Cognitive Load and Motor Adjustment Under Virtual Defensive Pressure

in Mixed Reality Sports Training

Hayato Saiki

, Kiyohiro Konno

, Kosuke Naruse

, Takahiro Shimizu

, Yasuhiro Suzuki

Seiji Ono

and Kenji Suzuki

Artiﬁcial Intelligence Laboratory, University of Tsukuba, Tsukuba, Japan

Doctoral Program in Physical Education, University of Tsukuba, Tsukuba, Japan

Master’s Program in Engineering Science, University of Tsukuba, Tsukuba, Japan

Doctoral Program in Psychology, University of Tsukuba, Tsukuba, Japan

Institute of Health and Sport Sciences, University of Tsukuba, Tsukuba, Japan

Keywords:

Mixed Reality, Defensive Pressure, Virtual Avatars, Basketball Training, Cognitive Load, Motor Behavior.

Abstract:

Defensive pressure experienced during competition exerts a signiﬁcant impact on athletes’ psychological states

and motor behaviors; however, reproducing such pressure in training environments remains a persistent chal-

lenge. This study developed a Mixed Reality (MR)-based training system that enables athletes to experience

realistic defensive pressure during actual physical movements, using virtual avatars that perform closeout

actions. By integrating motion capture technology with an optical see-through head-mounted display, we

constructed an environment where the perceived intensity of defensive pressure could be modulated by ma-

nipulating avatar height. Experiments involving nine skilled male basketball players compared three defensive

conditions—Free (no defender), Human defender, and Virtual defender (180 cm)—and further examined the

effects of avatar height variations (160 cm vs. 200 cm). Results indicated that virtual defenders, despite the

absence of physical contact, induced psychological load comparable to that of human defenders, as measured

by NASA-TLX scores and ball-holding duration. Notably, taller avatars elicited greater perceived pressure and

accelerated shot preparation, although no signiﬁcant differences in shooting accuracy were observed. These

ﬁndings suggest that MR-based training systems offer an effective means to systematically scale perceived

difﬁculty without compromising performance. Furthermore, the observed dissociation between subjective

workload and ﬁnal outcomes highlights the importance of incorporating multisensory feedback to enhance

ecological validity in future system designs.

1 INTRODUCTION

In athletic performance, the execution of motor skills

is closely related to the pressure perceived by ath-

letes. This pressure can take various forms, in-

cluding social evaluative stress caused by environ-

mental factors such as attention from spectators and

media, as well as spatial and physical constraints

uniquely imposed by defenders in team sports(Endo

et al., 2023). Among the various forms of pres-

sure, particular attention has been paid to spatial pres-

sure from defenders, which directly affects athletes’

physical control. In fact, numerous studies have

demonstrated that the performance of offensive ath-

letes is signiﬁcantly inﬂuenced by the pressure ex-

erted by opposing defenders, based on data collected

during actual matches across various sports(Sampaio

et al., 2016; Leander et al., 2024; Grifﬁn et al.,

2017). Traditionally, interventions to prevent the

decline of executive functions under psychological

pressure—particularly stress caused by environmen-

tal factors—have relied on Virtual Reality Exposure

Therapy (VRET)(Gerardi et al., 2010; Gerardi et al.,

2020). This method aims to reduce real-world stress

responses by allowing individuals to experience an-

ticipated stress-inducing scenarios in a virtual envi-

ronment beforehand. However, in sports contexts, the

ability to control physical movements under pressure

is critically important, and psychological habituation

alone offers limited effectiveness. Because VR en-

vironments make it difﬁcult for athletes to engage

in training that involves actual body movements, in-

Saiki, H., Konno, K., Naruse, K., Shimizu, T., Suzuki, Y., Ono, S. and Suzuki, K.

Cognitive Load and Motor Adjustment Under Virtual Defensive Pressure in Mixed Reality Sports Training.

DOI: 10.5220/0013661700003988

Paper published under CC license (CC BY-NC-ND 4.0)

In Proceedings of the 13th International Conference on Spor t Sciences Research and Technology Support (icSPORTS 2025), pages 13-24

ISBN: 978-989-758-771-9; ISSN: 2184-3201

Ball Position (x, y, z)

Virtual Defender

Motion Capture

Collision Detection

HMD

Appearance of Virtual Defender

a b

Avatar generated

from face photo

Figure 1: Overview of the proposed system and virtual defender. (a) Real-time motion capture and collision detection for

simulating defensive pressure in MR. (b) Virtual defender appearance generated from facial photo with adjustable body

attributes.

creasing attention has been given in recent years to

reproducing realistic match scenarios using AR (Aug-

mented Reality) and MR (Mixed Reality) technolo-

gies(Cheng et al., 2024; Wen et al., 2024). Never-

theless, very few studies have examined the psycho-

logical and physical effects that occur when athletes

perform real-world movements under pressure from

virtual defenders.

In this study, defensive pressure is deﬁned as the

actions taken by defenders to apply spatial pressure

on offensive players to restrict their movements and

prevent them from scoring. The role of a defender

is to read the opponent’s movements, quickly move

to the appropriate position, and block the opponent’s

passes or shots. The stronger the defensive pressure,

the more difﬁcult it becomes for the offensive athletes

to fully utilize their motor skills. To counteract de-

fensive pressure, athletes need to undergo appropriate

training(Sarah et al., 2020). It is important to reﬁne

movements and decision-making under pressure by

training while actually experiencing defensive pres-

sure through scenarios that may occur in matches.

Traditional training methods that athletes have

adopted to improve performance under pressure can

be broadly categorized into individual practice and

group practice. In individual practice, athletes of-

ten recreate critical in-game scenarios through mental

imagery or practice using static objects that simulate

defenders(Lindsay et al., 2023). In contrast, group

practice involves training with teammates or partners,

enabling the development of practical skills such as

decision-making and shooting under defensive pres-

sure(Wellington et al., 2023; Seyﬁ et al., 2018). How-

ever, both training approaches present certain limita-

tions. In individual practice, the recreation of pres-

sure situations heavily relies on the athlete’s imagi-

native ability, while group practice is constrained in

terms of the intensity and quality of defensive pres-

sure it can realistically simulate. In actual matches,

athletes often face defenders with greater physical

size and ability than their teammates, making it dif-

ﬁcult for conventional group training to reproduce the

high-intensity pressure that they may have never ex-

perienced before. Therefore, there is a growing need

for new training methods that allow athletes to phys-

ically experience defensive pressure in real time and

ﬂexibly adjust its intensity.

