The Stress Is Real: Physiological Measurement of League of Legends

Players Experience During a Live Esports Event

David Berga

1,2 a

, Eleonora De Filippi

1 b

, Arijit Nandi

1 c

, Eul

alia Febrer Coll

3 d

Marta Revert

e Bel

, Lautaro J. Russo

4 e

and Alexandre Pereda-Ba

nos

1 f

Eurecat, Centre Tecnol

ogic de Catalunya, Barcelona, Spain

Elisava, Barcelona School of Design and Engineering (UVic-UCC), Spain

eDojo - Electronic Dojo S.L., Barcelona, Spain

Mental Gaming S.L., Barcelona, Spain

dberga@elisava.net, {eleonora.deﬁlippi, arijit.nandi, alexandre.pereda}@eurecat.org,

Keywords:

Esports, League of Legends, Physiology, Arousal, Stress, Electrodermal Activity, Electrocardiogram,

Photoplethysmogram, EDA, GSR, ECG, PPG, Heart Rate.

Abstract:

Videogames and Esports experienced a huge growth in popularity lately and have opened a ripe new ﬁeld for

the study of human behavior. Esports gaming is an area in which videogame players need to cooperate and

compete with each other, inﬂuencing their cognitive load, processing, stress, and as well as social skills. In

this observational study we inquire whether variations in autonomic nervous system activity can be obtained

reliably during a live League of Legends (LoL) event, especially considering this is a preliminary study with a

limited participant sample. We found that game performance (winning or losing the game) signiﬁcantly affects

electrodermal activity and cardiac modulation, where players who lost the game showed higher stress-related

physiological responses, compared to players who won. We also found that speciﬁc important events in the

game, such as ”Killing,” ”Dying,” or ”Destroying the turret,” increased players’ electrodermal and cardiac

modulation compared with other less relevant events, such as ”Placing the guards” or ”Destroying the turret

plates.” Finally, by analyzing activity according to players’ roles, we found various notable activity trends.

Altogether, these (yet preliminary) results encourage further exploration of physiology-based applications for

LoL and Esports players on live events.

1 INTRODUCTION

In Esports gaming video game players must cooper-

ate and compete, often in highly demanding environ-

ments, affecting their cognitive load, emotional pro-

cessing, stress levels, and social skills, among other

aspects. Esports and videogames are on the rise, aided

by the increasing ease with which online gaming en-

vironments allow interacting with other players, and

they thus open a ripe area for neuroscience and exper-

imental psychology research in more realistic and mo-

tivating settings than usually found in laboratory ex-

https://orcid.org/0000-0001-7543-2770

https://orcid.org/0000-0001-5496-652X

https://orcid.org/0000-0003-4238-5183

https://orcid.org/0000-0003-1450-8086

https://orcid.org/0009-0005-4224-6724

https://orcid.org/0000-0002-1145-146X

periments (Pedraza-Ramirez et al., 2020). For exam-

ple, technological developments involving wearable

physiological, multimodal sensors, and interaction-

based metrics, allow for close monitoring of cogni-

tive emotional processes with high temporal resolu-

tion. This opens a landscape of possibilities not only

for basic research, but also for research on advanced

interaction techniques, such as the establishment of

BCI (brain-computer-interface) loops that can be used

for a wide variety of training, self-monitoring appli-

cations, and even to provide the audience with infor-

mation on the state of the players. Of course, there is

no question that games have become highly complex

in the environments they represent, the skills they re-

quire, or the rules and roles involved, thus, any such

applications would need to consider the speciﬁcs of

the game under consideration.To this end we mea-

sured several metrics of electrodermal activity and

cardiac modulation among competitive players dur-

Berga, D., De Filippi, E., Nandi, A., Coll, E., Bel, M., Russo, L. and Pereda-Baños, A.

The Stress Is Real: Physiological Measurement of League of Legends Players Experience During a Live Esports Event.

