ArticlePDF Available

Abstract and Figures

Research into the effects of action video gaming on cognition has largely relied on self-reported action video game experience and extended video game training. Only a few studies have focused on participants' actual gaming skills. However, whether superior players and average players have different executive control is still not fully demonstrated. This study had top-ranking League of Legends players (global top 0.17%; N = 35) and average-ranking League of Legends players (N = 35) perform two cognitive tasks that aimed to measure three aspects of executive functioning: cognitive flexibility, interference control, and impulsive control. We controlled self-reported gaming experience, so that top-ranking players and average-ranking players had similar years of play and hours of play per week. We found that compared to a group of average players, top players showed smaller task-switching costs and smaller response-congruency effects in a Stroop-switching test. In a continuous performance test, top players indicated higher hit rates and lower false alarm rates as compared to average players. These findings suggest that top players have better cognitive flexibility and more accurate control of interference in the context of task-switching. Moreover, top players exhibit better impulsive control. The present study provides evidence that players' gaming skills rather than gaming experience are related to cognitive abilities, which may explain why previous studies on self-reported gaming experience and those assessing supervised training and cognitive performance have shown inconsistent results.
Content may be subject to copyright.
Running head: EXPERT GAME SKILLS AND EXECUTIVE CONTROL
Time for a True Display of Skill: Top Players in League of Legends Have Better
Executive Control
Xiangqian Li1, Liang Huang1, Bingxin Li2*, Haoran Wang1, Chengyang Han3,4
1Department of Psychology, School of Social Development and Public Policy, Fudan
University, Shanghai, China
2Institute of Psychology, Chinese Academy of Sciences, Beijing, China
3Shenzhen Key Laboratory of Affective and Social Cognitive Science, Shenzhen University,
Shenzhen, China
4College of Psychology, Shenzhen University, Shenzhen, China
Length: Text 8227 words, 2 Tables, 4 Figures, 79 References
Xiangqian Li and Liang Huang share first authorship.
CORRESPONDENCE:
Bingxin Li
Institute of Psychology
Chinese Academy of Sciences,
Beijing, China
100101
E-mail: libx@psych.ac.cn
EXPERT GAME SKILLS AND EXECUTIVE CONTROL
1
Abstract
Research into the effects of action video gaming on cognition has largely relied on
self-reported action video game experience and extended video game training. Only a few
studies have focused on participants’ actual gaming skills. However, whether superior players
and average players have different executive control is still not fully demonstrated. This study
had top-ranking League of Legends players (global top 0.17%; N = 35) and average-ranking
League of Legends players (N = 35) perform two cognitive tasks that aimed to measure three
aspects of executive functioning: cognitive flexibility, interference control, and impulsive
control. We controlled self-reported gaming experience, so that top-ranking players and
average-ranking players had similar years of play and hours of play per week. We found that
compared to a group of average players, top players showed smaller task-switching costs and
smaller response-congruency effects in a Stroop-switching test. In a continuous performance
test, top players indicated higher hit rates and lower false alarm rates as compared to average
players. These findings suggest that top players have better cognitive flexibility and more
accurate control of interference in the context of task-switching. Moreover, top players
exhibit better impulsive control. The present study provides evidence that players’ gaming
skills rather than gaming experience are related to cognitive abilities, which may explain why
previous studies on self-reported gaming experience and those assessing supervised training
and cognitive performance have shown inconsistent results.
Key words: Executive control, gaming skills, task-switching, cognitive flexibility,
impulsive control
EXPERT GAME SKILLS AND EXECUTIVE CONTROL
2
1. Introduction
Action video games require players to vigilantly monitor the visual periphery while
responding quickly to or switching rapidly among multiple targets. In the past two decades,
numerous studies have reported that, compared to non-action video game players (nAVGP),
action video game players (AVGPs) have increased abilities that are cognitive in nature (see a
recent review by Bediou et al., 2018). Specifically, researchers found that experience and
training on action video games led to improved allocation of visual attention and visual
search (e.g., Castel, Pratt & Drummond, 2005; Green & Bavelier, 2003, 2006, 2007; Azizi,
Abel & Stainer, 2017; but see Unsworth et al., 2015). Other studies have suggested better
cognitive flexibility, interference control and impulsive control in AVGPs over nAVGPs
(Andrews & Murphy, 2006; Dobrowolski, Hanusz, Sobczyk, Skorko & Wiatrow, 2015;
Strobach, Frensch & Schubert, 2012). Nevertheless, these results are not without controversy.
1.1 Game Experience and Cognitive Flexibility
Cognitive flexibility is an important aspect of executive function, which denotes the
ability to flexibly switch between tasks or mental sets and avoid being stuck on ineffective
strategies (Dimond, 2013). By applying a task-switching test, Boot and colleagues (2008)
examined how video game experience impacted participants’ cognitive flexibility in a
longitudinal study and a cross-sectional study. In their cross-sectional study, it was found that
AVGPs outperformed nAVGPs, with AVGPs showing smaller task-switching costs. The
results from their longitudinal study in nAVGPs indicated that video game training did not
produce any significant effect on the ability to rapidly switch between two tasks. Specifically,
nAVGPs showed no improvements in task-switching performance after twenty-one hours of
video game training (Boot et al., 2008). However, as shown by Green and colleagues (2013;
EXPERT GAME SKILLS AND EXECUTIVE CONTROL
3
Experiment 4), fifty hours of action video training reduced the task-switching costs in
nAVGPs. Researchers concluded that task-switching ability may be improved in nAVGPs but
not to the same level as in AVGPs (Green et al., 2013). The switch benefit in AVGPs was not
limited to the response mode. AVGPs exhibited smaller task-switching costs compared to
nAVGPs in both the manual and vocal responding conditions (Green et al., 2013; Experiment
1).
However, not all studies have established smaller task-switching costs in AVGPs as
compared to nAVGPs (Cain, Landau & Shimamura, 2012; Karle, Watter & Shedden, 2010).
For example, Cain and colleagues (2012) tested AVGPs’ and nAVGPs’ cognitive flexibility
using a flanker-switching test, where participants had to alternate between a pro-response
arrow task (the easier task; left arrow press left; right arrow press right) and an anti-
response arrow task (the harder task; left arrow press right; right arrow press left). Cain
et al. (2012) found that for nAVGPs switching from the relatively harder task to the relatively
easier task caused greater task-switching costs than the other way around, whereas for
AVGPs such asymmetry was attenuated. Nevertheless, Cain et al. (2012) did not find any
significant differences in the overall task-switching costs between AVGP and nAVGP groups.
1.2 Game Experience and Interference Control
Apart from cognitive flexibility (measured by task-switching cost), task-switching
studies also examine control process, shedding light on the ability to resolve response conflict
as measured by the response-congruency effect. Compared to congruent or neutral targets
that associate with only one response, participants typically have delayed responses and make
more errors in incongruent trials due to the response conflicts triggered by two possible
target-response associations (Sudevan & Taylor, 1987; Schneider, 2015, 2018; Wendt &
Kiesel, 2008; see section 1.6.2 for more information).
EXPERT GAME SKILLS AND EXECUTIVE CONTROL
4
We know little about the relationship between response-congruency effects and action
video game experience/skills. Many existing game studies did not analyse response-
congruency effects even though their empirical designs afforded such measurement (Boot et
al., 2008; Green et al., 2013; Strobach et al., 2012). In the studies that have reported
response-congruency effects in the context of task-switching, results were inconsistent. In a
study by Andrews & Murphy (2006) AVGPs showed smaller response-congruency effects
compared to nAVGPs, but other researchers found no such differences between AVGPs and
nAVGPs (Cain et al., 2012; Dobrowolski et al., 2015).
The ability to control conflicts was also investigated in studies that compare AVGPs
and nAVGPs using other cognitive tasks. However, researchers found no evidence that
AVGPs were better than nAVGPs in terms of the ability to resolve interference from the
conflicting information. For example, the work of Gobet et al. (2014) indicated that both
AVGP and nAVGP groups produced similar amount of congruency effect in a Flanker test.
Consistent with Gobet et al. (2014), in Kowal, Toth, Exton and Campbell (2018), AVGPs
displayed similar Stroop effect to a group of nAVGPs.
1.3 Game Experience and Impulsive Control
Impulsive control, the ability to inhibit instigated and prepotent responses, has been
examined in recent studies on action video games and cognitive performance (Azizi et al.,
2018; Colzato et al., 2013; Decker & Gay, 2011; Deleuze, Christiaens, Nuyens & Billieux,
2017; Littel et al., 2012). For example, studies have shown that compared to non-players,
World of Warcraft (a massively multiplayer online role-playing game, MMORG) players
showed more risky response bias and were more likely to make a response in no-go trials in
which responses should be inhibited, suggesting worse impulsive control in MMORG players
(Decker & Gay, 2011; Littel et al., 2012). In a recent study, Azizi and colleagues (2018)
EXPERT GAME SKILLS AND EXECUTIVE CONTROL
5
provided seminal findings suggesting that although different genres of video game experience
associated with different levels of impulsive control, multi-genre gamers had higher false-
alarm rates and more risk-taking response bias compared to non-gamers in a continuous
performance test that was used to measure sustained attention and impulsive response
inhibition.
Contradictory results were reported in other studies indicating no differences between
AVGPs and nAVGPs in terms of impulsive control (Colzato et al., 2013; see also Dye, Green
& Bavelier, 2009; Metcalf & Pammer, 2014; Mack & Ilg, 2014). For example, Colzato et al.
(2013) compared performance between players who did not play First Person Shooting (FPS)
games and those who played FPS games in a stop-signal task. Colzato et al. (2013) found no
differences between players and non-players, with both groups showing similar RTs and ERs
in the trials that required participants to inhibit responding. Similarly, Metcalf and Pammer
(2014) indicated that non-addicted FPS players did not differ from non-players in the tasks
measuring impulsiveness (e.g., go/no-go task), although addicted FPS players had increased
impulsive responses in no-go trials and more missed detections in go trials.
1.4 Potential Issue on Assessing Game Experience
In short, a large body of previous studies examining the relationship between
executive functioning and action video game experience have produced inconsistent results:
Some empirical studies showed better executive functioning in AVGPs than in nAVGPs,
whereas others failed to find differences. Inconsistent evidence were also from systematic
reviews, meta-analysis studies and large scale correlation studies: Some studies indicated a
positive relationship with large effect size between video game experience and various
cognitive abilities (Powers, Brooks, Aldrich, Palladino, Alfieri, 2013; Toril, Reales &
EXPERT GAME SKILLS AND EXECUTIVE CONTROL
6
Ballesteros, 2014; Wang et al., 2016), but others did not find any meaningful relationships
(Hambrick et al., 2010; Sala, Tatlidil & Gobet, 2018; Unsworth et al., 2015).
One potential limitation could be that many previous studies mainly categorised
participants based on time spent on action video games (e.g., Azizi et al., 2017, Kowal et al.,
2018; Unsworth et al., 2015; Waris et al., 2019). For example, in Kowal et al. (2018),
participants were considered “hardcore players” if they reported having spent more than 23
hours per week in a game, or “casual players” if they reported having spent fewer than seven
hours per week in a game. Likewise, correlational studies applied surveys which asked
participants to indicate the average hours of various video games played per week as an index
of game experience (e.g., options ranging from never,0-1 hours,1-3 hours,3-5 hours,5-10
hours, to 10+ hours in Unsworth et al., 2015). In addition, previous training studies tried to
establish whether action video game play can cause increased cognitive performance by
asking participants to partake in a few hours of supervised video game training (e.g., Boot et
al., 2008; Green et al., 2013).