2 RELATED WORKS

2.1 Virtual Reality Exposure Therapy

In sports, performing exercises under defensive pres-

sure from the practice stage to become accustomed

to the defensive pressure expected in games is a

commonly used training method. This shares sim-

ilarities with exposure therapy, which is primarily

used to treat anxiety disorders, phobias, and PTSD

(Post-Traumatic Stress Disorder) (Richards and Rose,

1991). As a development of exposure therapy, many

studies have been conducted on VRET (Virtual Real-

ity Exposure Therapy), a treatment method that aims

to recreate anxiety-inducing situations in a VR envi-

ronment and gradually accustom individuals to these

situations (Gerardi et al., 2010; Gerardi et al., 2020).

Freeman et al. investigated the effectiveness of au-

tomatic cognitive intervention for acrophobia guided

by an avatar virtual coach. The results indicated that

participants in the VR treatment group experienced a

signiﬁcant reduction in acrophobia at the end of the

treatment and during follow-up, suggesting that auto-

mated psychotherapy by a VR coach has effects equal

to or greater than traditional treatment methods (Free-

man et al., 2018). In this way, using VR for treat-

ment to alleviate speciﬁc anxieties and fears has been

icSPORTS 2025 - 13th International Conference on Sport Sciences Research and Technology Support

proven effective in numerous studies. Therefore, if

external pressure can be deliberately reproduced in a

virtual environment and users can be gradually accli-

mated to it, the approach could be applied as a form

of training aimed at enhancing resilience to external

stressors encountered during actual competition.

2.2 Presence of Virtual Avatars

To effectively convey the spatial pressure of defense,

it is necessary to consider the sense of presence peo-

ple feel towards virtual avatars. The sense of presence

in virtual avatars has been extensively studied, par-

ticularly in the ﬁeld of social communication(Tianqi

et al., 2024; Yoon et al., 2016; Christos and Michael-

Grigoriou, 2022). In the study by Yoon et al., the

impact of avatar body part visibility and character

style on social presence was compared. The results

showed that realistic full-body avatars generated a

higher sense of social presence than any other con-

ditions(Yoon et al., 2019). Furthermore, Yoon et al.

also investigated the differences in presence caused

by varying levels of avatar transparency(Yoon et al.,

2023). Their study revealed that as transparency in-

creased, social presence decreased. Based on these

ﬁndings, it is suggested that to make people feel a

strong sense of presence towards AR-displayed de-

fense, it is necessary to use avatars with a realistic ap-

pearance and full-body visibility within the system,

and to employ devices capable of projecting AR ob-

jects with low transparency.

One phenomenon to be cautious of when using

realistic avatars is the Uncanny Valley(Mori, 1970).

Previous research has shown that when virtual avatars

are made to look realistic but exhibit unnatural move-

ments, it can cause users of the system to feel a sense

of eeriness(Angela et al., 2010). Therefore, in this

study, to avoid unintentionally causing discomfort to

athletes using the system, careful attention must be

paid to the movements of the avatars.

2.3 Sports Training with Virtual Avatar

Recent advances in immersive technologies have

drawn attention to the use of virtual agents in tacti-

cal training for team sports. Cheng et al. developed

a Mixed Reality (MR) system designed to bridge the

gap between tactical understanding and execution in

basketball(Cheng et al., 2024). By enabling athletes

to coordinate and confront virtual teammates and op-

ponents, the system enhances spatial and situational

awareness, allowing for repeated practice of complex

tactics that are often difﬁcult to achieve in conven-

tional training environments.

The application of virtual avatars has also pro-

gressed in the context of coaching. Wen et al. pro-

posed the Augmented Coach, which integrates 3D

volumetric video and spatial annotations to deliver

real-time visual feedback on posture and timing(Wen

et al., 2024). This approach enables personalized re-

mote coaching by projecting expert guidance directly

into the athlete’s environment.

In addition, studies have explored the psychologi-

cal dimensions of immersive sports training. Stinson

used a CAVE-based VR system to recreate penalty

kick scenarios in soccer and investigated athletes’

psychological responses under environmental pres-

sure(Stinson and Bowman, 2014). Their ﬁndings

demonstrated that both physiological and subjective

stress responses were signiﬁcantly induced even in

virtual settings, suggesting the potential of VR-based

resilience training.

While these studies have contributed valuable in-

sights into environmental simulation, remote instruc-

tion, and psychological stress, there remains a lack

of research that reproduces and modulates interper-

sonal pressure—speciﬁcally, the spatial and tactical

pressure exerted by opposing players—in virtual en-

vironments. Furthermore, many existing studies tar-

get novice participants or lack realistic situational se-

tups, highlighting the need for training systems that

more closely replicate actual competition settings.

This study aims to design and evaluate a training

environment grounded in real-game contexts by intro-

ducing virtual defenders into MR spaces that impose

interpersonal pressure. Through both quantitative and

qualitative analysis, we focus on the cognitive and

behavioral demands induced by the presence of op-

ponents—an aspect insufﬁciently addressed in prior

work—and propose a novel framework for sport-

speciﬁc MR training support.

3 PURPOSE

The objective of this study is to investigate how visu-

ally represented defensive pressure within a mixed re-

ality (MR) environment affects athletes’ preparatory

behavior and perceived cognitive load during actual

physical movement.

We hypothesize that realistic defensive actions

performed by virtual avatars can induce a level of

pressure comparable to that triggered by real human

defenders. Accordingly, this study examines how

skilled athletes respond—both cognitively and behav-

iorally—when facing defensive pressure from either

a human or a virtual MR-based defender. Further-

more, we assess whether manipulating simple visual

Cognitive Load and Motor Adjustment Under Virtual Defensive Pressure in Mixed Reality Sports Training

features of the virtual defender (e.g., height) can sys-

tematically modulate athletes’ perceived threat and

preparatory behavior.

The contributions of this study are fourfold:

• Empirical validation that virtual defend-

ers can induce cognitive pressure re-

sponses—demonstrating perceived workload

levels comparable to those induced by real human

defenders.

• Behavioral analysis of preparatory actions un-

der different defender types—revealing qualita-

tive differences in how athletes manage perceived

pressure when facing human versus virtual oppo-

nents.

• Demonstration of scalable pressure modula-

tion—showing that simple visual manipulations,

such as avatar height, can effectively tune per-

ceived defensive pressure.

• Proposal of a methodological frame-

work—enabling the investigation of dissoci-

ation between perception (subjective pressure),

preparation (ball-holding duration), and perfor-

mance (shooting accuracy) in ecologically valid

MR-based sports scenarios.