DOI: 10.5220/0012981700003828

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

In Proceedings of the 12th International Conference on Sport Sciences Research and Technology Support (icSPORTS 2024), pages 199-205

ISBN: 978-989-758-719-1; ISSN: 2184-3201

199

ing several sessions of League of Legends games that

took place in a live Esports event celebrated at a

dedicated gaming venue. Together with the physi-

ological metrics we obtained the activity logs from

the corresponding games, provided by the game de-

veloper (Riot Games), to study if and how different

events and roles elicit varying cognitive/emotional re-

sponses. aiming to ascertain the utility of such met-

rics for future biofeedback-based training of different

game skills in LoL. Research in the ﬁeld of psychol-

ogy has traditionally focused on three main forms of

both emotional and attentional responses: subjective

perceptions of the individual about their own state, ef-

fects on behavior, and changes in their physiological

patterns, such as acceleration or deceleration of heart

rate and increase in skin conductivity (Bradley and

Lang, 2000)(Mauss and Robinson, 2009). Each one

of these approaches comes with both advantages and

disadvantages. First, auto-informed methods, such as

questionnaires or interviews, for which participants

are directly asked to report their status are the only

way to access the individual’s subjective perception

but are also limited by the individual’s own ability

to introspect, since many psychological processes can

occur unconsciously or with low levels of conscious-

ness (Nisbett and Wilson, 1977). Moreover, another

limitation of this approach is that cognitive biases

(such as social desirability bias) may interfere with

the reports, thus making the information not entirely

reliable. For this reason, in the ﬁeld of experimen-

tal psychology research, the analysis of physiological

responses (for example, variations in heart rate, skin

conductivity or activation of facial muscles) has been

introduced as a way of obtaining information about

the cognitive and emotional processes of individuals

in an indirect way. On the other hand, the main disad-

vantage of the physiological methods with respect to

the self-reports concerns their ecological validity, so it

is recommended to combine both methodologies (Ca-

cioppo et al., 2017). Here, we used two main physio-

logical measures:

Electrodermal Activity (or EDA, also known as Gal-

vanic Skin Response or GSR) is a correlate of the

activation of the sympathetic branch of the autonomic

nervous system, which provides reliable information

with a high temporal resolution about participants’

physiological arousal, related to the intensity of ex-

perienced cognitive-emotional responses.

Cardiac Modulations (ECG or Electrocardiogram),

or Photoplethysmography (PPG) is a measure of

cardiovascular pulse (i.e. Heart Rate). This signal

is mediated by both the sympathetic and parasympa-

thetic systems, thus responding to both physiologi-

cal arousal and emotional regulation processes. This

is dependent on both the intensity of emotions and

their hedonic load, also with the appearance of cogni-

tive resources for stimulus processing. The variability

of heart rate can be analyzed by looking at the ratio

of sympathetic vs parasympathetic modulation within

the PPG signal.

Over the last decade, several physiological studies

have been carried out in the Esports ﬁeld (Kivikan-

gas et al., 2011)(Argasi

nski and Grabska-Gradzi

nska,

2017)(Alhargan et al., 2017) and have reported rele-

vant differences in participants’ affective states dur-

ing games. Among them, some studies have fo-

cused on the analysis of muscle signals (Electromyo-

graphy/EMG) (Ahsan et al., 2009), of brain signals

(Electroencephalography/EEG) (Hafeez et al., 2021),

or of facial gestures (Samara et al., 2017). The main

focus of these latter studies was to assess the emo-

tional state of participants from these measures, clas-

sifying each of the 7 affective categories of basic emo-

tions initially deﬁned by (Ekman, 2005): anger, sad-

ness, fear, disgust, joy, surprise, contempt, and neu-

tral. Most of these studies have used psychophysio-

logical tests such as the Russell test (Russell, 1980)

or the Self Assessment Mannekin (Bradley and Lang,

1994), as a guideline for the validation of physiologi-

cal measures that could affect the affective responses

of participants. In a recent review by (Leis and

Lautenbach, 2020), 17 studies were meta-analyzed

in Esports contexts for psychological and physiolog-

ical stress, and it was concluded that simply play-

ing in an Esports non-competitive environment pro-

duced no stress reactions, whereas in competitive en-

vironments several studies reported increases in anx-

iety levels, cortisol levels, and physiological sympa-

thetic activation, all three indicators of stress (Jones

et al., 2012)(Yaribeygi et al., 2017). However, stress

is not the only interesting indicator to consider in Es-

ports environments, since peripheral physiology can

also provide insight into various aspects of informa-

tion processing, such as emotion, engagement, bore-

dom or frustration.