Researchers typically assume that the longer the time spent playing action video
games, the better the players’ gaming skills should be. However, there is no guarantee that
more video game experience is always equivalent to becoming a game expert. If we consider
playing action video games as a self-regulated training process on gaming skills, then without
any objective control (e.g., players may have different motivations), gaming skills could
hardly be improved and the training effect could vary across individuals. In fact, even under
typical laboratory conditions, effects of game training are hard to verify due to limited
training durations, compliance with training instructions and various training methods (c.f.,
Green, Strobach & Schubert, 2014; Green et al., 2018).
There is no guarantee that a three-hour session of video game play would improve
gaming skills and/or cognitive performance compared to a one-hour training. Although
EXPERT GAME SKILLS AND EXECUTIVE CONTROL
7
studies demonstrated that gaming skills and gaming experience are positively correlated
(Bonny et al., 2016; Röhlcke et al., 2018), the general conclusion that the more one plays, the
higher gaming skills one can get may not apply to every player. It is still possible that people
who spend less time playing video games (e.g., casual players in Kowal et al., 2018) actually
have better gaming skills compared to those who play a great deal (e.g., hardcore players). In
other words, evaluating one’s gaming skills based on the hours of game playing may be
inappropriate. An imprecise classification of expertise may account for the discrepancy in the
results in past video-game literature (Latham, Patston & Tippett, 2013).
Appropriate classifying participants as an expert or novice game player may require
direct and objective assessment of their gaming skills (Latham et al., 2013; Sala et al., 2018).
Gaming skills can be considered as the expertise or talent needed to win a game. Ranking
systems in typical MOBA (multiplayer online battle arena) games such as Defense of the
Ancients (DOTA) and League of Legends (LOL) may provide a more objective measurement
of players gaming skills. MOBA game ranking scores are usually calculated based on ELO
rating (Elo, 1986) which has been used in many traditional sports (e.g., chess), and have been
used to indicate the relative skill levels of players. Taking LOL ranking as an example of the
ELO system, there are nine rank divisions (from Iron division to Challenger division), with
each division representing a different skill level (Figure 1). The system matches players of a
similar skill level within a rank division to play with and against each other. Players within
each division are ranked using a system of points called League Points: League Points
increase per win and decrease per loss. Players will be promoted to a higher division once
they accumulate enough League Points. In contrast, if players lose too many League Points,
they will be demoted. Therefore, we can expect that players in higher rank divisions should
have better gaming skills compared to players in lower rank division.
Recent research has utilized ranking scores in MOBA games to measure AVGPs’
EXPERT GAME SKILLS AND EXECUTIVE CONTROL
8
gaming skills (Bonny, Castaneda & Swanson, 2016; Gong, Ma, Liu, Yan & Yao, 2019;
Kokkinakis, Cowling, Drachen & Wade, 2017; Röhlcke, Bäcklund, Sörman & Jonsson,
2018). In Bonny et al. (2016), DOTA gaming skills were assessed by ranking scores obtained
on the DOTA ranking system. Their results indicated that players with greater DOTA game
expertise were more likely to have faster speed of processing on the task specifically relying
on spatial long-term memory (but see Röhlcke et al., 2017). Using a similar approach,
Kokkinakis et al. (2017) found that LOL gaming skills, as indicated by game rankings,
correlated with fluid intelligence. Empirical evidence was also provided by Gong et al. (2019)
who studied the impact of game expertise on human brain development using resting-state
fMRI. Gong et al. (2019) deliberately selected top-ranking LOL players with reasonably high
standard (top 1.77% of all players worldwide; but note that we used a higher criterion
including players ranking at top 0.15%), and lower-ranking players as a comparison. Their
results showed that compared to lower-ranking players, higher-ranking players had superior
local functional integration in the executive areas and higher levels of local functional
connectivity density in brain regions related to memory and planning, suggesting better
executive functioning in players with higher gaming skill levels (Gong et al., 2019).
Nevertheless, previous studies focusing on players’ gaming skills (assessed by game
ranking system) and cognitive abilities still have two potential limitations. Firstly, many
studies, except Gong et al. (2019), recruited AVGPs without intentionally selecting top-
ranking players, and only required experience in the game or attending/staffing the
tournament; participants in those studies were thus generally of average skill and could
hardly be considered experts (Bonny et al., 2016; Röhlcke et al., 2018; Kokkinakis et al.,
2017). Therefore, their results on gaming expertise and cognitive abilities may not generalize
well to players with the highest game ranking scores.
Secondly, in the previous studies that investigated cognitive abilities across game
EXPERT GAME SKILLS AND EXECUTIVE CONTROL
9
players without controlling game experience (Bonny et al., 2016; Röhlcke et al., 2018;
Kokkinakis et al., 2017), the relationship between gaming skills and cognitive abilities can
be confounded with time spent in the game. In other words, it was unclear if the enhanced
cognitive ability could be related to higher gaming skills or more gaming experience, or both.
Gong et al. (2019) did not control players’ gaming time either, which means any differences
between neural networks in higher-ranking players and lower-ranking players can be
confounded with gaming experience.
1.5 Aim of the Current Study
The present study sought to further investigate and elaborate on the relationship
between actual gaming skills as measured by game ranking system and executive functioning.
We examined whether those with superior gaming skills in an action video game (ranking at
global top 0.15%) would have better executive functions compared to average players after
controlling for self-reported time spent on the action video game. As in previous studies
(Gong et al., 2019; Kokkinakis et al., 2017), we focused on one particular action video
gameLeague of Legends (LOL)for three reasons. First, LOL is currently one of the most
popular action video games in the world. According to Riot (one of the largest American
video game developers and e-sports tournament organizers), as of September 2019 LOL has
nearly eight million concurrent players during peak hours of the day around the world
(https://na.leagueoflegends.com/en/news/game-updates/special-event/join-us-oct-15th-
celebrate-10-years-league).
In addition, the LOL ranking system provides an objective assessment of the actual
skill level of each participant, so that we can make sure those with higher ranks would have
improved game skills. Details of the LOL ranking system can be found on “League system”
(2019).
EXPERT GAME SKILLS AND EXECUTIVE CONTROL
10
Finally, playing MOBA games such as LOL requires a variety of cognitive abilities.
LOL involves two opposing teams of players. Players need to control a game character called
a “champion” with unique spells and battle against a team of other players in a preset battle
arena. In order to destroy the opponent’s base and win the game, players have to constantly
adjust their team and personal tactics according to current battle conditions, while ignoring
bait information that is used for distracting attention. Both of such filters require high
cognitive flexibility and interference control. Moreover, LOL players must not only make
quick decisions about when to cast their spells against opponents, but also when not to; both
types of decisions may require quick reaction time and efficient impulsive control.
We explored whether players who were close to the highest ranking score in LOL
performed differentlycompared to average-ranking LOL playersin cognitive experiments
requiring frequent switching between different tasks while resolving the interference and/or
the ability to resist impulsive responding.
1.6 Stroop-Switching Test
In the present study, we employed a Stroop-switching test, in which participants were
required to switch between the color-naming and word-reading tasks. This classic Stroop-
switching test has been employed by many researchers of cognitive flexibility during the
frequent switching and interference control when conflicts occur (e.g., Allport, Styles &
Hsieh, 1994; Monsell et al., 2000; Kalanthroff & Henrik, 2014; Saban, Gabay & Kalanthroff,
2018; Wu et al., 2015; Yeung & Monsell, 2003).
1.6.1 Cognitive flexibility
Cognitive flexibility in Stroop-switching experiments has been indicated by “task-
switching cost”: Response times (RTs) and error rates (ERs) are increased in trials where the
task switches from that of the preceding trial (e.g., color-naming word-reading) compared
EXPERT GAME SKILLS AND EXECUTIVE CONTROL
11
to task repetitions (e.g., color-naming color-naming; Allport et al., 1994; Kalanthroff &
Henrik, 2014; Saban, Gabay & Kalanthroff, 2018; Yeung & Monsell, 2003). It has been
suggested that fast disengaging from the previous task and efficient preparation for the task in
the upcoming trial would be associated with smaller task-switching costs and thereby with
better cognitive flexibility (e.g., Kramer, Cepeda, Cepeda, 2001; Kray et al., 2012; Mayr et
al., 2014).
Moreover, studies that employed Stroop-switching test have also reported
asymmetrical switching costs: Switching from the relatively harder color-naming task to the
relatively easier word-reading task caused greater task-switching costs than the other way
around (e.g., Allport, Styles & Hsieh, 1994; Wu et al., 2015; but see Monsell et al., 2000;
Yeung & Monsell, 2003). Schneider and Anderson (2010) suggested that the asymmetrical
switch costs arise from sequential difficulty effects. Since performance was impaired after a
difficult trial, response times were longer when switching to an easy task and when repeating
a difficult task leading to asymmetrical switch costs. In addition, asymmetrical switching
costs may reflect the proactive interference from previous tasks: the harder the previous task,
the stronger the proactive interference and thereby greater task-switching costs when
switching to trials with an easy task (Allport, Styles & Hsieh, 1994).
1.6.2 Interference in Stroop-Switching Test
Stroop-switching experiments can consist of three types of target stimuli: neutral
target stimuli, congruent target stimuli, and incongruent target stimuli (Kalanthroff & Henik,
2014; Steinhauser & Hubner, 2009). Neutral target stimuli (e.g., the word RED printed in
black, or meaningless symbols like &&& printed in red or green) have only one task-relevant
feature and are presented in one task; as such, they are mapped to only one response in the
experiment. In contrast, congruent and incongruent target stimuli have features for both tasks.
Congruent target stimuli (e.g., RED printed in red or REDred) have two task-relevant features
EXPERT GAME SKILLS AND EXECUTIVE CONTROL
12
that are mapped to the same response in the two tasks, whereas incongruent target stimuli
(e.g., GREENred) have two task-relevant features that require different responses in the two
tasks. In such an experimental setup, one would typically find two types of interference
effects: response-congruency effects and reverse-facilitation effects (Kalanthroff & Henik,
2014; Steinhauser & Hubner, 2009; see Appendix D for an illustration of both effects).
The response-congruency effects were measured by the performance differences
between trials with incongruent and neutral target stimuli. It has been found that participants
showed increased RTs and ERs in incongruent trials than in neutral trials because the former
is associated with two possible feature-response associations which resulted in interference
(see also Li, Li, Liu, Lages & Stoet, 2019a, 2019b; Guo, Li, Yu, Liu & Li, 2019; Sudevan &
Taylor, 1987; Schneider, 2015, 2018; Wendt & Kiesel, 2008).
The reverse-facilitation effects were measured by the performance differences
between trials with congruent and neutral target stimuli. Although both congruent and neutral
trials afford only one possible response in the experiment, researchers have shown that
neutral stimuli led to faster responses and fewer errors compared to congruent stimuli. This is
because compared to congruent stimuli, neutral stimuli are related to information in one task
causing smaller or no task interference during response retrieval (Kalanthroff & Henik, 2014;
Steinhauser & Hubner, 2009).
1.7 Continuous Performance Test
In the present study, in order to measure impulsive control in top- and lower-ranking
LOL players, we employed a go/no-go Continuous Performance Test (CPT), which has been
used to investigate impulsive performance and action video-game experience (e.g., Metcalf &
Pammer, 2014; Azizi et al., 2018). The CPT test requires participants to respond quickly to
the target stimuli in go trials (i.e., when letters other than “X” are presented) and to inhibit the
EXPERT GAME SKILLS AND EXECUTIVE CONTROL
13
response in no-go trials (i.e., when the letter “X” is presented). According to Egeland and
Kovalik-Gran (2010), four key variables are the mean RTs in the correct go trials; rate of
false alarms (i.e., probability of responses in no-go trials); hit rate (i.e., probability of
responses in go trials); and response bias, which takes both the hit rate and false alarm rate
into account, reflecting the tendency to respond to both targets and non-targets.