4 PROPOSED METHOD

4.1 Generation of Defensive Pressure by

Virtual Player

In this system, a virtual player with collision detec-

tion for the ball performs defensive actions against

a user wearing a see-through HMD (Head-Mounted

Display), thereby inducing pressure on the user. Fig-

ure 1 shows an overview of the proposed method.

The virtual player performs defensive actions against

a user performing speciﬁc offensive actions. The suc-

cess of the virtual player’s defensive interference is

calculated based on the real-time measured position

of the ball and the position of the virtual player in the

system. When the ball makes contact with the virtual

player, immediate feedback is provided to the user.

This requires the user to recognize the virtual player

as a defender, similar to a real human, and move in

a way that prevents their offensive actions from be-

ing obstructed. By implementing collision detection

for the virtual player and minimizing the differences

from the real world, the system allows the user to feel

pressure from the virtual player. This system is not

limited to speciﬁc scenes or sports and can be applied

to various training scenarios by intentionally inducing

pressure in interpersonal sports.

4.2 Collision Detection Between the Ball

and the Virtual Player

In this system, the moment when the constantly

tracked ball and the virtual player in the system come

into contact is detected. The entire surface of the vir-

tual player’s body is speciﬁed as a range capable of

detecting contact with any object, and based on the

position and size of the ball, feedback can be pro-

vided to the user. The physical characteristics and

movement speed of the virtual player can be arbitrar-

ily modiﬁed, allowing for the adjustment of pressure.

For instance, by increasing the height of the virtual

player to simulate a real-world match, or by setting a

higher movement speed to enhance the agility of the

defense, the pressure can be adjusted accordingly.

5 SYSTEM CONFIGURATION

5.1 Virtual Player

In this study, it is necessary to pay attention to the

physical characteristics of the avatars to investigate

the differences in pressure exerted by virtual players

versus real human defenders. If the appearance of the

avatars signiﬁcantly differs from that of the human de-

fenders, the players taking the shots might perceive

the defense differently, potentially affecting the eval-

uation of the pressure (Wang et al., 2013). Therefore,

in this study, we used an AI called Avaturn, which

can create 3D avatars from photographs, to generate

3D models of the human defenders for use in the ex-

periments (Avaturn, 2025). The height of the avatars

was adjusted on the platform used to create the sys-

tem to match that of the human defenders. Addition-

ally, the defensive movements are a critical aspect that

greatly affects the pressure. Therefore, during the ex-

periments, we measured the movements of the human

defenders using Optitrack motion capture system and

conﬁgured the system to minimize any discrepancies.

5.2 Motion Capture System

The position data of the ball was measured using the

Optitrack motion capture system. The motion cap-

ture cameras were set to a sampling rate of 120 Hz.

An octagonal basketball (Wilson) was used in the ex-

periment, with a total of 42 markers placed along the

edges of each face. Additionally, during the exper-

iments, participants wore motion capture suits (Op-

titrack) to measure their body movements. A skele-

ton model was created for each participant, allowing

icSPORTS 2025 - 13th International Conference on Sport Sciences Research and Technology Support

the system to automatically estimate the position of

occluded markers from other visible markers, even

when markers were obscured by the ball or the par-

ticipant’s body.

5.3 Head Mounted Display

For dynamic sports like basketball, a standalone

HMD is desirable. Considering the importance of

weight and ﬁeld of view (FOV) for physical activities,

the Magic Leap 2 (Magic Leap) was used. This de-

vice is lightweight at 260g and provides an adequate

FOV of 70 degrees (horizontal: 45 degrees, vertical:

55 degrees), ensuring sufﬁcient visibility for move-

ment. Additionally, as an optical see-through HMD,

it minimizes latency and discomfort when interacting

with the real world. The device also features a dim-

ming function for AR content, allowing it to display

immersive content to the user without transparency,

maintaining a sense of immersion (Yoon et al., 2023).

5.4 Large Space

In this study, the experiments were conducted using

”LargeSpace,” a large-scale immersive display owned

by the University of Tsukuba(Takatori et al., 2016).

”LargeSpace” consists of 12 projectors, 10 comput-

ers, and 20 motion capture cameras, allowing motion

capture measurements in a spacious area of 25 me-

ters in width, 7.7 meters in height, and 15 meters in

depth. In this space, a half-court basketball court was

created by laying polyline tape (molten) according to

the speciﬁcations set by the FIBA (International Bas-

ketball Federation). Additionally, a freestanding bas-

ketball hoop (Spalding) was installed within the same

space.

6 EXPERIMENT

6.1 Experimental Design

6.1.1 Participant and Collaborator

In this study, there were two types of individuals in-

volved: experiment participants and an experiment

collaborator. The experiment participants refer to the

nine individuals who were measured in this study. All

participants were male (age: M = 21.0, SD = 2.05;

height: M = 173.5 cm, SD = 2.89; basketball expe-

rience: M = 11.8 years, SD = 2.04). To control for

factors that could affect pressure evaluation, partic-

ipants were required to have at least seven years of

basketball experience under the guidance of a coach

Passer Shooter

Defender

4.0 m

3.5 ~ 4.5 m

Figure 2: Shooting task setup with variable defensive pres-

sure distances.

and a height between 170 cm and 180 cm. The experi-

ment collaborator refers to the individual who played

the role of the defender. All participants faced the

same defender (male, age = 23, height = 180 cm,

basketball experience = 14 years). This experiment

was approved by the institutional ethics review board

(Approval ID: 2024R892). All participants provided

written informed consent prior to participation and

received compensation after completing the experi-

ment.

6.1.2 Target Offensive Action

In this experiment, we evaluate the pressure felt by

participants during shooting when faced with defense.

Speciﬁcally, participants perform three-point shots in

a scenario called ”closeout” (Hopla, 2012). A ”close-

out” refers to a situation where the defender runs to-

wards the offensive player who has just caught the

ball, requiring the offensive player to execute a quick

shooting motion compared to a free state without any

defense. The actual experimental setting is shown

in Figure 2. In the condition with defense, an au-

dible signal is given before each shot, and simulta-

neously, the experimenter passes the ball to the par-

ticipant. At the same time, the defensive collabo-

rator starts running from a designated position and

attempts to block the shot, applying pressure on the

shooting participant. To prevent the shooting partici-

pant from becoming accustomed to the defender’s ac-

tions, the defensive collaborator randomly starts from

three different positions with varying distances. The

running distances are Hard: 3.5m, Medium: 4m, and

Easy: 4.5m from the shooter. To ensure consistent

running speeds, the defensive collaborator practiced

the closeout 30 times from each position prior to the

experiment and subsequently measured the time for

15 trials from the signal to reaching 1m in front of

the three-point line. The results showed that for the

Hard distance, the average time was 1.2522 s with a

standard deviation of 0.0864 s; for the Medium dis-

Cognitive Load and Motor Adjustment Under Virtual Defensive Pressure in Mixed Reality Sports Training

tance, the average time was 1.4748 s with a standard

deviation of 0.1062 s; and for the Easy distance, the

average time was 1.6042 s with a standard deviation

of 0.0775 s. For the virtual player’s closeout action,

we applied the animation that most closely matched

the average running time for each difﬁculty level and

set it to randomly vary the speed within the standard

deviation. Lastly, in the condition without defense,

an audible signal is given before each shot, the ex-

perimenter passes the ball to the participant, and the

participant takes the shot.