In our study, we collected both EDA and ECG data

while users were engaged in a widely played desk-

top videogame ”League of Legends”. The aim of this

work was to evaluate the affective responses by ana-

lyzing distinct EDA and ECG metrics depending on

game performance (winning or losing). Moreover,

we investigated whether different game events (e.g.

”Killing”, ”Dying”, ”Kill Assist”, ”Destroy Turret”,

”Destroy Turret Plate” or ”Placing Ward”) elicited

distinct physiological responses. Lastly, we wanted to

icSPORTS 2024 - 12th International Conference on Sport Sciences Research and Technology Support

200

elucidate whether different player roles (Jungle, Mid-

dle, Utility, Bottom, and Top) also impact physiolog-

ical measures.

2 METHODS

2.1 Experimental Design

Sensors and Software. For our experiments we

used a Shimmer3

pack with 5 simultaneous GSR+

Units providing Galvanic Skin Response for acquisi-

tion of Electrodermal Activity (EDA), as well as Car-

diac Pulse (PPG) estimating heart rate variations. We

developed our own Python tools (available in github

)

for capturing data and sending it through Bluetooth

(using pyserial and pylsl) and synchronizing that data

with events of the game (using riotwatcher) for later

statistical analysis for EDA (Ledalab

V3.4.9) and

PPG (HeartPy

) respectively.

League of Legends Skills and Mechanics. League

of Legends

is a Multiplayer Online Battle Arena

(MOBA) videogame developed by Riot Games

2009. It has become one (if not) the most popular

game of the current generation of online videogames

and has a signiﬁcant impact on the economy of many

other industrial and technology sectors (i.e. brand-

ing, consumables, social networks and media con-

tent). The original League of Legends game type

mechanics (Summoner Rift) consists of a competi-

tion between 2 teams of 5 players in which each team

has to destroy the enemy base or ”Nexus”. The battle

zone is composed by 3 lanes (top, mid and bot) as well

as a jungle (in between these lanes). Each team lane

has 4 turrets and an inhibitor, and two more turrets

in the Nexus. However, before reaching the enemy

base, each team needs to destroy the turrets and the

inhibitor of one of the enemy lanes. The usual team-

work consists of two players (Carry/Bottom and Util-

ity/Support) controlling the bot lane, the Top/Tank

controlling the top lane, the Middle controlling the

mid lane and the Jungle that moves around all the

lanes. As in other MOBAs, players will need to co-

operate and sometimes play aggressively to kill the

https://shimmersensing.com/

https://github.com/dberga/riotwatcher-shimmer-pyn

put

http://www.ledalab.de/

https://pypi.org/project/heartpy/

https://www.leagueoflegends.com/

https://www.riotgames.com/en

https://www.bcg.com/publications/2023/how-Esports

-will-become-future-of-entertainment

opponent’s champions and overcome the enemy posi-

tions. Riot users (summoners) have a speciﬁc experi-

ence level

(given its in-game time) and rank

(given

its actual performance in ranked games).

Game Sessions and Subjects. A total of 4 sessions

have been performed in the Asobu Esports Experi-

ence venue

with 12 participants (contacted and se-

lected by United Gamers Academy

) in the game-

play experimentation. The participants were recruited

through questionnaires reporting age, gender, skill

level, rank level and game preferences. Only play-

ers with higher rank levels above silver III were se-

lected for the study. Subjects’ age was between 18

and 25 years old (4 women and 12 men). Average

player level was 216 (with lowest 82 and biggest 402)

and corresponded to silver-gold S12 competitive rank

gamers. Due to the limitation in the quantity of sen-

sors available, we captured data from 12 participants

on 4 sessions of playing a speciﬁc team during Sum-

moner’s Rift gameplay (avg time 30-45 min), later

ﬁltered on 7 with enough valid events for statisti-

cal comparison. We cut the recording of these par-

ticipants from the start to the end of the game and

we set speciﬁc window times for each event (i.e. 5

sec). This data is synchronized with events down-

loaded from riotwatcher api

. Some events avail-

able for capture are ”killing”, ”dying”, ”kill assist”,

”special killing”, ”item purchased”, ”level up”, ”ward

placed”, ”building kill”, ”champion transform”, ”tur-

ret plate destroyed” and ”elite monster kill”. After

gameplay we annotated riot’s metadata for each par-

ticipant such as game session data (total kills/deaths

or damage done/received), win or loss condition and

player roles (top, mid, bot, utility and jungle).