Response bias (β) in the CPT test is estimated based on signal detection theory and is
calculated using the formulas shown below. Hit rate (H) is calculated as the proportion of
responses to target stimuli in go trials, and false alarm rate (F) is calculated as the proportion
of responses in no-go trials. Larger βindicates more conservative responding, while smaller β
indicates more risky responding.
1.8 Hypotheses
This study sought to investigate whether top-ranking AVGPs would show superior
performance in the two cognitive tasks, when compared to average-ranking AVGPs with
similar game experience. According to the previous brain imaging results (Gong et al., 2019),
higher ranking in LOL was found to be associated with better executive control. Therefore,
we hypothesized that players who were close to the highest ranking score in LOL may
outperform average players in the tests that measure different aspects of executive
functioning (i.e., cognitive flexibility, interference control and impulsive control).
Specifically, we predicted that compared to average-ranking LOL players, top-
ranking LOL players would produce smaller task-switching costs associated with better
cognitive flexibility in the Stroop-switching test. In addition, top players would show smaller
EXPERT GAME SKILLS AND EXECUTIVE CONTROL
14
response-congruency effects and reverse facilitation effects associated with more efficient
interference control. For the CPT test, we predicted that top-ranking LOL players would
exhibit lower false alarm rate and larger response bias associated with better impulsive
control compared to average-ranking players.
2. Method
2.1 Research ethics
The present study was approved by the Fudan University Department of Psychology
Ethics committee and complied with the Declaration of Helsinki. Prior to the study, all
participants were informed about the procedures of the experiments. All participants provided
written informed consent before taking part in the present study.
2.2 Participants
Top LOL Players. We first recruited thirty-five LOL expert players (2 females; mean
age = 22.8 years, SD = 4.42) ranked higher than the Diamond tier (i.e., Master, GrandMaster
and Challenger tiers) to participate in the present experiments. Fewer than 0.2% of the LOL
players have a chance to reach beyond the Diamond tier (Figure 1); therefore, we called them
“top-ranking LOL players” or “top players.” Top players were recruited through
advertisements posted in online game communities, and via word of mouth.
EXPERT GAME SKILLS AND EXECUTIVE CONTROL
15
Figure 1. Distributions of the nine LOL rank divisions. The histogram displays the player
distribution by divisions. Iron is the lowest tier and Challenger is the highest tier. Data were
gathered by “Rank distribution” (2019) considering all players around the world.
Average LOL Players. We recruited a large group of average-ranking LOL players
(N > 50). Average players had LOL ranks ranging from Iron to Diamond tier, and were
assigned to a waiting list. After we collected all top players, 35 average-ranking LOL players
(four females; mean age = 23.1 years, SD = 4.59) were invited from the waiting list (N> 50):
We deliberately selected a group of players whose game experience, age and gender
distribution were not significantly different from the top players (Table 1).
All participants provided their LOL account names before taking part, so that
researchers could verify their rank division in the game using a third-party smartphone app
called WeGame. All participants reported that LOL was the only action video game they
played on a weekly basis. All reported having normal or corrected-to-normal visual acuity.
Participants were paid 100 (≈ $12) for their participation.
EXPERT GAME SKILLS AND EXECUTIVE CONTROL
16
Demographics
Top players
Average players
Sig.
difference
Gender
M/F = 33/2
M/F = 31/4
p> .05
Age (years)
22.8 (SD = 4.42)
23.1 (SD = 4.55)
p> .05
Educational Levels
5.08 (SD = 1.12)
5.71 (SD = 0.93)
p= .012
LOL Game
Experience
Years
5.07 (SD = 1.67)
5.14 (SD = 1.57)
p> .05
Hours Per Week
13.26 (SD = 8.51)
14.69 (SD = 8.62)
p> .05
Rank Division
8.31 (SD = 0.90)
3.83 (SD = 0.92)
p< .001
2.3 Apparatus
Both the Stroop-switching test and the CPT test were programmed using PsyToolkit
(Stoet, 2010, 2017). All stimuli were presented at the center of a 21-inch Dell computer
monitor with black background. A QWERTY keyboard was used to record participants’
responses with ±1 ms precision. In the Stroop-switching test, participants gave left and right
responses by pressing the “A” key or “L” key on the keyboard with their left and right index
finger, respectively. In the CPT test, participants pressed the spacebar on the keyboard to give
their response.
2.4 Stroop-Switching Test
2.4.1 Stimuli
Our Stroop-switching test was identical to Kalanthroff and Henrik (2014) except that
the word-reading task cue and all target stimuli were written in a different language. As per a
EXPERT GAME SKILLS AND EXECUTIVE CONTROL
17
previous study that showed language barrier might obscure the task-switching cost (Li et al.,
2019a), we applied Chinese characters as the task cue and target stimuli in a sample of
Chinese participants. Two critical Chinese characters were 绿(GREEN) and (RED)
displayed in red and green, so that 绿(GREEN) displayed in green and (RED) displayed in
red were congruent target stimuli and they were incongruent target stimuli if 绿(GREEN)
was shown in red and (RED) was shown in green. The experiment also included two
simplified Chinese characters 绿(GREEN) and (RED) as well as two traditional Chinese
characters (GREEN) and (RED) displayed in white serving as neutral target stimuli in
the word-reading task; and two meaningless characters and displayed in red and green
serving as neutral target stimuli in the color-naming task. The size of each character was 38
mm × 38 mm. In total, there were twelve combinations of Chinese characters and colors: two
congruent stimuli, two incongruent stimuli and eight neutral stimuli (four for each task; see
Figure 2b). A circular ring consisting of two white Chinese characters 绿(GREEN) and
(RED) served as the word-reading task cue and a circular ring consisting of colored dots
served as the color-naming task cue. The diameter of each task cue was 71mm (Figure 2a). In
the word-reading task, participants had to decide whether a target stimulus was word red or
green (red press the left key; green press the right key). In the color-naming task,
participants had to decide whether a target stimulus was displayed in red or green (red
press the left key; green press the right key).
2.4.2 Procedure
In each trial, a circular task cue and a target Chinese character appeared
simultaneously and were located at the center of the screen. The target always appeared
inside the task cue, with both staying on the screen until a response was made or a maximum
of 2500 ms was exceeded. A correct response would trigger the next trial after an intertrial
EXPERT GAME SKILLS AND EXECUTIVE CONTROL
18
interval of 300 ms. If the participant failed to respond within 2500 ms, a “time out” message
appeared for 3 s. If the participants made an incorrect response, a “mistake” message
appeared for 3 s. The Stroop-switching experiment consisted of three training blocks of 32
trials followed by two experimental blocks of 128 trials. Note that the present Stroop-
switching test was demanding because both the task cue and target stimuli were presented
simultaneously. In a pilot study we found that training with three blocks of 32 trails helped
participants to better recall the task rule and show reasonable accuracy.
Figure 2. Illustration of the task rules and target stimuli in the Stroop-switching test. (a) The
color-naming task cue was a circular ring consisting of red and green dots, and the word-
reading task cue was a circular ring consisting of Chinese characters in white. (b) Target
stimuli were twelve Chinese characters displayed in red, green, and white. LEFT and RIGHT
correspond to pressing the “A” and “L” keys on the QWERTY keyboard, respectively.
EXPERT GAME SKILLS AND EXECUTIVE CONTROL
19
2.5 Continuous Performance Test (CPT)
2.5.1 Stimuli and Procedure
Our CPT test was similar to the CPT in previous studies (e.g., Azizi et al., 2018),
except that we provided feedback immediately following either correct, incorrect or delayed
response. This is because in LOL players receive immediate feedback on almost all of their
actions. We also applied a narrower response window (450 ms), which was extremely
demanding, in order to simulate the rapid response mode in the LOL game.
The target stimuli were green (RGB 0,255,0) English letters (32 mm × 30 mm). In
each trial of the CPT test, the stimuli appeared in the center of the screen, one at a time (as
one trial), for 250 ms. The trials were presented in 18 consecutive blocks where each block
contained inter-stimulus intervals (ISI) of 1000, 2000, or 4000 ms in a random order.
Transitions from one block to the next occurred unannounced and without delay. Participants
were asked to press the spacebar on the keyboard as quickly as possible whenever letters
except “X” show up on the screen (go trial) and to suppress the response when “X” shows up
(no-go trial). If participants respond too slow in a go trial (> 450 ms) a “timeout” message
appeared for 300 ms. If the participants could not suppress the response in the no-go trials
(i.e., participants pressed the spacebar within 700 ms), a “mistake” message appeared for 300
ms. The CPT test consisted of a block of 10 training trials and 18 blocks of 20 experimental
trials.
2.6 General Procedure
Experiments took place in a psychology lab at Fudan University. After participants
provided demographics information, they sat with their eyes at a distance of approximately
80 cm from the computer screen. All participants were asked to complete the Stroop-
EXPERT GAME SKILLS AND EXECUTIVE CONTROL
20
switching test before the continuous performance test. They received verbal and on-screen
instructions before each test. All participants took a five-minute break between the tests.
3. Results
The data that support the findings of this study are openly available in Open Science
Framework at https://osf.io/mygz7/?view_only=ef6eb4e8d88b4653909776fdbbecfa01.
3.1 Stroop-Switching Test
All participants performed better than chance, with error rates lower than 15% in the
Stroop-switching test. Error trials were excluded from all RT analyses. For the analyses of the
Stroop-switching experiment, all training trials, the first trial of each block, and the trial
immediately following incorrect trials were excluded. If participants made a mistake in trial
n–1, trial n cannot be categorized as a task-switch trial or task-repeat trial. As a result, we
excluded 9.12% of trials. Mean RTs and ERs for each trial condition and group of players are
shown in Figure 3, Figure 4 and listed in the Appendix A.
We first applied a 2 (Trial transition: repeat, switch) × 2 (Task: color-naming, word-
reading) × 3 (Congruency: congruent, incongruent, and neutral) × 2 (Player group: top,
average) mixed-factors multivariate analyses of variance (MANOVA), with Player group as a
between-subjects factor and other factors as within-subject factors. Two dependent variables
were RTs and ERs. We listed the results of the MANOVA in Appendix B. In short, the main
effects of Trial transitions, Congruency and Player groups were significant (p<.001), but the
main effect of Task was not significant (p= .643) with no differences between the color-
naming and word-reading task. There were also two-way and three-way interactions (p
<.001). In order to further understand whether top and average LOL players differed in the
Stroop-switching performance, we submitted their RT and ER data in two separate analyses
EXPERT GAME SKILLS AND EXECUTIVE CONTROL
21
of variance (ANOVA) with the design identical to the MANOVA. The results of the
ANOVAs are summarized in Table 2 and illustrated in Figure 3 and Figure 4.