6.1.3 Evaluation Criteria

Across the two experiments conducted in this study,

we evaluated three primary aspects to assess the psy-

chological and behavioral effects of defensive pres-

sure: ball-holding duration, NASA-TLX, and shoot-

ing performance.

Ball-holding duration was used as a behavioral

indicator of hesitation or caution under pressure. It

was computed based on the 3D distance between the

ball marker and the nearest hand marker. Speciﬁcally,

holding was deﬁned as the period starting when the

ball ﬁrst entered within 24.5 cm of the hand marker

and ending when it ﬁrst moved beyond 24.5 cm. The

threshold of 24.5 cm was chosen to correspond to the

ofﬁcial diameter of a basketball. Longer holding du-

rations were interpreted as increased cognitive load or

decision-making latency, particularly in the presence

of a defender.

NASA-TLX was used to assess subjective men-

tal workload. After each condition, participants com-

pleted the validated Japanese version of the NASA-

TLX(Hart and Staveland, 1988; Haga and Mizukami,

1996), which consists of six subscales: mental de-

mand, physical demand, temporal demand, perfor-

mance, effort, and frustration. Each item was rated on

a scale from 0 to 100. To calculate the overall work-

load score, we used the original weighted NASA-

TLX method, in which participants ﬁrst performed

pairwise comparisons to assign relative weights to the

subscales. The ﬁnal score was then computed as a

weighted average of the six subscale ratings. This

measure captured participants’ internal states, such as

mental demand and frustration, which are not always

externally observable through behavior.

Shooting performance served as an outcome-

based measure reﬂecting task success. Each shot was

scored using a 4-point system based on established

methods in prior studies(Fazel et al., 2018). A score

of 3 was given for a successful shot that did not touch

the rim, 2 for a successful shot that hit the rim, 1 for

a missed shot that hit the rim, and 0 for a missed shot

that did not touch the rim. The total score across all

shooting attempts in each condition was used as the

ﬁnal performance measure.

Together, these three indicators—behavioral de-

lay, perceived mental load, and motor perfor-

mance—were used to comprehensively assess the in-

ﬂuence of defensive pressure under both real and vir-

tual conditions. Comparisons were made not only

between defense and no-defense conditions, but also

between human and virtual defenders, allowing us to

examine whether virtual agents can elicit responses

similar to real human opponents.

6.2 Pressure Comparison Between

Human and Virtual Defenders

The objective of this experiment was to evaluate the

differences in psychological and behavioral pressure

exerted by human and virtual defenders during shoot-

ing actions. To this end, we conducted an experiment

with nine skilled basketball players under three differ-

ent conditions.

6.2.1 Procedure

Prior to the start of the experiment, participants com-

pleted 30 practice three-point shots. They were then

instructed on how to use the mixed reality (MR) sys-

tem and were shown how the ”Blocked” feedback

would be displayed when a shot was blocked. There-

after, following the experimental setup described in

Section 6.1, each participant performed ﬁfteen three-

point shots under the following three conditions:

• Free: No defender was present. Participants were

allowed to shoot without any obstruction.

• Human: A real human defender with a height of

180 cm performed a closeout to contest the shot.

• Virtual (V-180): A virtual defender with the

same height, appearance, and motion as the Hu-

man condition contested the shot within a mixed

reality environment.

The order of the three conditions was randomized

for each participant to control for order effects.

Data corresponding to the three evaluation

criteria—ball-holding duration, subjective workload

(NASA-TLX), and shooting performance—were col-

lected and analyzed after each condition. These data

served as the basis for assessing the psychological and

behavioral effects of defensive pressure.

6.2.2 Result

The results demonstrated clear differences in both

subjective and behavioral pressure across the three

icSPORTS 2025 - 13th International Conference on Sport Sciences Research and Technology Support

conditions (Free, Human, V-180). These results are

visualized in Fig.4, and detailed data for the NASA-

TLX subscales are summarized in Table1.

Ball-Holding Duration: A Friedman test re-

vealed a signiﬁcant effect of defender type on ball-

holding duration (χ

(2) = 9.56, p = 0.0084). Partic-

ipants held the ball signiﬁcantly longer in the Free

condition (1.08 ± 0.13 s) than in the Human (0.88

± 0.07 s; p = 0.0078) and Virtual (0.98 ± 0.11 s;

p = 0.0273) conditions. The Human condition also

resulted in shorter durations than the Virtual condi-

tion (p = 0.0117).

Subjective Mental Workload (NASA-TLX):

NASA-TLX scores differed signiﬁcantly among the

three conditions (χ

(2) = 14.00, p = 0.0009). The

Free condition (37.56 ± 15.30) was rated signiﬁcantly

lower in mental workload than both the Human (63.44

± 15.79; p = 0.0039) and Virtual (58.26 ± 16.16;

p = 0.0039) conditions. No signiﬁcant difference

was found between the Human and Virtual conditions

(p = 0.3008), suggesting comparable perceived pres-

sure from both.

Shooting Score: No signiﬁcant difference in

shooting performance was found among the condi-

tions (χ

(2) = 3.00, p = 0.2231). The average scores

were 3.81 ± 0.50 for Free, 3.42 ± 0.69 for Human,

and 3.40 ± 0.49 for Virtual. Although a Wilcoxon test

revealed a marginally signiﬁcant difference between

Free and Virtual (p = 0.0273), no difference was ob-

served between Free and Human (p = 0.1275) or Hu-

man and Virtual (p = 0.7991).

2.0 m

1.6 m

Figure 3: Virtual Defender Height Variations (160 cm vs.

200 cm).

6.3 Differences in Defensive Pressure by

Adjusting Physical Features

In the second experiment, the effect of a vir-

tual defender’s physical characteristics—speciﬁcally

height—on perceived and behavioral pressure was ex-

amined.

6.3.1 Procedure

The procedure was identical to that of the ﬁrst exper-

iment, except for the defender height conditions. As

illustrated in Figure 3, each participant performed ﬁf-

teen three-point shots under the following two condi-

tions:

• V-160: A virtual defender with a height of 160 cm

performed a closeout block motion in the mixed

reality environment.