2.2 Physiological Data Processing

GSR Preprocessing. We have processed raw GSR

data with Ledalab to extract the following measures:

nrSCRs (total skin conductance number ”#” of re-

sponses above threshold), Latency (delay/surpassed

time ”s” to elicit EDA with respect to the event), Am-

plitude (mean activity ”mV” inside the event win-

dow), PhasicMax (max phasic value ”mV” from the

gap with respect the response and the event window)

https://leagueoflegends.fandom.com/wiki/Experience

(summoner)

https://leagueoflegends.fandom.com/wiki/Rank (Leag

ue of Legends)

https://asobuEsports.com/

https://unitedgamers.pro/

https://developer.riotgames.com/

The Stress Is Real: Physiological Measurement of League of Legends Players Experience During a Live Esports Event

201

and Tonic (max tonic activity ”mV” with respect win-

dow). See Ledalab’s documentation

for more de-

tails. Previous literature in electrodermal physiology

has shown EDA can be a reliable quantiﬁer of sympa-

thetic dynamics (Posada-Quintero et al., 2016), mean-

ing higher EDA correlated with higher sympathetic

(stress/alert) levels.

The Matlab-based toolbox ”Ledalab” (Benedek

and Kaernbach, 2010) was used for the GSR signal

preprocessing and analysis. First, we carried out a

preliminary visual examination to look for periodic

drift in the signal, which reﬂects artifacts, and we re-

sampled the raw signal to 50Hz using Neurokit2

The following preprocessing operations were then

carried out using Ledalab toolbox: low-pass Butter-

worth ﬁltering with a cutoff frequency of 5 Hz, and

smoothing to eliminate any remaining artifacts. Fi-

nally, we performed an event-related analysis utiliz-

ing the Continuous Decomposition Analysis (CDA)

to extract the features indicating Skin Conductance

Responses (SCRs). By extracting the phasic (driver)

information underlying EDA, this approach attempts

to obtain the signal features of the underlying sudo-

motor nerve activity. Skin conductance data is decon-

voluted by the overall response shape, considerably

enhancing temporal accuracy. This method enables

the extraction of continuous phasic and tonic activity

based on traditional deconvolution within a predeter-

mined time window, which for us corresponded to a

window comprising the three seconds before an event

marker to the ﬁve following seconds. The number of

SCRs within the response window, response latency

for the ﬁrst SCR, mean SCR amplitudes, maximum

phasic, and average tonic activity within the speciﬁed

window were therefore collected for each event de-

scribed in the previous section.

PPG Data Processing. We have processed raw

PPG data with Heartpy to obtain the BPM (”#” beats

per minute), IBI (time ”ms” of interbeat interval or

R-R), SDNN (standard deviation of intervals ”ms” be-

tween adjacent beats of the IBI of normal sinus beats),

SDSD (standard deviation of successive differences

between adjacent R-R intervals ”ms”) and RMSSD

(root mean square of successive differences between

adjacent R-R intervals ”ms”). The latter metrics

(SDSD and RMSSD) are related to the measurement

of HRV (heart rate variability). Indeed, higher HRV

(higher values of SDSD or RMSSD) can represent

parasympathetic/vagal modulation (associated with a

state of relaxation), while a lower HRV (lower values

http://www.ledalab.de/documentation.htm

https://neuropsychology.github.io/NeuroKit/

for SDSD or RMSSD) represents sympathetic/ﬂight-

or-ﬁght modulation (being stressed or alert; (Valenza

et al., 2018)). Here we have to point out that studies

on HRV are commonly analyzed over large timeline

streams of heart rate data (about 5 min or more; (Shaf-

fer and Ginsberg, 2017)) documentation explains the

aforementioned metrics. However, our measurements

of HRV are considering 5 to 10-second time windows

according to the League of Legends fast-paced events.