RT
ER
Effect
F
df
p
η2p
F
df
p
η2p
Player group (G)
0.24
1, 68
.623
.004
15.67
1, 68
<.001
.187
Trial transition (TT)
60.62
1, 68
<.001
.471
157.38
1, 68
<.001
.698
Congruency (C)
414.75
2, 136
<.001
.859
144.22
2, 136
<.001
.678
Task (TK)
<.01
1, 68
.962
<.001
5.86
1, 68
.018
.079
G×TT
12.48
1, 68
<.001
.155
8.09
1, 68
.006
.106
G×C
0.13
2, 136
.876
.002
10.02
2, 136
<.001
.128
G×TK
0.13
1, 68
.718
.002
1.83
1, 68
.179
.026
TT×C
14.14
2, 136
<.001
.172
71.90
2, 136
<.001
.514
TT×TK
12.07
1, 68
<.001
.151
8.93
1, 68
.004
.116
C×TK
8.68
2, 136
<.001
.113
10.69
2, 136
<.001
.136
G×TT×C
3.76
2, 136
<.001
.052
1.74
2, 136
.179
.025
G×TT×TK
0.27
1, 68
.599
.004
0.33
1, 68
.565
.005
G×C×TK
0.45
2, 136
.654
.006
1.14
2, 136
.322
.017
TT×C×TK
2.22
2, 136
.112
.032
16.25
2, 136
<.001
.193
G×TT×C×TK
0.52
2, 136
.595
.007
2.11
2, 136
.126
.030
3.1.1 Analysis of RT
The mean RTs were similar for top (948 ms) and average players (963 ms), and in the
color-naming task (958) and word-reading task (956 ms). We found significantly longer RT
in switch trials (992 ms) than in repeat trials (917 ms). In the following, post-hoc pairwise t-
test comparisons were used for multiple comparisons after Holm (1979). The results showed
that there were significant RT differences between neutral (874 ms) and congruent trials (910
EXPERT GAME SKILLS AND EXECUTIVE CONTROL
22
ms), p<.001, d= -0.70; between incongruent trials (1105 ms) and congruent trials, p<.001, d
= 2.41; and between neutral trials and incongruent trials, p<.001, d= -2.83.
There was a significant interaction between Trial transition and Task type. Participants
showed larger task-switching costs when switching to the color-naming task (switch - repeat
= 94 ms) compared to switching to the word-reading task (switch - repeat = 54 ms); the
difference was significant, p= .003, d= 0.36. There was a significant interaction between
Trial transition and Congruency. The task-switching costs were smaller in the trials with
congruent target stimuli (switch - repeat = 32 ms) compared to trials with neutral target
stimuli (switch - repeat = 95 ms), p< .001, d= -0.44, and trials with incongruent target
stimuli (switch - repeat = 94 ms), p<.001, d= -0.74. However, switch costs were not
statistically different between neutral and incongruent conditions (p>.05).
As predicted, we found a significant two-way interaction between Player group and
Trial transition; task-switching costs were smaller in top group (switch - repeat = 37 ms) than
in average group (switch - repeat = 91 ms). The three-way interaction between Player group,
Trial transition and Congruency was also significant. In order to better understand whether
there were group differences in different conditions, we compared the task-switching costs
(switchrepeat) between top players and average players in congruent, neutral, and
incongruent conditions. The results showed that top players had smaller task-switching costs
compared to average players in the congruent condition (10 ms vs. 54 ms, p= .043, d=
0.49), neutral condition (77 ms vs. 126 ms, p= .041, d=0.57), and incongruent condition
(40 ms vs. 150 ms, p= .006, d=0.77). The switching-cost difference between groups of top
and average players was larger in the incongruent condition compared to the neutral and
congruent conditions. There was also a significant interaction between task and congruency.
However, such interaction was not predicted. No other effects in RT reached statistical
significance.
EXPERT GAME SKILLS AND EXECUTIVE CONTROL
23
Figure 3. RT results of the Stroop-switching experiment. (a) Bar charts display mean RTs in
each trial condition (repeatcongruent, switchcongruent, repeatneutral, switchneutral,
repeatincongruent, switchincongruent). The error bars indicate ±1 SEM. (b) The violin plot
illustrates RT distribution for the congruent, neutral, and incongruent condition in each player
group (top players, average players). Jittered dots inside the violin plots represent average
RTs for each participant. The black horizontal bar and the box around it represent the mean
and 50 % CI of the mean in each condition, respectively. (c) The violin plot illustrates RT
task-switching costs (TSC) for the congruent, neutral, and incongruent condition in each
player group. The difference between top and average players in the TSC was denoted by
symbol “Δ”. Con = Congruent; Neu = Neutral; Inc = Incongruent; Rep = Repeat; Swi =
Switch. **p< .01, *p< .05.
3.1.2 Analysis of ER
An equivalent four-way ANOVA with mixed effects was conducted on mean ERs
between and within conditions. The results of this analysis are summarized in Table 2 and
EXPERT GAME SKILLS AND EXECUTIVE CONTROL
24
illustrated in Figure 4. Top players had lower ER (5.02%) compared to average players
(8.81%). ERs were higher in switch trials (8.04%) than in repeat trials (5.09%), and in the
word-reading task (7.21%) than in the color-naming task (6.63%). For the main effect of
Congruency, pairwise t-test comparisons showed a significant ER difference between neutral
(3.15%) and congruent trials (2.26%), p=.023, d= 0.28; between incongruent (15.35%) and
congruent trials, p< .001, d= 1.66; and between incongruent and neutral trials, p< .001, d=
1.55.
As predicted, there were significant two-way interactions between Player group and Trial
transition, and between Player group and Congruency. Top players (2.32%) showed smaller
task-switching costs (switchrepeat) compared to average players (4.29%), p= .020, d=
0.57.
Moreover, in order to better understand whether there were group differences in the
effect of Congruency, we analyzed the reverse-facilitation effects (congruentneutral) and
response-congruency effects (incongruentneutral) in both player groups. The results showed
that the reverse-facilitation effects were significant in the group of top players (1.18%, p
= .036, d= –0.42), but were not significant in the group of average players (.59%, p= .344,
d= –0.12). The response-congruency effects were significant in both the top players (9.44%,
p< .001, d= 1.26) and average players (14.95%, p< .001, d= 1.26); however, the effect was
smaller in the top players than in the average players, p=.003, d=0.74. Further analysis
showed that top players had significantly lower ER compared to average players in congruent
trials (1.07% vs 3.45%, p= .005, d=0.77), and incongruent trials (11.71% vs 19.00%, p
< .001, d=0.93), but in neutral trials the group difference was not statistically significant
(2.25% vs 4.05%, p= 0.109, d=0.38).
We found a significant two-way interaction between Trial transition and Congruency.
Task-switching costs were larger in incongruent condition (switchrepeat = 6.27%) than in
EXPERT GAME SKILLS AND EXECUTIVE CONTROL
25
neutral condition (switchrepeat = 2.27%), p< .001, d= 0.43, and congruent condition
(switchrepeat = 1.44%), p< .001, d= 0.51. Task switch costs were similar in congruent and
neutral conditions, p= .137.
The interaction between Trial transition and Task type was also significant. Participants
showed significant task-switching costs when switching to the color-naming task (switch
repeat = 3.94%, p<.001, d= 0.84) and when switching to the word-reading task (switch
repeat = 2.68 %, p<.001, d= 0.50), but the difference between the two tasks did not reach
statistical significance, p= .124.
The significant two-way interaction between Task and Congruency, and the significant
three-way interaction between Trial transition, Congruency and Task were not predicted. No
other effects in ER reached statistical significance.
Figure 4. ER results of the Stroop-switching experiment. The bar graphs display ERs of each
trial condition (repeatcongruent, switchcongruent, repeatneutral, switchneutral, repeat
incongruent, switchincongruent). The error bars indicate ±1 SEM. Con = Congruent; Neu =
Neutral; Inc = Incongruent; Rep = Repeat; Swi = Switch.
3.2 Continuous Performance Test
In the CPT test, we focused on four dependent variables: mean RT, the hit rate in the
go trials, the rate of false alarms in the nogo trials, and the response bias. The descriptive data
EXPERT GAME SKILLS AND EXECUTIVE CONTROL
26
and correlations between the four dependent variables were listed in Appendix C. All training
trials and go trials with RTs that were shorter than 150 ms were excluded. The results of the
CPT test are illustrated in Figure 5.
We conducted a one-way MANOVA to test the difference between top and average
players on the linear combination of the four dependent variables. The results showed a
significant multivariate effect for the four dependent variables as a group in relation to the
Player groups. In general, top players performed differently to average players, F(3, 68) =
3.09, p= .007, Pillai’s Trace = .19, η2p= .193. We studied the univariate effects in order to
examine whether top and average players differed on each dependent variable (Figure 5). We
found that the group of top players did not differ from the group of average group in mean
RT (F< 1, p> .05), and response bias, F(1, 68) = 1.01, p= .319. However, top players had a
significantly lower false alarm rate (38.46%) compared to average players (48.72%), F(1, 68)
= 8.34, p= .005, η2p= .109. Moreover, top players had significantly higher hit rate (91. 28%)
compared to average players (87.54%), F(1, 68) = 9.30, p= .003, η2p= .120.
EXPERT GAME SKILLS AND EXECUTIVE CONTROL
27
Figure 5. Results of the Continuous Performance Test. (a) The violin plot illustrates RT
distribution in go trials for the top and average players. The jittered dots inside each bean
represent averaged RTs of each participant. The black horizontal bar and the box around it
represent the mean RT and 50% CI of the mean RT in each group, respectively. (b) The
violin plot illustrates the distribution of response bias (β) for the top and average players. (c)
The violin plot illustrates the distribution of false-alarm rate for the top and average players.
(d) The violin plot illustrates the distribution of hit rate for the top and average players. **p
< .01.
EXPERT GAME SKILLS AND EXECUTIVE CONTROL
28
4. Discussion
The main purpose of the present study was to investigate whether video game experts
(i.e., top-ranking LOL players) have superior cognitive ability than average players after
controlling for self-reported gaming times. We applied a Stroop-switching test in order to
examine players’ abilities of flexible switching and interference control. Moreover, we also
applied a continuous performance test in order to examine players’ impulsive control. We
found that top-ranking players had better performance in both the Stroop-switching test and
continuous performance test compared to average-ranking players.
4.1 Cognitive Flexibility and Gaming Skill
In line with our hypothesis, both the top- and average-ranking LOL players showed
significant RT and ER task-switching costs, but the task-switching costs were smaller in the
group of top-ranking players. Importantly, in top-ranking players the decreases in switching
costs were not accompanied by the increased error rates given that top players showed overall
fewer errors compared to average-ranking players. Therefore, the advantage of top-ranking
players in the Stroop-switching test cannot be explained by the speed-accuracy trade off.
Top-ranking players might have better cognitive flexibility, possibly with fast task-set
reconfiguration when switching between tasks (Grange & Houghton, 2014; Kiesel et al.,
2010; Vandierendonck et al., 2010). Our results are inline with the previous fMRI study
(Gong et al., 2019) that better action video gaming skills were linked to improved executive
control abilities such as cognitive flexibility.
It is less likely that we can attribute the flexible switching in the top-ranking players
to extended gaming experience alone, because there were no differences in the gaming tenure
and times (i.e., years of play and hours of play per week) between top and average players. It
has been suggested that task-switching costs are sensitive to training and can be reduced
significantly after a few experimental sessions (e.g., reduced switching costs after six training
EXPERT GAME SKILLS AND EXECUTIVE CONTROL
29
sections in Wendt, Kähler, Luna-Rodriguez & Jacobsen, 2017; see also Strobach, Liepelt,
Schubert & Kiesel, 2011). However, in a study by Stoet and Snyder (2007) who asked four
participants to practice task switching for more than 20,000 trials, researchers found that 1
participant showed reduced task-switching costs whereas the cost was roughly constant or
even increased in other participants. Perhaps there is an interplay between task-switching
strategies and duration of training in the effect of task-switching training. Similarly, it is also
possible that game skills interplay with game experience, influencing the training effect of
game playing. That is, only top-ranking players can have improved cognitive flexibility and
facilitated task-switching performance as gaming experience increases (see Section 4.4 for
further discussion).
4.1.1 Asymmetrical switching costs
Inconsistent with our predictions, top players indicated asymmetrical task-switching
costs similar to average players. Unexpectedly, both groups had smaller task-switching costs
when switching from the relatively easier word-reading task to the relatively harder color-
naming task than the other way around, which contradicted previous results (e.g., Allport,
Styles & Hsieh, 1994; Wu et al., 2015; but see Monsell, et al., 2000; Yeung & Monsell, 2003).