• V-200: A virtual defender with a height of 200 cm

performed the same motion in the same environ-

ment.

All other aspects of the procedure, including

the randomized order of conditions, motion cap-

ture recording, extraction of ball-holding duration,

NASA-TLX evaluation, and 4-point shot scoring

method, were consistent with the ﬁrst experiment.

6.3.2 Result

The results of this experiment are visualized in Fig.4,

and detailed data for the NASA-TLX subscales are

summarized in Table1. Taller or human defenders

were associated with higher perceived pressure and

shorter preparation time, but did not lead to dif-

ferences in actual scoring performance. In addi-

tion to the within-experiment comparison between V-

160 and V-200, we conducted an exploratory cross-

condition analysis involving Human, V-160, V-180,

and V-200, as all participants experienced these con-

ditions under a consistent protocol.

Ball-Holding Duration: A Friedman test showed

a signiﬁcant effect of defender type on ball-holding

duration (χ

(3) = 13.40, p = 0.0038). Wilcoxon

tests revealed that participants held the ball signif-

icantly longer in V-160 (0.97 ± 0.11 s) than in V-

200 (0.85 ± 0.11 s; p = 0.0039), and also longer in

V-160 compared to Human (p = 0.0078) and V-180

(p = 0.6523). Human and V-180 conditions both led

to signiﬁcantly shorter durations compared to V-160.

Subjective Mental Workload (NASA-TLX): A

Friedman test found a signiﬁcant difference among

conditions (χ

(3) = 8.60, p = 0.0351). V-200 re-

sulted in signiﬁcantly higher workload scores (79.07

± 9.56) compared to V-160 (53.30 ± 16.30; p =

0.0078), V-180 (p = 0.0273), and Human (p =

0.0391). No signiﬁcant difference was observed be-

tween Human, V-160, and V-180.

Shooting Score: No signiﬁcant difference in

shot accuracy was found across the four conditions

(χ

(3) = 2.10, p = 0.5512). All pairwise Wilcoxon

tests returned non-signiﬁcant results (all p > 0.35),

Cognitive Load and Motor Adjustment Under Virtual Defensive Pressure in Mixed Reality Sports Training

indicating that increased pressure did not translate

into lower scoring performance.

7 DISCUSSION

7.1 Reproducibility of Defensive

Pressure in Mixed Reality

This study demonstrated that spatial pressure from

virtual defenders in a mixed reality (MR) environment

can induce a psychological burden on athletes equiv-

alent to that caused by human defenders. This is sup-

ported by the ﬁnding that virtual defenders—designed

to replicate the appearance, height, and movement

of human counterparts—elicited no signiﬁcant dif-

ferences in subjective mental workload compared to

real human defenders. This indicates that virtual

characters can serve as cognitively demanding agents

and suggests that MR can be expanded from merely

replicating training environments to actively design-

ing pressure conditions.

However, a signiﬁcant difference in ball-holding

duration during shooting was observed between the

Virtual and Human defender conditions. Despite re-

porting comparable subjective pressure, participants

tended to hold the ball slightly longer when facing

virtual defenders. This suggests that subtle discrep-

ancies in bodily sensation speciﬁc to MR environ-

ments—as well as missing or altered environmen-

tal cues such as auditory or tactile feedback—may

have inﬂuenced motor behavior(Kilteni et al., 2012;

Gonzalez-Franco and Lanier, 2017). For example,

factors such as visual latency, depth misperception, or

lack of material realism could have slightly delayed

participants’ decision-making and response timing,

leading to longer preparation phases.

Additionally, analysis of the NASA-TLX sub-

scales revealed that Frustration was reported to be

higher under the virtual defender condition compared

to the human defender condition. This suggests that

even if the overall level of perceived spatial pressure

was similar, emotional stress and feelings of incon-

gruity may have been heightened in the virtual envi-

ronment. Subtle discrepancies from real-world sen-

sations and the lack of natural interactive responses

may have unconsciously triggered irritation among

the participants(Kilteni et al., 2012), which in turn

could have inﬂuenced preparatory motor behavior,

such as the observed prolongation of ball-holding du-

ration.

This dissociation between subjective experience

and observable behavior offers critical insights for im-

proving the design of MR training systems. Even if

visual representations of defensive actions convinc-

ingly recreate subjective pressure, athletes’ timing

and motor accuracy may not fully align with real-

world responses. While one participant stated that

“the pressure induced by the virtual defender felt al-

most the same as from a human defender,” another

participant noted, “In an actual game, if my shot is

blocked, I fail. But in MR, I can still shoot even if

the defender visually blocks me—so it felt unnatural

to play against a virtual defender.”

This comment highlights an important design con-

sideration for current MR systems.: regardless of how

visually realistic they appear, the absence of real con-

sequences (e.g., losing possession) can reduce the ath-

lete’s sense of accountability for their actions, leading

to responses that diverge from those observed in ac-

tual competitive contexts. Therefore, in addition to

achieving visual and motor ﬁdelity, MR systems must

integrate sensory feedback and multisensory cues to

bridge the gap between perception and action.

For example, Chelladurai et al. reported that the

integration of haptic feedback and spatial audio en-

abled users to have a more natural and realistic ex-

perience(Chelladurai et al., 2024). Such multimodal

system design is expected to bring MR experiences

closer to the physicality of actual sports performance,

thereby enhancing the ﬁdelity and generalizability of

training.

7.2 Scalable Cognitive Load via Visual

Properties

In the second experiment, we examined how a sin-

gle visual characteristic, speciﬁcally the height of the

defender, inﬂuenced both cognitive and behavioral

responses. The results revealed clear differences in

NASA-TLX scores and ball-holding durations, sug-

gesting that participants evaluated the physical threat

posed by the defender based on visual information

prior to physical interaction. This evaluation appears

to inﬂuence the temporal aspects of decision-making.

Similarly, prior research by Mousas et al. demon-

strated that variations in the appearance and motion

of virtual characters could modulate users’ emotional

reactivity, supporting the notion that visual features

alone can systematically shape perceived pressure in

immersive environments(Mousas et al., 2018).

In particular, the virtual defender with a height

of 200 cm elicited signiﬁcantly higher cognitive load

compared to the 160 cm and 180 cm conditions, and

also led to a signiﬁcantly shorter ball-holding dura-

tion than the 160 cm condition. Although the height

increase from 180 cm to 200 cm was the same 20

cm as that from 160 cm to 180 cm, clear changes

icSPORTS 2025 - 13th International Conference on Sport Sciences Research and Technology Support

NASA-TLX

Total Score

Free

Human

V-180

V-160

V-200

120

150

n.s.

