 minutes. Training was integrated into the participants’ existing weekly training programs (no additional training load was added) to ensure ecological validity and minimize disruption to their competition preparation. All training sessions were conducted at the Beijing Boxing Team’s home training facility (Shichahai Sports School, Beijing), under the supervision of a certified boxing coach who was trained on the study protocol.

The key difference between groups was the visual condition during training: the experimental group trained under stroboscopic conditions using Senaptec Strobe glasses (Senaptec Inc., Beaverton, OR, USA), while the control group trained under normal visual conditions using identical non-stroboscopic glasses (transparent lenses) to control for placebo effects (i.e., participants were unaware of their group assignment). To assess the strength of the placebo effect, participants were asked post-intervention whether they believed they were in the experimental group (stroboscopic training) or the control group (non-stroboscopic training). No significant differences were observed in the accuracy of group assignment between the two groups (χ²=0.31, p=0.578), indicating that the placebo control was effective. The Senaptec Strobe glasses are lightweight, dual-lens glasses that alternate between transparent and opaque states via Bluetooth control using the Senaptec Strobe smartphone application. The glasses have been validated for use in sports training and have been shown to produce consistent stroboscopic effects (Zwierko et al., 2023; Hülsdünker et al., 2021).

2.6.1 Stroboscopic training progression

The stroboscopic training protocol was adapted from Hülsdünker et al. (2021), with progressive adjustments to strobe frequency (Hz) and duty cycle (the ratio of lens open time to closed time) to gradually increase task difficulty over the 6-week period. Strobe frequency refers to the number of times the lenses alternate between transparent and opaque per second (lower frequency = longer periods of occlusion = greater difficulty), while duty cycle refers to the percentage of time the lenses are transparent during each cycle (higher duty cycle = longer transparent periods = lower difficulty). The progression of strobe settings is presented in Table 3.

WeekSessions/weekStrobe frequency (Hz)Duty cycle (% open)Block structureTotal strobe exposure/session1310702.5 min on / 2.5 min off × 717.5 min239652.5 min on / 2.5 min off × 717.5 min338602.5 min on / 2.5 min off × 717.5 min437552.5 min on / 2.5 min off × 717.5 min536502.5 min on / 2.5 min off × 717.5 min635502.5 min on / 2.5 min off × 717.5 min

Stroboscopic training progression across the 6-week intervention.

Strobe frequency decreased weekly (10 Hz → 5 Hz), progressively increasing occlusion difficulty. Duty cycle decreased weekly (70% → 50%), reducing lens-open time. Both groups completed the same training drills; only the experimental group wore stroboscopic glasses.

To minimize fatigue and adaptation, strobe exposure was limited to 2.5-minute intervals, separated by 2.5-minute rest periods (during which participants removed the glasses and performed light active recovery exercises, e.g., jogging in place, arm circles). This interval training approach is consistent with previous stroboscopic training studies (Zwierko et al., 2023; Hülsdünker et al., 2021) and ensures that participants maintain optimal focus and effort during each training block.

2.6.2 Training components

The training program consisted of three sequentially performed components, designed to progressively build from general visual reaction skills to boxing-specific visuomotor integration. Each component was 10–15 minutes in duration, with a 2-minute transition period between components. Specific illustrations of the exercises are shown in Figure 1.

A graphical illustration of the study protocols with examples of the exercises. (A) Program A: General Visual Reaction Drills (tennis ball reaction training). (B) Program B: FITLIGHT Reaction Drills (light-based reaction training). (C) Program C: Boxing-Specific Reactive Drills (precision punching, fast reaction pad drills, and integrated defense-counter drills).

1. Program A: General Visual Reaction Drills (10 minutes).

This component focused on improving spatial awareness, rhythm control, and basic hand-eye coordination, using 6.7 cm diameter tennis balls that require rapid visual processing. Two drills were performed (5 minutes each):

Drill 1: Self-directed Toss and Catch. The participant tossed tennis balls against a flat wall and caught them while visually tracking the ball’s trajectory. This drill was designed to enhance visual tracking ability, basic hand-eye coordination, spatial awareness, and rhythm control. Each set consisted of 15 throws, with 30 seconds of rest between sets.

Drill 2: Catching Tossed Balls. Research staff stood 3–5 meters in front of the participant and tossed tennis balls toward them. The participant’s task was to predict the landing point, visually track the ball, and catch it. This drill targeted dynamic visual acuity, targeting accuracy, anticipation, and simple reaction speed. Each set consisted of 10 tosses, with 30 seconds of rest between sets.

2. Program B: FITLIGHT Reaction Drills (10 minutes).

This component used the FITLIGHT system (FITLIGHT Sports Technology Inc., Edmonton, Canada), consisting of at least 8 wireless, programmable light panels. The system was used to improve visual attention, focus, and dynamic visual-motor coordination. Two drills were performed (5 minutes each):

Drill 1: Basic Light Reaction. Eight light panels were arranged around the participant in a dynamic training area. The panels were programmed to illuminate randomly, and the participant’s task was to quickly touch any randomly lit FITLIGHT device as quickly as possible. This drill aimed to improve simple reaction speed, initiation speed, and visual attention. Each set consisted of 20 light activations, with 30 seconds of rest between sets.