Task-switching costs have at least two origins (cf., Grange & Houghton, 2014; Kiesel
et al. 2010; Vandierendonck et al., 2010). Firstly, task-switching costs can be due to proactive
interference from the task set activated in the previous trial with a different task (Allport et al.,
1994; Allport & Wylie, 2000; Waszak & Hommel, 2007). For example, in the present study,
when participants switched from a color-naming task to a word-reading task, the color feature
became irrelevant and interfered with the processing of the relevant word feature, thereby
delaying responses. In addition, task-switching costs can reflect the time needed to
reconfigure or to retrieve a relevant task set from memory before participants could give a
EXPERT GAME SKILLS AND EXECUTIVE CONTROL
30
response in task-switching trials (Meiran, 2000; Monsell & Mizon, 2006; Rogers & Monsell,
1995).
The phenomenon that switching from a harder task to a relatively easier task causes
larger task-switching costs rather than the other way around has been attributed to the larger
and persisting proactive interference elicited by the harder task compared to the easier task
(Allport et al., 1994; Wu et al., 2015). However, the task cues in the present study were very
transparent in task meaning: The word-reading task cue was a circular ring consisting of the
words “green” and “red” in white in Chinese, and the color-naming task cue was a circular
ring consisting of green and red dots. Those cues may have helped participants focus on the
task-relevant feature in each trial. Therefore, the amount of interference from the irrelevant
task may be reduced during task-switching (Gade & Steinhauser, 2019). Alternatively, it is
likely that in our study, the task-switching costs are primarily caused by the additional task-
reconfiguration process in switch trials. Task-reconfiguration may take longer when
switching to a hard task, associated with large RT switching costs. Similar results have been
reported by Yeung & Monsell (2003), suggesting that when the interference between tasks
was reduced, participants showed reversed asymmetrical task-switching costs, which means
larger switching costs when switching to the color-naming task compared to the word-reading
task.
In previous task-switching studies that investigated the task-switching performance in
the neutral target condition and the bivalent target condition, researchers typically found
smaller task-switching costs for neutral target stimuli compared to bivalent target stimuli (e.g.,
Allport & Wylie, 2000; Rogers & Monsell, 1995; Wylie & Allport, 2000). However, in our
study participants showed task-switching costs in trials with neutral target stimuli (95 ms) at a
similar level to trials with incongruent target stimuli (96 ms). An in-depth discussion is
beyond the scope of the present study, but perhaps our neutral target stimuli contained
EXPERT GAME SKILLS AND EXECUTIVE CONTROL
31
redundant information irrelevant to both tasks. For example, the structure of the two
meaningless Chinese characters was very complex (Figure 2b), which may also attract
attention and could delay the task-set reconfiguration process.
4.2 Resolving Interference and Gaming Skills
We found differences between top-ranking players and average-ranking players in ER
response-congruency effects, which was in line with our prediction. The response-
congruency effects reflect the involvement of cognitive efforts in resolving interference
between two potential feature-response associations in incongruent trials (Li et al., 2019a,
2019b; Sudevan & Taylor, 1987; Schneider, 2015, 2018; Wendt & Kiesel, 2008). Top-
ranking LOL players appeared to have better abilities to control and resolve interference
during response selections, inconsistent with a number of previous studies (Cain et al., 2012;
Gobet et al., 2014; Kowal et al., 2018; but see Andrews & Murphy, 2006). However, we
found that the advantages of top players were limited in accuracy. In other words, although
top-ranking players made fewer errors compared to average-ranking players in trials with
incongruent targets, both groups of players responded more slowly in incongruent trials than
in neutral trials. Top players may take a long time to complete the rule-based feature-
categorization process before they made a response in some trials, whereas average players
may even occasionally forget the task rules so that they failed to categorize the relevant target
feature and therefore made frequent errors in a large proportion of incongruent trials.
Another important finding was that top players responded with higher accuracy in the
congruent condition compared to the neutral condition, showing a significant facilitation
effect. In contrast, average players showed no ER differences between congruent and neutral
conditions. The ER results were inconsistent with previous results suggesting that participants
typically have higher ERs in congruent trials than in neutral trials because of the extra task
EXPERT GAME SKILLS AND EXECUTIVE CONTROL
32
interferences: Congruent stimuli consisted of features and afforded responses for the two
tasks, whereas neutral stimuli appeared in only one task (Kalanthroff & Henik, 2014;
Steinhauser & Hubner, 2009). Among top-ranking players, both relevant and irrelevant
features of a congruent target were processed efficiently and may have given rise to a
synergistic effect on response retrieval that facilitated response accuracy. However, average
players made generally more mistakes than top players in each trial condition, which may
account for the null effect. Note that the difference between top- and average-ranking players
was limited in accuracy. Both groups had similar RT results and showed delayed responses in
congruent trials than in neutral trials, replicating previous studies indicating a reverse-
facilitation effect (Kalanthroff & Henik, 2014; Steinhauser & Hubner, 2009).
Overall, our results suggested that top players responded more accurately in the task-
switching experiment compared to average players. We suggest that top players have better
cognitive ability to resolve task interference exhibiting higher accuracy but similar speed of
response to average players.
4.3 Impulsive Control and Gaming Skills
We used a continuous performance test to measure participants’ impulsive control.
The results showed that top- and average-ranking LOL players had no difference in RTs in
the go trials. Normally in CPT testing, participants can respond until the next target/non-
target onset or having a response window of at least 1000 ms (Azizi et al., 2018; Egeland &
Kovalik-Gran, 2010). With a narrower response window of 450 ms, our participants
experienced more time pressure and may be more motivated to respond faster in each go trial,
which may attenuate the RT difference between top and average players. In addition, we
found no difference between groups on response bias; this finding provided evidence that top
players and average players had similar response tendencies.
EXPERT GAME SKILLS AND EXECUTIVE CONTROL
33
However, compared to average players, top players had significantly higher hit rates
and lower false alarm rates, suggesting that top-ranking players were more able to
differentiate between targets and non-targets which may be associated with better impulsive
control. Our results were inconsistent with previous studies either showing no differences
between AVGPs and nAVGPs (Metcalf & Pammer, 2014; Colzato et al., 2013), or showing
that AVGPs (Decker & Gay, 2011; Littel et al. 2012) and players with multi-genre gaming
experience (inculding action video games; Azizi et al., 2018) had worse impulsive control
compared to nAVGPs.
4.4 Top Players and Top Performance
Two possible interpretations were proposed for the results of the present study. First,
it is possible that only individuals who have superior cognitive abilities (e.g., cognitive
flexibility, interference, and impulsive control) can achieve a higher ranking in the game or
become a top-ranking player. In other words, video game training and experience may not be
the key point; all top players may simply be the cognitively gifted individuals. Boot and
colleagues (2008) proposed a similar explanation, arguing that playing video games may not
improve cognitive abilities; instead, people with better cognitive abilities may be more likely
to play video games because they can do better than others in the game, reflecting a self-
selected process into video game play.
The second potential explanation could be that cognitive flexibility, interference, and
impulsive control can be better trained and improved by video game experience in players
who have a strong motivation to constantly hone their game skills and earn promotions. This
is because winning at the highest level of competition would require players to mobilize a
variety of cognitive resources. In contrast, if a player only wants to relax and pass the time by
playing games, the cognitive resources will not be fully mobilized and cognitive abilities may
not be enhanced.
EXPERT GAME SKILLS AND EXECUTIVE CONTROL
34
Taking impulsive control as an example, at the primary/average level of the
competitions in LOL, false alarms or impulsive behaviors (e.g., wasting a spell against an
unimportant target) are not punished by opponents because the spells in LOL have relatively
short cooldown times. However, during the highest level of competition, top players have to
always take advantage of opponents’ cooling down time despite its brevity for harassing or
attacking. Therefore, reducing impulsive behaviors and casting spells only on the most
valuable targets are important goals for top players. If players do not have the motivation to
win a game and get top ranks, they would not tend to improve their gaming techniques
related to impulsive control. Accordingly, we suggest that players’ motivation toward video
games may determine how games impact their cognition.
Although top players can be highly motivated to get higher rankings in the game,
their motivation to game playing might be different from those with video game addiction.
Previous studies have already shown that game addiction can moderate the impact of gaming
experience on cognition in a negative way: Addicted video game players tend to have worse
cognitive abilities compared to non-addicted players and non-players (Metcalf & Pammer,
2014; see a review by Ioannidis et al., 2019). For example, Metcalf and Pammer (2014)
showed that addicted FPS game players had higher levels of disinhibition and higher trait
impulsivity compared to non-addicted players and non-players whereas no differences were
found between non-addicted FPS players and non-players in neuropsychological deficits
associated with impulsivity. According to Metcalf and Pammer (2014) results, it is unlikely
that top players in our study were addicted to the game because we found that top players had
better executive control compared to average players. Future research should compare
between top-ranking players and addicted players in order to understand their motivation and
attitudes towards video games playing.
EXPERT GAME SKILLS AND EXECUTIVE CONTROL
35
Whether individual differences in players’ cognitive functions predetermined the
variances in game skills, or players’ motivation to win a game can help to improve cognitive
functions, or both, warrants further investigation. For example, future training studies may
ask nAVGPs varying in cognitive functioning to practice MOBA games (e.g., LOL) for some
period of time, whilst controlling motivation levels. If the game skills are related to cognitive
functions, those with better cognitive ability would show more improved performance in the
MOBA games as training duration increases, indicating a higher probability of reaching high
ranks. Similarly, by controlling cognitive ability, one can also investigate how players’
motivation interplay with the training effect on video games.
Our results also provide some insights for the e-sports industry: Perhaps performance
in similar cognitive experiments could be used as an important criterion when selecting
professional game players in the future.
4.5 Conclusion and Future Directions
Few researchers to date have studied whether players with different game skill levels
would differ in their cognitive abilities (e,g., Bonny et al., 2016; Gong et al., 2019;
Kokkinakis et al., 2017; Röhlcke et al., 2018). This may be due in part to the difficulty of
finding experts who are willing to participate in research (Gobet et al., 2014). The present
study provided evidence that top-ranking LOL players exhibited greater cognitive flexibility,
improved skills at resolving interference, and better impulsive control, even after we
controlled for self-reported gaming tenure and time. Since these cognitive abilities are core
aspects of executive functions (Diamond, 2013), we suggest that top players’ superior
performance in cognitive tasks may be related to better executive functioning (see also Bonny
et al., 2016; Gong et al., 2019). Moreover, our results indicated that compared to a few weeks
of game training or extended gaming experience, players who are highly skilled in video
EXPERT GAME SKILLS AND EXECUTIVE CONTROL
36
games may be more likely to have improved cognitive abilities.
One limitation of the present study is that we did not control players’ gaming
experience other than with League of Legends (LOL). Although all players reported LOL
was the only action video game they played on a weekly basis, it is still possible that skills
learned in other games can also interfere with our results. Previous studies suggested that
game genres may require or train different cognitive abilities (Dobrowolski et al., 2015; Azizi
et al., 2018; Deleuze et al., 2017); future studies can therefore focus on top-ranking players in
different video game genres and their cognitive performance.
The present study primarily aimed to explore whether cognitive abilities were
improved in AVGPs with superior game skills but similar hours of playing to a group of
AVGPs with average game skills. Therefore, we did not include nAVGPs (i.e., participants
without action video game experience). Future studies can compare the difference between
nAVGPs, average- and top-ranking AVGPs to further isolate the impact of game skills from
game experience on executive functioning. Nevertheless, our results highlight the importance
of having a robust classification of people with gaming skills and those without in future
experiments (cf. Latham et al., 2013; Sala et al., 2018). This may provide an avenue for better
understanding of how action video game training could help boost certain cognitive abilities.