Processing and analysis of raw PPG data

were conducted using the Python-based toolkit

”Heartpy”(Van Gent et al., 2019), specialized for the

analysis of PPG signal as compared to ECG. At ev-

ery heartbeat, blood perfuses via the capillaries and

arteries, causing the skin to become discolored. The

PPG detects this discoloration. The systolic peak, di-

astolic notch, and diastolic peak make up the signal.

First, as we did with the GSR signal, we resampled

the raw PPG signal to 50Hz using Neurokit2. Then,

we run the processing algorithm that comes with the

Heartpy toolkit and which allows for the peak detec-

tion to extract reliable time-domain measures, such as

beats per minute (BPM), and Interbeat Intervals (IBI).

Furthermore, for each event, we extracted measures

that reﬂect Heart Rate Variability (HRV) such as the

RMSSD (root mean square of successive differences)

and the SDSD (standard deviation of successive dif-

ferences).

3 RESULTS

We performed data curation for our statistical analy-

sis using data from 7 participants (a total of 2 game

sessions with recorded measures in which 3 partic-

ipants played twice) with enough event samples for

later analysis and processing. Some of the data not

mentioned in participants results was discarded for

the ﬁnal evaluation due to the incorrect samples from

the sensor data and/or given the lack of relevant events

in the gameplay to sync with the sensor data (e.g.

given too much time being dead without interacting

with game objects nor players). Here we select the

players’ sensor data collected with enough samples

to retrieve individual PPG and EDA patterns to make

valid statistical evaluations.

3.1 Physiological Results: Skin

Conductance

In Table 1 we show mean statistics of nrSCR, La-

tency, Amplitude, PhasicMax, and Tonic values of

players that win the gameplay and lose the game-

play. Similarly, in Table 2 we show statistics for

icSPORTS 2024 - 12th International Conference on Sport Sciences Research and Technology Support

202

Table 1: Win and Loss mean GSR metrics by stacking all events in one statistic. Observations from 7 participants playing

during 2 sessions. nrSCR: skin conductance responses over threshold. *p < 0.05 between all events.

Result nrSCR Latency Amplitude PhasicMax Tonic

WIN 1.85±1.73 -0.15±1.93 0.27±0.63 0.55±1.17 12.58±7.05

LOSS 2.60±2.21 -0.76±1.70 0.19±0.41 0.37±0.55 6.88±4.74

TOTAL 2.37±2.10 -0.57±1.79 0.22±0.49 0.42±0.79 8.61±6.12

Table 2: Event mean GSR metrics from events ”Killing”, ”Dying”, ”Placing Ward”, ”Destroying Turret” and ”Destroying

Turret Plate”. Observations from 7 participants playing during 2 sessions. nrSCR: skin conductance responses over threshold.

*p < 0.05 between win/lose.

EVENT nrSCR Latency Amplitude PhasicMax Tonic N

KILL 1.23±1.17 0.05±2.11 0.03±0.04 0.10±.07 7.76±3.23 9

DIE 2.42±2.10 -0.19±0.74 0.32±0.63 0.50±.86 12.20±5.82 35

PLACE WARD 2.32±2.16 -0.30±2.15 0.23±0.45 0.15±0.14 12.06±5.41 75

DES.TURRET 2.18±2.21 -0.61±1.54 0.16±0.23 0.25±0.44 9.75±6.27 79

DES.PLATE 2.18±1.89 -0.46±1.84 0.29±0.59 0.46±0.68 3.09±1.77 49

events ”Killing”, ”Dying”, ”Place Ward”, ”Destroy

Turret” and ”Destroy Turret Plate”. We expand these

statistics in Supplementary Material-Table 6 ﬁltering

player roles in the game.

Given the Chi-squared measured distributions

(non-parametric) we performed Friedman’s tests over

win-loss and event data for each GSR metric. On

analyzing winning or losing the match (Tables 1-2),

we observed that the nrSCR, Amplitude, and Pha-

sicMax activities were signiﬁcantly higher during the

”Killing” event for players (p = .046, χ

= 4.000).