Shooting Score

Total Score

Free

Human

V-180

V-160

V-200

Shooting Duration

Time (s)

Free

Human

V-180

V-160

V-200

0.5

0.8

1.4

1.1

1.7

Figure 4: Comparative analysis of behavioral and cognitive metrics across all experimental conditions: shooting duration (s),

NASA-TLX total score, and shooting score. (*p < .05, **p < .01; n.s. = not signiﬁcant).

Table 1: NASA-TLX Scores (Mean ± SD) by Condition and Subscale.

Subscale Free Human V-160 V-180 V-200

Mental Demand 22.78 ± 12.77 62.22 ± 22.10 45.00 ± 22.78 48.89 ± 21.76 76.67 ± 12.75

Physical Demand 25.00 ± 12.99 58.33 ± 24.24 41.67 ± 23.05 44.44 ± 19.44 83.33 ± 7.50

Temporal Demand 23.33 ± 13.46 61.67 ± 27.50 27.22 ± 20.48 50.00 ± 21.79 70.00 ± 28.61

Performance 48.33 ± 26.58 51.67 ± 28.39 58.33 ± 25.98 51.11 ± 23.15 55.00 ± 18.87

Effort 43.33 ± 30.00 61.11 ± 21.76 59.44 ± 19.91 55.00 ± 23.05 73.33 ± 18.37

Frustration 23.89 ± 23.42 66.67 ± 17.14 47.22 ± 30.83 61.67 ± 19.84 78.89 ± 20.58

in cognitive and behavioral responses were observed

only in the former case. This suggests, as one partic-

ipant noted—“I normally don’t play against players

who are 200 cm tall, so I felt strong pressure”—that

it was not merely the visual size of the defender,

but rather the surpassing of an internalized reference

norm that acted as a qualitatively distinct threat stim-

ulus. Indeed, Shen et al. demonstrated that in VR

environments, increases in object size—quantiﬁed by

omnidirectional ﬁeld of view occupancy—were sig-

niﬁcantly associated with elevated anxiety levels and

heart rate, supporting the idea that surpassing a cer-

tain size threshold may trigger qualitatively different

cognitive responses(Shen et al., 2022).

Furthermore, detailed analysis of the NASA-TLX

subscales revealed qualitative differences in the na-

ture of cognitive load across conditions. For exam-

ple, Temporal Demand remained particularly low un-

der the 160 cm condition, whereas Effort and Frus-

tration sharply increased under the 200 cm condition.

This suggests that players, who normally rely on au-

tomatized shooting skills, had to exert more conscious

control under stronger perceived pressure, increasing

cognitive and emotional load.

Additionally, the response patterns for Effort and

Frustration did not follow a linear progression, but

instead exhibited a sharp increase speciﬁcally when

moving from 180 cm to 200 cm. This suggests the

presence of a threshold effect in the perception of vir-

tual threat, wherein cognitive and emotional loads es-

calate nonlinearly once the defender’s height or vi-

sual salience surpasses a certain perceptual boundary.

Such nonlinear changes in cognitive and behavioral

responses near the perceptual threshold are consis-

tent with the ﬁndings by Tseng et al., who demon-

strated that perceptual manipulations in VR environ-

ments can lead to abrupt behavioral changes once a

sensory threshold is surpassed(Tseng et al., 2022).

Their work showed that small undetected manipula-

tions could accumulate and trigger nonlinear shifts

in user behavior upon crossing a critical perceptual

boundary, aligning with the present results.

On the other hand, no signiﬁcant difference was

observed between the V-160 and V-180 conditions.

One possible explanation is that the height difference

between these two avatars may have been relatively

minor in the context of the participants’ own heights,

which were all within the 170 cm range. In contrast,

the 200 cm avatar may have exceeded a perceptual

threshold, thereby functioning as a qualitatively dis-

tinct stimulus that provoked a heightened sense of

threat and a stronger cognitive-motor response.

These ﬁndings indicate that, within mixed reality

training environments, it is possible to design not only

motor difﬁculty but also the level of perceived pres-

sure in a gradual and scalable manner by skillfully

manipulating the visual features of virtual defenders.

7.3 Dissociation Between Perceptual

Load and Final Performance

Outcome

In this study, while signiﬁcant differences were ob-

served in subjective workload (NASA-TLX scores)

and preparatory behavior (ball-holding duration), no

substantial statistical differences were found in the

ﬁnal performance outcome, namely shooting accu-

Cognitive Load and Motor Adjustment Under Virtual Defensive Pressure in Mixed Reality Sports Training

racy. These results suggest that although pressure was

perceptually experienced and inﬂuenced participants’

motor strategies, skilled athletes were able to suppress

these effects at the performance level.

This dissociation indicates that skilled athletes

may possess adaptive control strategies that allow

them to maintain task performance even under pres-

sure. While the perception of pressure affected mo-

tor preparation—manifested as a reduction in ball-

holding duration—it did not lead to performance de-

terioration. This could be attributed to motor skill au-

tomatization and the stability of sensorimotor routines

in experienced players.

Such ﬁndings support the utility of mixed reality

(MR) as a tool for simulating performance pressure.

Speciﬁcally, for skilled athletes, MR-based training

can elicit changes in subjective and behavioral re-

sponses without excessively impairing actual perfor-

mance, enabling effective and targeted pressure train-

ing.

Similar patterns have been reported in VR-based

sports training research. For example, Gray demon-

strated that VR batting training, when adaptively de-

signed to match the performer’s skill level, not only

improved virtual task performance but also success-

fully transferred to real-world batting outcomes and

long-term competition levels (Gray, 2017). These

ﬁndings further reinforce the potential of MR systems

to elicit meaningful cognitive and motor adaptations

that generalize beyond the training context.

Furthermore, the absence of signiﬁcant differ-

ences in shooting accuracy across the Free, Human,

and Virtual conditions suggests that participants were

able to execute similar avoidance strategies and mo-

tor decisions under virtual defensive pressure as they

would under real human defense. This indicates that

the use of virtual defenders in MR can provide a

sufﬁciently realistic context for training pressure re-

silience, offering a viable alternative to human-based

defensive drills—especially for situations requiring

repetition, customization, or safety.

This advantage is further enhanced by the unique

capabilities of MR systems, which allow for ﬁne-

grained modulation of perceived pressure (e.g., avatar

height, proximity, or timing) in ways that are difﬁ-

cult to achieve with human defenders. This precision

enables the design of targeted training interventions

that go beyond conventional practice, especially for

exploring edge cases or rare in-game situations.