EXPERT GAME SKILLS AND EXECUTIVE CONTROL
37
Reference
Andrews G., Murphy K. (2006). Does video-game playing improve executive function? In:
Vanchevsky MA (Eds). Frontiers in Cognitive Psychology. (pp. 145–161). Nova Science
Publishers.
Allport, A., Styles, E. A., & Hsieh, S. (1994). Shifting intentional set: Exploring the dynamic
control of tasks. In C. Umilta & M. Moscovitch (Eds.), Conscious and nonconscious
information processing: Attention and performance XV (pp. 421–452). Cambridge, MA:
MIT Press.
Allport, A., & Wylie, G. (2000). Task-switching, stimulus–response bindings, and negative
priming. In S. Monsell & J. Driver (Eds.), Control of cognitive processes: Attention and
performance XVIII (pp. 36–70). Cambridge, MA: MIT Press.
Azizi, E., Abel, L., & Stainer, M. (2017). The influence of action video game playing on eye
movement behaviour during visual search in abstract, in-game and natural scenes.
Attention, Perception, & Psychophysics, 79(2), 484-497..
Azizi, E., Stainer, M., & Abel, L. (2018). Is experience in multi-genre video game playing
accompanied by impulsivity? Acta Psychologica,190, 78-84. doi:
10.1016/j.actpsy.2018.07.006
Bediou, B., Adams, D., Mayer, R., Tipton, E., Green, C., & Bavelier, D. (2018). Meta-
analysis of action video game impact on perceptual, attentional, and cognitive skills.
Psychological Bulletin, 144(1), 77-110. doi: 10.1037/bul0000130
Bonny, J. W., Castaneda, L. M., & Swanson, T. (2016). Using an international gaming
tournament to study individual differences in MOBA expertise and cognitive skills. In
Proceedings of the SIGCHI conference on human factors in computing systems. San Jose,
CA, (pp. 3473–3484). http://dx.doi.org/10.1145/2858036.2858190
Boot, W. R., Kramer, A. F., Simons, D. J., Fabiani, M., & Gratton, G. (2008). The effects of
video game playing on attention, memory, and executive control. Acta Psychologica, 129,
387–398. http://dx.doi.org/10.1016/j.actpsy.2008.09.005
Cain, M. S., Landau, A. N., & Shimamura, A. P. (2012). Action video game experience
reduces the cost of switching tasks. Attention, Perception & Psychophysics, 74, 641–647.
http://dx.doi.org/10.3758/s13414-012-0284-1
Castel, A. D., Pratt, J., & Drummond, E. (2005). The effects of action video game experience
on the time course of inhibition of return and the efficiency of visual search. Acta
Psychologica, 119, 217–230. http://dx.doi.org/10.1016/j.actpsy.2005.02.004
Colzato, L. S., van den Wildenberg, W. P. M., Zmigrod, S., & Hommel, B. (2013). Action
video gaming and cognitive control: Playing first person shooter games is associated
with improvement in working memory but not action inhibition. Psychological Research,
77, 234–239. http://dx.doi.org/10.1007/s00426-012-0415-2.
Decker, S. A., & Gay, J. N. (2011). Cognitive-bias toward gaming-related words and
disinhibition in World of Warcraft gamers. Computers in Human Behavior, 27, 798–810.
Deleuze, J., Christiaens, M., Nuyens, F., & Billieux, J. (2017). Shoot at first sight! First
person shooter players display reduced reaction time and compromised inhibitory control
in comparison to other video game players. Computers in Human Behavior, 72, 570–576.
Diamond, A. (2013). Executive Functions. Annual Review of Psychology, 64, 135–168.
Dobrowolski, P., Hanusz, K., Sobczyk, B., Skorko, M., & Wiatrow, A. (2015). Cognitive
enhancement in video game players: The role of video game genre. Computers in Human
Behavior,44, 59-63. doi: 10.1016/j.chb.2014.11.051
Dye, M. W. G., Green, C. S., & Bavelier, D. (2009b). Increasing speed of processing with
EXPERT GAME SKILLS AND EXECUTIVE CONTROL
38
action video games. Current Directions in Psychological Science, 18(6), 321–326.
Egeland, J., & Kovalik-Gran, I. (2010). Measuring several aspects of attention in one test:
The factor structure of Conners’s continuous performance test. Journal of Attention
Disorders, 13(4), 339–346.
Elo, A. (1986). The rating of chessplayers, past and present. New York: Arco Pub.
Gade, M., & Steinhauser, M. (2019). The impact of cue format and cue transparency on task
switching performance. Psychological Research. doi: 10.1007/s00426-019-01150-0
Grange, J., & Houghton, G. (2014). Task switching and cognitive control. Oxford, UK:
Oxford University Press. http://dx.doi.org/10.1093/acprof:
osobl/9780199921959.001.0001
Green, C. S., & Bavelier, D. (2003). Action video game modifies visual selective attention.
Nature, 423, 534–537. http://dx.doi.org/10.1038/nature01647
Green, C. S., & Bavelier, D. (2006). Enumeration versus multiple object tracking: The case of
action video game players. Cognition, 101, 217–245.
http://dx.doi.org/10.1016/j.cognition.2005.10.004
Green, C. S., & Bavelier, D. (2007). Action-video-game experience alters the spatial
resolution of vision. Psychological Science, 18, 88–94. http://dx.doi.org/10.1111/j.1467-
9280.2007.01853.x
Green, C. S., Bavelier, D., Kramer, A. F., Vinogradov, S., Ansorge, U., Ball, K. K., . . .Witt,
C. M. (2018). Improving Methodological Standards in Behavioral Interventions for
Cognitive Enhancement. Journal of Cognitive Enhancement, 3(1),2–29.
Green, C. S., Kattner, F., Eichenbaum, A., Bediou, B., Adams, D. M., Mayer, R. E., &
Bavelier, D. (2017). Playing some video games but not others is related to cognitive
abilities: A critique of Unsworth et al. Psychological Science, 28, 679–682, 2015.
http://dx.doi.org/10.1177/0956797616644837
Green, C., Strobach, T., & Schubert, T. (2013). On methodological standards in training and
transfer experiments. Psychological Research, 78(6), 756-772. doi: 10.1007/s00426-013-
0535-3
Green, C. S., Sugarman, M. A., Medford, K., Klobusicky, E., & Daphne Bavelier (2013). The
effect of action video game experience on task switching. Computers in Human Behavior,
28, 984–994. http://dx.doi.org/10.1016/j.chb.2011.12.020
Gobet, F., Johnston, S. J., Ferrufino, G., Johnston, M., Jones, M. B., Molyneux, A., . . .
Weeden, L. (2014). “No level up!”: No effects of video game specialization and
expertise on cognitive performance. Frontiers in Psychology, 5, 1337.
http://dx.doi.org/10.3389/fpsyg.2014.01337
Gong, D., Ma, W., Liu, T., Yan, Y., & Yao, D. (2019). Electronic-Sports Experience Related
to Functional Enhancement in Central Executive and Default Mode Areas. Neural
Plasticity, 2019, 1-7. doi: 10.1155/2019/1940123
Guo, D., Li, B., Yu, Y., Liu, X., & Li, X. (2019). Exploring the Limitations of the Shielding
Function of Categorization Rules in Task-Switching. Frontiers in Psychology, 10. doi:
10.3389/fpsyg.2019.01212
Hambrick, D. Z., Oswald, F. L., Darowski, E. S., Rench, T. A., & Brou, R. (2010). Predictors
of multitasking performance in a synthetic work paradigm. Applied Cognitive
Psychology, 24, 1149–1167. http://dx.doi.org/10.1002/acp.1624
Holm, S. (1979). A simple sequentially rejective multiple test procedure. Scandinavian
Journal of Statistics,6, 65–70
Ioannidis, K., Hook, R., Goudriaan, A., Vlies, S., Fineberg, N., Grant, J., & Chamberlain, S.
(2019). Cognitive deficits in problematic internet use: meta-analysis of 40 studies.
British Journal of Psychiatry, 215(5), 639-646. doi: 10.1192/bjp.2019.3
EXPERT GAME SKILLS AND EXECUTIVE CONTROL
39
Karle, J., Watter, S., & Shedden, J. (2010). Task switching in video game players: Benefits of
selective attention but not resistance to proactive interference. Acta Psychologica,134(1),
70-78. doi: 10.1016/j.actpsy.2009.12.007
Kramer, A., Cepeda, N., & Cepeda, M. (2001). Methylphenidate effects on task-switching
performance in attention-deficit/hyperactivity disorder. Journal of The American
Academy of Child & Adolescent Psychiatry,40(11), 1277-1284.
http://dx.doi.org/10.1097/00004583-200111000-00007
Kray, J., Karbach, J., Haenig, S., & Freitag, C. (2012). Can task-switching training enhance
executive control functioning in children with attention deficit/-hyperactivity disorder?
Frontiers in Human Neuroscience,5. http://dx.doi.org/10.3389/fnhum.2011.00180
Kalanthroff, E., & Henik, A. (2014). Preparation time modulates pro-active control and
enhances task conflict in task switching. Psychological Research,78(2), 276-288. doi:
10.1007/s00426-013-0495-7
Kiesel, A., Steinhauser, M., Wendt, M., Falkenstein, M., Jost, K., Philipp, A.M. & Koch, I.
(2010). Control and interference in task switching—A review. Psychological Bulletin,
136, 849-874. http://dx.doi.org/10.1037/a0019842
Kokkinakis, A., Cowling, P., Drachen, A., & Wade, A. (2017). Exploring the relationship
between video game expertise and fluid intelligence. PLOS ONE,12(11), e0186621.
http//doi: 10.1371/journal.pone.0186621
Kowal, M., Toth, A., Exton, C., & Campbell, M. (2018). Different cognitive abilities
displayed by action video gamers and non-gamers. Computers in Human Behavior, 88,
255-262. doi: 10.1016/j.chb.2018.07.010
Latham, A., Patston, L., & Tippett, L. (2013). Just how expert are “expert” video-game
players? Assessing the experience and expertise of video-game players across “action”
video-game genres. Frontiers in Psychology, 4. http://doi: 10.3389/fpsyg.2013.00941
League system. (2019, July 17). Retrieved from
https://leagueoflegends.fandom.com/wiki/League_system
Li, Q. (2002).
转型时期的中国社会分层结构
[The stratification of Chinese society in the
transitional period].Harbin, China: Heilongjiang People’s Press (p182).
Li, X., Li, B., Liu, X., Lages, M., & Stoet, G. (2019a). Task-switching costs disappear if non-
Chinese participants respond to Chinese characters. Journal of Experimental Psychology:
Learning, Memory, and Cognition. http://dx.doi.org/10.1037/xlm0000692
Li, B., Li, X., Liu, X., Lages, M., Stoet, G (2019b). Target-response associations can
eliminate task-switching costs but not response congruency effects. Frontiers in
Psychology, 10, 40.
Littel, M., Berg, I., Luijten, M., Rooij, A. J., Keemink, L., & Franken, I. H. A. (2012). Error
processing and response inhibition in excessive computer game players: An event related
potential study. Addiction Biology,17(5), 934–947.
Mack, D. J., & Ilg, U. J. (2014). The effects of video game play on the characteristics of
saccadic eye movements. Vision Research, 102, 26–32.
Mayr, U., LaRoux, C., Rolheiser, T., Osternig, L., Chou, L., & van Donkelaar, P. (2014).
Executive Dysfunction Assessed with a Task-Switching Task following Concussion.
Plos ONE, 9(3), e91379. doi: 10.1371/journal.pone.0091379
Meiran, N. (2000). Modeling cognitive control in task-switching. Psychological Research,
63(3-4), 234-249. http://dx.doi.org/10.1007/s004269900004
Metcalf, O., & Pammer, K. (2014). Impulsivity and related neuropsychological features in
regular and addictive first person shooter gaming. Cyberpsychology, Behavior and Social
Networking,17(3), 147–152.