Additionally, nrSCR and Amplitude values were also

signiﬁcantly elevated during the ”Destroying Turret”

event (p = .020, χ

= 5.444), while nrSCR activity

alone showed a signiﬁcant increase during the ”De-

stroying Plate” event (p = .008, χ

= 7.143), with

Amplitude showing a trend towards signiﬁcance (p

= .071, χ

= 3.266). Tonic activity differed signiﬁ-

cantly only in relation to the ”Dying” event (p = .035,

= 4.455) and ”Placing Ward” event (p = .002, χ

10.000) between winning and losing conditions.

When comparing GSR activity distributions

across all events for winning players, we found that

Latency was signiﬁcantly shorter (p = .024, χ

11.265), Amplitude was signiﬁcantly higher (p =

.041, χ

= 9.959), and Tonic activity was signiﬁcantly

greater (p = .010, χ

= 13.28). PhasicMax showed a

trend towards higher values, though it did not reach

statistical signiﬁcance (p = .092, χ

= 8.000). In con-

trast, there were no signiﬁcant differences in GSR ac-

tivity across events for players who lost the game.

3.2 Physiological Results: Heart Rate

In Table 3 and Supplementary Material-Table 5 we

show mean statistics of pulse metrics according to win

condition, event, and role.

After performing Friedman tests over PPG met-

rics for all events, we found that SDSD was signif-

icantly higher when winning the game (p = .016,

= 10.371). BPM was signiﬁcantly lower and IBI

was signiﬁcantly longer (p = .041, χ

= 8.28). We

also tested for differences between winning and los-

ing the game for each speciﬁc event. For ”Destroy-

ing Turret Plate”, SDSD was signiﬁcantly higher (p

= 5.32×10

−4

, χ

= 12.0) and RMSSD was signiﬁ-

cantly increased (p = .004, χ

= 8.333). During the

”Dying” event, SDSD was signiﬁcantly elevated (p

= .011, χ

= 6.4) and RMSSD was also higher (p =

.002, χ

= 10). Additionally, SDSD showed a signiﬁ-

cant increase when ”Placing a Ward” (p = .011, χ

6.4).

4 CONCLUSIONS

This study shows the potential of using physiologi-

cal measurements (EDA and ECG) to monitor cog-

nitive and emotional processes in complex game en-

vironments such as League of Legends, and supports

the idea that these metrics could enable biofeedback

based loops for interaction, training, and showcasing

purposes. In the study, we characterized physiolog-

ical responses depending on performance, events as

well as participants’ roles in the game.

The Stress Is Real: Physiological Measurement of League of Legends Players Experience During a Live Esports Event

203

Table 3: Win and Loss mean PPG metrics by stacking all events in one statistic. Observations from 7 participants. BPM:

Beats per minute; IBI: interbeat interval; SDNN: deviation between adjacent beats; SDSD: deviation of differences between

R-R intervals; RMSSD: successive differences between R-R intervals. *p < 0.05 between all events.

Result BPM IBI SDNN SDSD RMSSD

WIN 172±113 531±333 129±48 *100±47 202±87

LOSS *94±50 *743±212 77±52 57±44 117±92

TOTAL 114±80 688±265 90±56 68±49 140±98

Table 4: Event mean PPG metrics from events ”Killing”, ”Dying”, ”Placing Ward”, ”Destroying Turret” and ”Destroying

Turret Plate”. Observations from 7 participants playing during 2 sessions. BPM: Beats per minute; IBI: interbeat interval;

SDNN: deviation between adjacent beats; SDSD: deviation of differences between R-R intervals; RMSSD: successive differ-

ences between R-R intervals. N are event occurrences. *p < 0.05 between win/lose.

EVENT BPM IBI SDNN SDSD RMSSD N

KILL 69±10 892±139 73±46 57±31 99±54 7

DIE 103±66 726±249 87±54 *67±46 *132±86 37

PLACE WARD 122±75 648±279 92±54 *76±54 143±94 31

DES.TURRET 118±95 678±260 95±56 72±52 146±96 57

DES.PLATE 117±78 673±276 *90±60 *64±47 140±111 71

In most cases, we found signiﬁcant differences

in EDA (for nrSCR, Amplitude, and PhasicMax ac-

tivity) during ”Killing”, ”Destroying Turret” or ”De-

stroying Turret Plate” between players that are win-

ning the game and players that are losing the game,

by showing more relaxed states for winning players.