8 LIMITATIONS

Participant Characteristics. All participants in this

study were nine skilled male basketball players with

over eight years of competitive experience and a

height range of 170–180 cm. While this design en-

sured consistency in skill and physical attributes, it

also introduced selection bias that may limit the gen-

eralizability of our ﬁndings. In particular, the small

sample size poses statistical limitations. This homo-

geneous and limited sample precludes examination of

how athletes of different genders, skill levels, or phys-

ical statures might perceive and respond to virtual de-

fensive pressure.

Limited Scope of Pressure Factors. The type

of pressure manipulated in this study was primarily

spatial and physical—based on the defender’s appear-

ance and proximity. However, real-game pressure

also includes temporal constraints, audience effects,

and situational stress, none of which were represented

in the experimental design. Accordingly, the pressure

experienced in this MR system reﬂects only a subset

of the multifaceted stressors encountered during ac-

tual competitive scenarios.

Constraints on Measurement Indices. The eval-

uation of pressure effects in this study was limited

to three indices: subjective workload (NASA-TLX),

preparatory movement (ball-holding duration), and ﬁ-

nal outcome (shooting accuracy). While these are

meaningful and practical measures, they do not cap-

ture the full range of cognitive and motor changes un-

der pressure.

Simpliﬁed Task Design. The target action in

this study was restricted to three-point shooting in a

”closeout” situation. The task did not involve more

complex offensive behaviors such as dribbling, pass-

ing, or decision-making under multiple response op-

tions. As a result, the system may not fully replicate

the dynamic and strategic demands of in-game scenar-

ios involving high cognitive load and action selection.

Variability Due to Human Execution. In the

Human defender condition, the closeout movement

was manually executed by the defender, resulting in

slight trial-to-trial variability in timing and running

trajectory. Additionally, in all conditions, the passes

delivered to participants were manually thrown by

an experimenter. These sources of human error may

have introduced noise into the measurements of ball-

holding duration and perceived mental workload.

icSPORTS 2025 - 13th International Conference on Sport Sciences Research and Technology Support

9 FUTURE WORKS

Expansion to Decision-Making Tasks. While this

study focused on a single shooting action, actual

gameplay requires players to make rapid decisions

among multiple options such as shooting, driving, or

passing. As a next step, we plan to implement multi-

option tasks where the player selects the appropriate

offensive action based on the virtual defender’s move-

ment, distance, and spatial context.

Development of Multi-User Training Environ-

ments. The current system is designed for a single

user. In the future, we aim to construct a synchro-

nized multi-user MR training environment in which

multiple players, each wearing an HMD, can inter-

act with virtual players in the same shared space.

This would allow for team-based tactical training and

the development of coordinated actions such as posi-

tioning and spacing, thereby extending the system’s

utility from individual skill acquisition to group-level

strategy training.

Application to Remote Training. By capturing

the player’s motion in real time using motion capture

or IMU sensors and reﬂecting it in a virtual avatar,

it becomes possible to construct a training system in

which athletes can practice together in a shared virtual

space despite being geographically apart. For exam-

ple, remote coaches could observe and provide feed-

back on live movement data, or distributed athletes

could engage in synchronized tactical practice, paving

the way for a distributed sports education platform.

Integration of Interactive Virtual Player Con-

trol. Currently, the virtual defender operates using

pre-recorded motion data. Future implementations

may incorporate interactive control, allowing users

(e.g., coaches or training partners) to operate the vir-

tual defender in real time via controllers or user in-

terfaces. This functionality would enable the virtual

defender to respond dynamically to each scenario, en-

hancing the variability and adaptability of training ex-

ercises.

10 CONCLUSION

This study demonstrated that virtual defenders in

mixed reality (MR) can exert psychological and be-

havioral pressure on skilled basketball players compa-

rable to that of real human defenders. Notably, avatars

replicating human appearance, height, and motion

induced similar subjective workload levels (NASA-

TLX), suggesting that realistic pressure can be ef-

fectively reproduced in MR environments. Adjust-

ing the virtual defender’s height alone signiﬁcantly

inﬂuenced ball-holding behavior and perceived pres-

sure, indicating that visual cues can be used to sys-

tematically scale difﬁculty without physical contact.

While differences in cognitive load and preparation

time were observed, shooting performance remained

stable, implying that skilled athletes employed adap-

tive motor strategies even under MR-induced pres-

sure. These ﬁndings suggest that MR can serve not

only as a simulation tool but also as an effective en-

vironment for training psychological resilience and

pressure tolerance without degrading performance.

REFERENCES

Angela, T., Mark, G., and Andrew, W. (2010). Uncanny be-

haviour in survival horror games. Journal of Gaming

& Virtual Worlds, 2(1):3–25.

Avaturn (2025). Avaturn: Realistic 3d avatar creator. Ac-

cessed: 2025-05-03.

Chelladurai, P. K., Li, Z., Weber, M., Oh, T., and Peiris,

R. L. (2024). SoundHapticVR: Head-Based Spatial

Haptic Feedback for Accessible Sounds in Virtual Re-

ality for Deaf and Hard of Hearing Users. In Proceed-

ings of the 26th International ACM SIGACCESS Con-

ference on Computers and Accessibility, pages 1–17.

ACM.

Cheng, L., Jia, H., Yu, L., Wu, Y., Ye, S., Deng, D., Zhang,

H., Xie, X., and Wu, Y. (2024). Viscourt: In-situ guid-

ance for interactive tactic training in mixed reality.

In Proceedings of the 37th Annual ACM Symposium

on User Interface Software and Technology (UIST),

pages 1–14. ACM.

Christos, K. and Michael-Grigoriou, D. (2022). Social in-

teraction with agents and avatars in immersive virtual

environments: A survey. Frontiers in Virtual Reality,

page 786665.

Endo, T., Sekiya, H., and Raima, C. (2023). Psycholog-

ical pressure on athletes during matches and prac-

tices. Asian Journal of Sport and Exercise Psychol-

ogy, 3:161–170.

Fazel, F., Morris, T., Watt, A., and Maher, R. (2018).

The effects of different types of imagery delivery on

basketball free-throw shooting performance and self-

efﬁcacy. Psychology of Sport and Exercise, 39:29–37.

Freeman, D., Haselton, P., Freeman, J., Spanlang, B.,

Kishore, S., Albery, E., Denne, M., Brown, P., Slater,

M., and Nickless, A. (2018). Automated psycholog-

ical therapy using immersive virtual reality for treat-

ment of fear of heights: a single-blind, parallel-group,

randomised controlled trial. The Lancet Psychiatry,

5(8):625–632.