EXPERT GAME SKILLS AND EXECUTIVE CONTROL
40
Monsell, S., & Mizon, G. (2006). Can the task-cuing paradigm measure an endogenous task-
set reconfiguration process? Journal of Experimental Psychology: Human Perception
and Performance,32(3), 493-516. http://dx.doi.org/10.1037/0096-1523.32.3.493
Powers, K. L., Brooks, P. J., Aldrich, N. J., Palladino, M. A., & Alfieri, L. (2013). Effects of
video-game play on information processing: A meta analytic investigation. Psychonomic
Bulletin & Review, 20, 1055–1079. http://dx.doi.org/10.3758/s13423-013-0418-z
Rank distribution (2019, Jule 17). Retrieved from
https://www.leagueofgraphs.com/rankings/rank-distribution
Redick, T. S., Unsworth, N., Kane, M. J., & Hambrick, D. Z. (2017). Don’t shoot the
messenger: Still no evidence that video-game experience is related to cognitive
abilities—A reply to Green et al. Psychological Science, 28, 683–686.
http://dx.doi.org/10.1177/0956797617698527
Röhlcke, S., Bäcklund, C., Sörman, D., & Jonsson, B. (2018). Time on task matters most in
video game expertise. PLOS ONE,13(10), e0206555. doi:
10.1371/journal.pone.0206555
Saban, W., Gabay, S., & Kalanthroff, E. (2018). More than just channeling: The role of
subcortical mechanisms in executive functions – Evidence from the Stroop task. Acta
Psychologica,189, 36-42. doi: 10.1016/j.actpsy.2017.03.001
Sala, G., Tatlidil, K., & Gobet, F. (2018). Video game training does not enhance cognitive
ability: A comprehensive meta-analytic investigation. Psychological Bulletin, 144(2),
111-139. doi: 10.1037/bul0000139
Schneider, D. W. (2015). Isolating a mediated route for response congruency effects in task
switching. Journal of Experimental Psychology: Learning, Memory, and Cognition, 41,
235–245.
Schneider, D. W. (2018). Categorization difficulty modulates the mediated route for response
selection in task switching. Psychonomic Bulletin & Review, 25, 1958–1967.
Stoet, G. (2010). PsyToolkit: A software package for programming psychological
experiments using Linux. Behavior Research Methods,42(4), 1096–1104.
Stoet, G. (2017). PsyToolkit. Teaching of Psychology,44(1), 24–31.
Stoet, G., & L. H, Snyder. (2007). Extensive practice does not eliminate human switch costs.
Cognitive,Affective, & Behavioral Neuroscience, 7 (3), 192–197
Sudevan, P., & Taylor, D. (1987). The cueing and priming of cognitive operations. Journal of
Experimental Psychology: Human Perception and Performance,13(1), 89–103.
Steinhauser, M., & Hubner, R. (2009). Distinguishing response conflict and task conflict in
the Stroop task: Evidence from ex-Gaussian distribution analysis. Journal of
Experimental Psychology: Human Perception and Performance, 35, 1398–1412
Strobach, T., Frensch, P. A., & Schubert, T. (2012). Video game practice optimizes executive
control skills in dual-task and task switching situations. Acta Psychologica, 140(1), 13–
24.
Strobach, T., Liepelt, R., Schubert, T., & Kiesel, A. (2011). Task switching: effects of
practice on switch and mixing costs. Psychological Research, 76(1), 74-83. doi:
10.1007/s00426-011-0323-x
Toril, P., Reales, J. M., & Ballesteros, S. (2014). Video game training enhances cognition of
older adults: A meta-analytic study. Psychology and Aging, 29, 706–716.
http://dx.doi.org/10.1037/a0037507
Unsworth, N., Redick, T. S., McMillan, B. D., Hambrick, D. Z., Kane, M. J., & Engle, R. W.
(2015). Is playing video games related to cognitive abilities? Psychological Science, 26,
759 –774. http://dx.doi.org/10.1177/0956797615570367
EXPERT GAME SKILLS AND EXECUTIVE CONTROL
41
Vandierendonck, A., Liefooghe, B., & Verbruggen, F. (2010). Task switching: Interplay of
reconfiguration and interference control. Psychological Bulletin,136(4), 601-626.
http://dx.doi.org/10.1037/a0019791
Wang, P., Liu, H. H., Zhu, X. T., Meng, T., Li, H. J., & Zuo, X. N. (2016). Action video
game training for healthy adults: A meta-analytic study. Frontiers in Psychology, 7, 907.
http://dx.doi.org/10.3389/fpsyg.2016.00907
Waris, O., Jaeggi, S., Seitz, A., Lehtonen, M., Soveri, A., & Lukasik, K. et al. (2019). Video
gaming and working memory: A large-scale cross-sectional correlative study. Computers
in Human Behavior, 97, 94-103. doi: 10.1016/j.chb.2019.03.005
Waszak, F., & Hommel, B. (2007). The costs and benefits of cross-task priming. Memory &
Cognition,35(5), 1175-1186. http://dx.doi.org/10.3758/bf03193487
Wendt, M., & Kiesel, A. (2008). The impact of stimulus-specific practice and task
instructions on response congruency effects between tasks. Psychological Research,
72(4), 425–432.
Wendt, M., Kähler, S., Luna-Rodriguez, A., & Jacobsen, T. (2017). Adoption of Task-
Specific Sets of Visual Attention. Frontiers in Psychology, 8. doi:
10.3389/fpsyg.2017.00687
Wu, S., Hitchman, G., Tan, J., Zhao, Y., Tang, D., Wang, L., & Chen, A. (2015). The neural
dynamic mechanisms of asymmetric switch costs in a combined Stroop-task-switching
paradigm. Scientific Reports, 5(1). doi: 10.1038/srep10240
Yeung, N., & Monsell, S. (2003). Switching between tasks of unequal familiarity: The role of
stimulus-attribute and response-set selection. Journal of Experimental Psychology:
Human Perception and Performance, 29, 455–469.
EXPERT GAME SKILLS AND EXECUTIVE CONTROL
42
Appendix A
Stroop-Switching: Mean (SD) of RT ms and ER % of Each Trial Condition and
Player Group
Congruent
Neutral
Incongruent
RT ms
ER %
RT ms
ER %
RT ms
ER %
Color-Naming Task
TP Rep
868(205)
0.55(1.92)
835(176)
1.01(3.90)
1073(212)
6.76(10.63)
AP Rep
850(174)
1.61(4.38)
815(132)
2.61(6.70)
1040(184)
14.90(11.70)
TP Swi
912(191)
1.25(2.33)
921(175)
2.83(4.39)
1127(209)
13.20(12.34)
AP Swi
930(163)
5.14(5.48)
979(179)
5.77(8.86)
1212(203)
22.75(13.87)
Word-Reading Task
TP Rep
928(238)
0.92(2.36)
814(196)
1.59(3.14)
1079(245)
11.54(9.18)
AP Rep
933(174)
2.72(4.40)
827(152)
2.74(4.88)
1041(180)
14.79(12.02)
TP Swi
901(224)
1.21(2.34)
884(206)
3.61(5.71)
1104(240)
13.75(12.53)
AP Swi
947(197)
4.00(5.93)
925(181)
5.07(6.79)
1185(232)
22.81(13.55)
Note. TP = Top player; AP = Average player; Rep = Repeat; Swi = Switch
EXPERT GAME SKILLS AND EXECUTIVE CONTROL
43
Appendix B
Stroop-Switching Test: Results of mixed-measures MANOVAs on the linear
combination of RT and ER data, using Trial transition (repeat, switch), Task (color-
naming, word-reading), Congruency (congruent, incongruent, neutral) as within-
subject factors, and Player group (top, average) as a between-subjects factor
Effect
F
df
p
Pillai’s Trace
η2p
Player group (G)
7.86
2, 67
<.001
.190
.190
Trial transition (TT)
49.09
2, 67
<.001
.594
.594
Congruency (C)
63.72
4, 272
<.001
.968
.484
Task (TK)
0.44
2, 67
.643
.013
.013
G×TT
7.62
2, 67
.001
.185
.185
G×C
3.94
4, 272
.004
.109
.055
G×TK
1.00
2, 67
.372
.029
.029
TT×C
12.42
4, 272
<.001
.309
.154
TT×TK
6.48
2, 67
.003
.162
.162
C×TK
4.92
4, 272
.001
.135
.067
G×TT×C
2.12
4, 272
.078
.061
.030
G×TT×TK
0.21
2, 67
.811
.006
.006
G×C×TK
0.35
4, 272
.844
.010
.005
TT×C×TK
1.30
4, 272
.271
.037
.019
G×TT×C×TK
0.99
4, 272
.413
.029
.014
Note: In the multivariate context, sometimes Pillai's Trace is equal to η2p
EXPERT GAME SKILLS AND EXECUTIVE CONTROL
44
Appendix C
CPT: Mean (SD) of RT ms, Response Bias, False Alarms % and Hit Rate % for Each
Player Group and the Correlations between the Four Dependent Variables
1
2
3
Top players
Mean (SD)
Average players
Mean (SD)
1. RT (ms)
-
337 (30)
334 (30)
2. Response Bias
.17
-
.57 (.12)
.54 (.09)
3. False Alarm Rates (%)
- .34**
-.61***
38.62 (17.32)
48.72 (11.29)
4. Hit Rates (%)
.16
-.35***
-.40***
91.27 (4.28)
87.54 (5.76)
Note. ** p<.01; *** p<.001
EXPERT GAME SKILLS AND EXECUTIVE CONTROL
45
Appendix D
Schematic illustrations of three types of target stimuli and the associated response-
congruency effect and reverse-facilitation effect in Stroop-switching test
Figure D1. (a) Reverse-facilitation effect = congruent trial - neutral trial; Response-
congruency effect = incongruent trial - neutral trial. (b) Example of the congruent target
stimuli, in which the ink color and the word meaning are mapped to the same response in the
color-naming and word-reading tasks. (c) Examples of the neutral target stimuli, in which
only the color (upper) or word printed in black (lower) are displayed and only appear in one
task. (d) Examples of the incongruent target stimuli, the ink color and the meaning of the
word are mapped to different responses in the color-naming and word-reading tasks.
Please note that the neutral target stimulus REDblack also has a color because it is
printed in black. However, in the present study, the color-naming task rule only required
participants to judge between two colors: green and red, which means black is not included.
As a result, the target REDblack did not appear in the color-naming task and should be
EXPERT GAME SKILLS AND EXECUTIVE CONTROL
46
considered as a neutral target stimulus appearing only in the word-reading task.
Article
Full-text available
Working memory (WM), inhibitory control (IC) and cognitive flexibility (FC) are core executive functions (EFs) that influence individuals' cognition and behavior. The stimulation of these EFs can contribute to an individual's more adaptive development and can be performed through digital games. In this context, League of Legends (LoL) is distinguished by its popularity. The aim of this article was to evaluate which nuclear EFs (MT, CI and FC) are stimulated by the LoL game. The research had a qualitative approach and to reach the objective, two researchers interacted with LoL for an average of seven hours. The researchers find that the three nuclear EFs are stimulated. It is possible that LoL activities that require nuclear EFs are broader and more diversified than those reported. However, due to the research method, the identification of stimulated EFs is limited by the perception of researchers.