One should consider that more relaxed state cannot

be achieved when there is greater activity on GSR.

Moreover, when players were winning the game, they

showed distinct patterns of physiological activity de-

pending on the events in the game (e.g., ”Killing”,

”Destroying Turret”, ”Destroying Plate”). In contrast,

we did not ﬁnd any signiﬁcant difference between

these events for players that were losing the game, as

they remained overall with lower HRV modulation.

This can hinder the possibility that players that per-

form badly show similar physiological states (being

alert and/or excited) across the game, while players

that perform well have distinct physiological behav-

ior during the course of game events.

For the case of PPG, similarly to the aforemen-

tioned, SDSD was signiﬁcantly distinct for players

that were winning the game between different events.

On the other side, we found that only IBI and BPM

measures showed signiﬁcant differences for players

that were losing the game. Overall, players that per-

formed better (winning) showed signiﬁcantly higher

parasympathetic modulation (i.e., relaxation) than the

ones that were losing. These results suggest that poor

game performance induces higher stress or to be in a

state of alert, while players that perform better tend

to remain in more relaxed states. Furthermore, the

analysis for speciﬁc events, like ”Dying”, ”Destroy-

ing Turret Plate” or ”Placing Ward”, has shown that

players have distinct values of SDSD and RMSSD,

with Killing” or ”Dying” events inducing higher sym-

pathetic modulation (lower HRV).

Study Limitations and Future Work. Despite the

lack of physiological samples for participants and

game sessions we obtained enough measurements to

pinpoint differences in-game performance and events,

conﬁrming the potential for using these measures in

live events. By having a higher number of participants

and game sessions we would suggest undergoing sim-

ilar studies, not only for analyzing physiology over

game performance and events but also for conducting

an in-depth analysis of game roles, champions, player

level, and type of match (beyond League of Legends’

summoner’s rift) in relation with EDA and ECG mea-

surements. Further exploration in this context would

expand both research and development capabilities on

evaluating physiological responses in eSports.

4.1 Funding

This study has been possible through the Grant

IRC 2020 (Reference ACE033/21/000046) funded by

ACCI

O (Catalan Agency for Business Competitive-

ness), from the project ”Esports-LAB” lead by IN-

DESCAT (Associaci

o Catalana Cl

uster de la Ind

ustria

de l’Esport), partners with Generalitat de Catalunya,

EsportCat and Fundaci

o Catalana per l’Esport.

icSPORTS 2024 - 12th International Conference on Sport Sciences Research and Technology Support

204

4.2 Conﬂicts of Interest

The authors declare that they have no known com-

peting ﬁnancial interests or personal relationships that

could have appeared to inﬂuence the work reported in

this paper. Experiment participants signed an autho-

rization form prior to the study to authorize the usage

of the captured physiological data as well as perfor-

mance from their Riot Gamertag and remained anony-

mous according to the Spanish national law LOPD

(Ley Org

anica de Protecci

on de Datos de Car

acter

Personal).

REFERENCES

Ahsan, M. R., Ibrahimy, M. I., and Khalifa, O. O. (2009).

Emg signal classiﬁcation for human computer interac-

tion: a review. volume 33, pages 480–501.

Alhargan, A., Cooke, N., and Binjammaz, T. (2017). Multi-

modal affect recognition in an interactive gaming en-

vironment using eye tracking and speech signals. In

Proceedings of the 19th ACM International Confer-

ence on Multimodal Interaction. ACM.

Argasi

nski, J. K. and Grabska-Gradzi

nska, I. (2017). Pat-

terns in serious game design and evaluation applica-

tion of eye-tracker and biosensors. In Artiﬁcial Intel-

ligence and Soft Computing, pages 367–377. Springer

International Publishing.

Benedek, M. and Kaernbach, C. (2010). A continuous mea-

sure of phasic electrodermal activity. Journal of neu-

roscience methods, 190(1):80–91.

Bradley, M. M. and Lang, P. J. (1994). Measuring emotion:

The self-assessment manikin and the semantic differ-

ential. Journal of Behavior Therapy and Experimental

Psychiatry, 25(1):49–59.