Gerardi, M., Cukor, J., Difede, J., Rizzo, A., and Rothbaum,

B. O. (2010). Virtual reality exposure therapy for post-

traumatic stress disorder and other anxiety disorders.

Current Psychiatry Reports, 12:298–305.

Gerardi, M., Cukor, J., Difede, J., Rizzo, A., and Rothbaum,

B. O. (2020). Virtual reality exposure therapy for so-

Cognitive Load and Motor Adjustment Under Virtual Defensive Pressure in Mixed Reality Sports Training

cial anxiety disorder: a systematic review and meta-

analysis. Psychological Medicine, 50(15):2487–2497.

Gonzalez-Franco, M. and Lanier, J. (2017). Model of il-

lusions and virtual reality. Frontiers in Psychology,

8:1125.

Gray, R. (2017). Transfer of training from virtual to real

baseball batting. Frontiers in Psychology, 8:2183.

Grifﬁn, J. A., McLellan, C. P., Presland, J., Woods, C. T.,

and Keogh, J. W. (2017). Effect of defensive pres-

sure on international women’s rugby sevens attacking

skills frequency and execution. International Journal

of Sports Science & Coaching, 12(6):716–724.

Haga, S. and Mizukami, N. (1996). Japanese version of

nasa task load index: Sensitivity of its workload score

to difﬁculty of three different laboratory tasks. The

Japanese Journal of Ergonomics, 32(2):71–79.

Hart, S. G. and Staveland, L. E. (1988). Development of

nasa-tlx (task load index): Results of empirical and

theoretical research. In Hancock, P. A. and Meshkati,

N., editors, Human Mental Workload, pages 139–183.

North-Holland, Amsterdam.

Hopla, D. (2012). Basketball Shooting. Human Kinetics,

Champaign, Illinois.

Kilteni, K., Groten, R., and Slater, M. (2012). The sense of

embodiment in virtual reality. Presence: Teleopera-

tors and Virtual Environments, 21(4):373–387.

Leander, F., Leon, F., Stefan, A., Darko, J., and Matthias,

K. (2024). The keys of pressing to gain the ball–

characteristics of defensive pressure in elite soccer us-

ing tracking data. Science and Medicine in Football,

8(2):161–169.

Lindsay, R. S., Larkin, P., Kittel, A., and Spittle, M.

(2023). Mental imagery training programs for devel-

oping sport-speciﬁc motor skills: a systematic review

and meta-analysis. Physical Education and Sport Ped-

agogy, 28(4):444–465.

Mori, M. (1970). The uncanny valley. Energy, 7:33–55.

Mousas, C., Anastasiou, D., and Spantidi, O. (2018). The

effects of appearance and motion of virtual characters

on emotional reactivity. Computers in Human Behav-

ior, 86:99–108.

Richards, D. A. and Rose, J. S. (1991). Exposure therapy

for post-traumatic stress disorder. British Journal of

Psychiatry, 158:836–840.

Sampaio, J., Leser, R., Baca, A., Calleja-Gonzalez, J.,

Coutinho, D., Gonc¸alves, B., and Leite, N. (2016).

Defensive pressure affects basketball technical actions

but not the time-motion variables. Journal of Sport

and Health Science, 5(3):375–380.

Sarah, B., Juliana, T., Laura, D., M, P. G., Andr

e, A.,

Juan, M., Tomaz, R., and Mauro, C. (2020). Physical

and physiological demands of basketball small-sided

games: the inﬂuence of defensive and time pressures.

Biology of Sport, 37(2):131–138.

Seyﬁ, S., Fatih, Y. M., and Ahmet, U. (2018). The ef-

fects of rapid strength and shooting training applied

to professional basketball players on the shot percent-

age level. Universal Journal of Educational Research,

6(7):1569–1574.

Shen, J., Kitahara, I., Koyama, S., and Li, Q. (2022). Size

does matter: An experimental study of anxiety in vir-

tual reality. In Proceedings of the 28th ACM Sym-

posium on Virtual Reality Software and Technology

(VRST), pages 1–2. ACM.

Stinson, C. and Bowman, D. A. (2014). Feasibility of train-

ing athletes for high-pressure situations using virtual

reality. IEEE Transactions on Visualization and Com-

puter Graphics, 20(4):606–615.

Takatori, H., Enzaki, Y., Yano, H., and Iwata, H. (2016).

Development of a large-immersive display “larges-

pace”. Transactions of the Virtual Reality Society of

Japan, 21:493–502.

Tianqi, H., Yue, L., and Hai-Ning, L. (2024). Avatar type,

self-congruence, and presence in virtual reality. In

Proceedings of the Eleventh International Symposium

of Chinese CHI, pages 61–72.

Tseng, P.-H., Kellogg, R., and Lopes, P. (2022). The dark

side of perceptual manipulations in virtual reality. In

Proceedings of the 2022 CHI Conference on Human

Factors in Computing Systems, pages 1–18, New Or-

leans, LA, USA.

Wang, Y., Geigel, J., and Herbert, A. (2013). Reading per-

sonality: Avatar vs. human faces. In 2013 Humaine

Association Conference on Affective Computing and

Intelligent Interaction, pages 479–484.

Wellington, R., Gilbert, F., Felipe, S., and Leonardo, L.

(2023). Integrated evaluation of team strategy, training

practices and game performance of a basketball team.

International Journal of Sports Science & Coaching,

18(1):197–206.

Wen, J., Gold, L., Ma, Q., and LiKamWa, R. (2024). Aug-

mented coach: Volumetric motion annotation and vi-

sualization for immersive sports coaching. In 2024

IEEE Conference on Virtual Reality and 3D User In-

terfaces (VR), pages 137–146. IEEE.

Yoon, B., e. Shin, J., i. Kim, H., Oh, S. Y., Kim, D., and

Woo, W. (2023). Effects of avatar transparency on so-

cial presence in task-centric mixed reality remote col-

laboration. IEEE Transactions on Visualization and

Computer Graphics, 29(11):4578–4588.

Yoon, B., i. Kim, H., Lee, G. A., Billinghurst, M., and Woo,

W. (2016). Avatar realism and social interaction qual-

ity in virtual reality. In 2016 IEEE Virtual Reality

(VR), pages 277–278.

Yoon, B., i. Kim, H., Lee, G. A., Billinghurst, M., and Woo,

W. (2019). The effect of avatar appearance on so-

cial presence in an augmented reality remote collab-

oration. In 2019 IEEE Conference on Virtual Reality

and 3D User Interfaces (VR), pages 547–556.

icSPORTS 2025 - 13th International Conference on Sport Sciences Research and Technology Support