Article
Full-text available
Applying categorization rules narrows attention toward the relevant features of a target and helps participants to ignore the irrelevant features of the target. This is called the shielding function of categorization rules. Here we explored the limitation of the shielding function in two task-switching experiments. In Experiment 1, we assigned each target a single digital numeral as an additional feature in addition to conventional bivalent features as in the previous task-switching experiments with bivalent tasks. In the first two stages of Experiment 1, half of the participants learned the numeral-response associations and the other half used an alternative numeral-categorization rule to infer the response. Without participants applying conventional task-switching rules, the switching costs disappeared. Moreover, when participants performed tasks by numeral-response associations the bivalent features interfered with response retrieval and caused response-congruency effects, whereas when participants applied the numeral-categorization rule, the bivalent features were shielded away and thereby the response-congruency effects disappeared. In the third stage, in which all participants applied task-switching rules by discriminating between bivalent features (i.e., filling and orientations), we found task-switching costs and response-congruency effects. In Experiment 2, new bivalent features produced stronger interference compared to Experiment 1. As a consequence, participants in both the association group and the numeral-categorization rule group showed significant response-congruency effects in the first two stages, where task-switching rules were not introduced. It follows that the shielding function of categorization rules has limits—strong interference from bivalent features can break down the shielding function. In addition, participants in the association group showed task-switching costs without being informed about the task-switching rules. We propose that strong proactive interference can produce task-switching costs even without the use of task-switching rules.
Article
Full-text available
Background Excessive use of the internet is increasingly recognised as a global public health concern. Individual studies have reported cognitive impairment in problematic internet use (PIU), but have suffered from various methodological limitations. Confirmation of cognitive deficits in PIU would support the neurobiological plausibility of this disorder. Aims To conduct a rigorous meta-analysis of cognitive performance in PIU from case–control studies; and to assess the impact of study quality, the main type of online behaviour (for example gaming) and other parameters on the findings. Method A systematic literature review was conducted of peer-reviewed case–controlled studies comparing cognition in people with PIU (broadly defined) with that of healthy controls. Findings were extracted and subjected to a meta-analysis where at least four publications existed for a given cognitive domain of interest. Results The meta-analysis comprised 2922 participants across 40 studies. Compared with controls, PIU was associated with significant impairment in inhibitory control (Stroop task Hedge's g = 0.53 (s.e. = 0.19–0.87), stop-signal task g = 0.42 (s.e. = 0.17–0.66), go/no-go task g = 0.51 (s.e. = 0.26–0.75)), decision-making (g = 0.49 (s.e. = 0.28–0.70)) and working memory (g = 0.40 (s.e. = 0.20–0.82)). Whether or not gaming was the predominant type of online behaviour did not significantly moderate the observed cognitive effects; nor did age, gender, geographical area of reporting or the presence of comorbidities. Conclusions PIU is associated with decrements across a range of neuropsychological domains, irrespective of geographical location, supporting its cross-cultural and biological validity. These findings also suggest a common neurobiological vulnerability across PIU behaviours, including gaming, rather than a dissimilar neurocognitive profile for internet gaming disorder.
Article
Full-text available
In task-switching experiments with bivalent target stimuli, conflicts during response selection give rise to response-congruency effects. Typically, participants respond more slowly and make more errors in trials with incongruent targets that require different responses in the two tasks, compared to trials with congruent targets that are associated with the same response in both tasks. Here we investigate whether participants show response-congruency effects when task rules are not made explicit. In two experiments, we assigned task-irrelevant features to each bivalent target. When participants were instructed to apply the task rules, they showed significant task-switching costs as well as response-congruency effects. Importantly, when the same participants did not know the task rules and responded without applying the task rules, they showed response-congruency effects but no switch costs. The significant congruency effects suggest that associations between bivalent target features and responses can be formed passively, even when participants do not follow the task rules and use task-irrelevant target features to make a response.
Article
Full-text available
In experiments with univalent target stimuli, task-switching costs can be eliminated if participants are unaware of the task rules and apply cue-target-response associations. However, in experiments with bivalent target stimuli, participants show task-switching costs. Participants may exhibit switch costs even when no task rules are provided in the instructions because they can infer the task rules. We tested this prediction by controlling the meaningfulness of cues and targets and therefore the ability to apply the task rules in 2 groups of participants. We compared the performance of Chinese and non-Chinese participants, who responded to Chinese numerals in an odd/even and high/low number task. In Experiment 1, Chinese participants, who knew Chinese characters and understood the task rules, showed task-switching costs. Non-Chinese participants on the other hand, who did not know Chinese characters, exhibited no switch costs. They applied a "target-first" strategy which means that they processed the target stimulus before the cue. In Experiment 2, we confirmed the absence of task-switching costs in Chinese participants using traditional Chinese numerals as target stimuli. Further, to determine how the target-first strategy affects switch costs, we manipulated the sequence of cue and target presentations. We conclude that taskswitching costs can be eliminated more easily than previously thought, even for bivalent stimuli. The occurrence of task-switching costs depends on the approach used by participants and this may answer the puzzling question why humans typically do show task-switching costs whereas pigeons and monkeys do not.
Article
Full-text available
Cues help in retrieving and implementing task-sets, that are actual representations of the to-be performed task in working memory. However, whereas previous studies revealed that the effectiveness of selecting and implementing task-sets based on cues depends on the type of cue (i.e., transparent words vs. arbitrary shapes), it is still unclear which characteristics of cues are responsible for these differences and whether the impact of the cue is bound to task-set retrieval only or also impacts task-set representations. For instance, the amount of interference during actual task performance has been reported to alter dependent on cue type as do preparation gains such as the reduction of switch cost. To investigate the effectiveness of cue characteristics (i.e., cue transparency and cue format), we manipulated those within- and betweenparticipants in three experiments. Main dependent measures were switch costs in reaction times and error rates that occur when participants have to switch task-sets, and thus update working memory content. Our results consistently show beneficial effects of transparent cues for the reduction of switch cost. The influence of cue format was manifest in within-participants manipulation only and was mainly found in error rates. Overall, our data suggest that the amount of interference experienced in actual taskperformance can be significantly modulated dependent on cue type, suggesting flexible adaptation of the cognitive system to contextual information.
Article
Full-text available
Electronic-sports (e-sports) is a form of organized, online, multiplayer video game competition, which requires both action skills and the ability and process of forming and adapting a strategy (referred to as strategization hereafter) to achieve goals. Over the past few decades, research has shown that video gaming experience has an important impact on the plasticity of the sensorimotor, attentional, and executive brain areas. However, little research has examined the relationship between e-sports experience and the plasticity of brain networks that are related to strategization. Using resting-state fMRI data and the local functional connectivity density (lFCD) analysis, this study investigated the relationship between e-sports experience (League of Legends [LOL] in this study) and brain plasticity by comparing between top-ranking LOL players and lower-ranking (yet experienced) LOL players. Results showed that the top-ranking LOL players had superior local functional integration in the executive areas compared to lower-ranking players. Furthermore, the top-ranking players had higher lFCD in the default mode areas, which have been found related to various sub-processes (e.g., memory and planning) essential for strategization. Finally, the top-ranking players' lFCD was related to their LOL expertise rank level, as indicated by a comprehensive score assigned by the gaming software based on players' gaming experience and expertise. Thus, the result showed that the local functional connectivity in central executive and default mode brain areas was enhanced in the top-ranking e-sports players, suggesting that e-sports experience is related to the plasticity of the central executive and default mode areas.
Article
Full-text available
There is substantial interest in the possibility that cognitive skills can be improved by dedicated behavioral training. Yet despite the large amount of work being conducted in this domain, there is not an explicit and widely agreed upon consensus around the best methodological practices. This document seeks to fill this gap. We start from the perspective that there are many types of studies that are important in this domain—e.g., feasibility, mechanistic, efficacy, and effectiveness. These studies have fundamentally different goals, and, as such, the best-practice methods to meet those goals will also differ. We thus make suggestions in topics ranging from the design and implementation of control groups, to reporting of results, to dissemination and communication, taking the perspective that the best practices are not necessarily uniform across all study types. We also explicitly recognize and discuss the fact that there are methodological issues around which we currently lack the theoretical and/or empirical foundation to determine best practices (e.g., as pertains to assessing participant expectations). For these, we suggest important routes forward, including greater interdisciplinary collaboration with individuals from domains that face related concerns. Our hope is that these recommendations will greatly increase the rate at which science in this domain advances.
Article
Full-text available
In this study, we investigated whether working memory capacity (WMC), personality characteristics (grit) and number of matches played (time on task) can predict performance score (matchmaking rating [MMR]) in experienced players of a popular video game called Dota 2. A questionnaire and four online-based cognitive tasks were used to gather the data, and structural equation modelling (SEM) was used to investigate the interrelationships between constructs. The results showed that time on task was the strongest predictor of MMR, and grit also significantly influenced performance. However, WMC did not play a substantial role in predicting performance while playing Dota 2. These results are discussed in relation to sample characteristics and the role of deliberate practice and skill acquisition within the domain of playing Dota 2. Further, we suggest that future research investigates the social aspects of attaining skill, the relationship between personality and performance, and the qualitative aspects of time spent on a task.
Article
Studies have indicated that video gaming is positively associated with cognitive performance in select cognitive domains, but the magnitudes of these associations have been called into question, as they have frequently been based on extreme groups analyses that have compared video gamers with non-gamers. When including the whole range of participants, and not just extreme cases, these effects were observed to reduce markedly (Unsworth et al., 2015). To further study this issue, we compared the associations between video gaming and aspects of working memory (WM) performance in an extreme groups design to those of a design that includes the full range of participants in a large adult sample (n = 503). WM was measured with three composite scores (verbal WM, visuospatial WM, n-back). The extreme groups analyses showed that video gamers performed better than non-gamers on all three WM measures, while the whole sample analyses indicated weak positive associations between the time spent playing video games and visuospatial WM and n-back performance. Thus, study design modulated the effects, but two of the three associations between WM and video gaming were consistent across both analysis techniques. A separate study confirmed that our questionnaire-based estimate of gaming hours was reliable when compared with one-week diaries of videogame playing. While the present cross-sectional results preclude causal inferences, possible mechanisms of WM – videogame playing associations and future research directions are discussed. Overall, our results indicate that cognition – videogame playing relationships, albeit weak, are not solely due to recently discussed methodological artefacts concerning the particular analytical approach and survey reliability.
Article
Developing impulsivity has been one of the main concerns thought to arise from the increasing popularity of video gaming. Most of the relevant literature has treated gamers as pure-genre players (i.e. those who play only a specific genre of game). However, it is not clear how impulsivity is associated with different genres of games in multi-genre gamers, given that there is increasing diversity in the games played by individuals. In this study, we compared 33 gamers to 23 non-gamers in a go/no-go task: the Continuous Performance Test (CPT). To evaluate whether or not impulsivity occurs as a trade-off between speed and accuracy, we emphasised fast performance to all participants. Then, to examine the ability to predict impulsivity from game genre-hours, we fitted separate multiple regression models to several dependent variables. As an additional measure, we also compared groups in an antisaccade task. In the CPT, gamers showed a trend towards significantly faster reaction time (RT), accompanied by higher false alarm rate (FAR) and more risk-taking response bias (β), suggesting impulsive responses. Interestingly, there was a significant negative correlation between RT and FAR across all participants, suggesting an overall speed-accuracy trade-off strategy, perhaps driven by the emphasis on speed during task instruction. Moreover, time spent on role playing games (RPG) and real-time strategy (RTS) games better predicted FAR and β than did time spent on action and puzzle games. In the antisaccade task; however, gamers showed a shorter antisaccade latency but a comparable error rate in comparison with non-gamers. There was no specific game genre which could predict performance in the antisaccade task. Altogether, there was no evidence of oculomotor impulsivity in gamers; however, the CPT results suggested the presence of impulsive responses in gamers, which might be the result of a speed-accuracy trade-off. Furthermore, there was a difference in game genres, with time spent on RPG and RTS games being accompanied by greater probability of impulsive responses. Training studies are required to investigate the causality of different video game genres on the development of impulsivity.