Bradley, M. M. and Lang, P. J. (2000). Measuring emo-

tion: Behavior, feeling, and physiology. In Lane, R.

D. R., Nadel, L., Ahern, G. L., Allen, J., and Kasz-

niak, A. W., editors, Cognitive Neuroscience of Emo-

tion, pages 25–49. Oxford University Press.

Cacioppo, J. T., Chen, H. Y., and Cacioppo, S. (2017).

Reciprocal inﬂuences between loneliness and self-

centeredness: A cross-lagged panel analysis in a

population-based sample of african american, his-

panic, and caucasian adults. Personality and Social

Psychology Bulletin, 43(8):1125–1135.

Ekman, P. (2005). Basic emotions. In Handbook of Cogni-

tion and Emotion, pages 45–60. John Wiley & Sons,

Ltd.

Hafeez, T., Saeed, S. M. U., Arsalan, A., Anwar, S. M.,

Ashraf, M. U., and Alsubhi, K. (2021). EEG in game

user analysis: A framework for expertise classiﬁcation

during gameplay. PLOS ONE, 16(6):e0246913.

Jones, B. J., Tan, T., and Bloom, S. R. (2012). Minire-

view: Glucagon in stress and energy homeostasis. En-

docrinology, 153(3):1049–1054.

Kivikangas, J. M., Chanel, G., Cowley, B., Ekman, I.,

Salminen, M., J

arvel

a, S., and Ravaja, N. (2011). A

review of the use of psychophysiological methods in

game research. Journal of Gaming & Virtual Worlds,

3(3):181–199.

Leis, O. and Lautenbach, F. (2020). Psychological and

physiological stress in non-competitive and competi-

tive esports settings: A systematic review. Psychology

of Sport and Exercise, 51:101738.

Mauss, I. B. and Robinson, M. D. (2009). Measures of emo-

tion: A review. Cognition & Emotion, 23(2):209–237.

Nisbett, R. E. and Wilson, T. D. (1977). The halo effect: Ev-

idence for unconscious alteration of judgments. Jour-

nal of Personality and Social Psychology, 35(4):250–

256.

Pedraza-Ramirez, I., Musculus, L., Raab, M., and Laborde,

S. (2020). Setting the scientiﬁc stage for esports psy-

chology: a systematic review. International Review of

Sport and Exercise Psychology, 13(1).

Posada-Quintero, H. F., Florian, J. P., Orjuela-Ca

on, A. D.,

and Chon, K. H. (2016). Highly sensitive index of

sympathetic activity based on time-frequency spectral

analysis of electrodermal activity. American Journal

of Physiology-Regulatory, Integrative and Compara-

tive Physiology, 311(3):R582–R591.

Russell, J. A. (1980). A circumplex model of affect. Journal

of Personality and Social Psychology, 39(6):1161–

1178.

Samara, A., Galway, L., Bond, R., and Wang, H. (2017).

Affective state detection via facial expression analysis

within a human computer interaction context. Journal

of Ambient Intelligence and Humanized Computing,

10(6):2175–2184.

Shaffer, F. and Ginsberg, J. P. (2017). An overview of heart

rate variability metrics and norms. Frontiers in Public

Health, 5.

Valenza, G., Citi, L., Saul, J. P., and Barbieri, R. (2018).

Measures of sympathetic and parasympathetic auto-

nomic outﬂow from heartbeat dynamics. Journal of

Applied Physiology, 125(1):19–39.

Van Gent, P., Farah, H., Van Nes, N., and Van Arem, B.

(2019). Heartpy: A novel heart rate algorithm for

the analysis of noisy signals. Transportation research

part F: trafﬁc psychology and behaviour, 66:368–378.

Yaribeygi, H., Panahi, Y., Sahraei, H., Johnston, T. P., and

Sahebkar, A. (2017). The impact of stress on body

function: a review. EXCLI Journal; 16:Doc1057;

ISSN 1611-2156.

SUPPLEMENTARY MATERIAL

See Tables 5 and 6 in https://drive.google.com/file/d/

1zByOo59gS2x0cGhZLhY5akaVFhTGlAQi/view?u

sp=sharing.

The Stress Is Real: Physiological Measurement of League of Legends Players Experience During a Live Esports Event

205