The Relationship of Players to Their Avatars in MMORPGs: Differences between Adolescents, Emerging Adults and Adults

Lukas Blinka
Institute for Research of Children, Youth and Family, Faculty of Social Studies,

Masaryk University, Brno, Czech Republic


Typically, MMORPG (Massively Multiplayer Online Role-Playing Games) are persistent virtual worlds falling within the fantasy genre, in which the player controls his or her character which subsequently becomes a part of the online fantasy world. The study focuses on the aspects of player's relationship with the character while searching for differences in these aspects between various age groups, especially adolescents (12 – 19 years) and young adults (20 – 27 years). A sample containing 532 players was obtained upon contacting players in discussion fora, where they were asked to fill in a web-based questionnaire in English. A battery of 12 questions was tested dealing with the various aspects of relationship between the players and their avatars. The results identified three dominant factors. The first one dealt with identification, i.e. not distinguishing the player from the avatar and the younger the player the stronger the identification. The immersion factor, i.e. daydreaming and emotional feelings towards avatar, was found to be important on the same level for all age cathegories. Among adolescents and emerging adults, the results show a higher similarity in the identification and compensation between the player and the avatar, while the relationship (and its components) was found to be weaker among adults.

Key words: MMORPG, avatar, adolescence, emerging adulthood


MMORPG is a new phenomenon with growing importance, which is implied, for instance, by the data on the most popular of these games, World of Warcraft, which has gained over 10 million players in the last two years following its launch (Blizzard entertainment, 2008). According to the data provided by the “Entertainment Software Association” (2007), between 2006 and 2007, the number of players doubled, while an even stronger growth is expected in the forthcoming year. Yee (2006) claims the players' average age to be 25 years, while the average intensity of playing is reported to be 23 hours per week. Griffiths et al. (2004) found out that the average time of playing the game stretches to approximately two years among adolescents (under 20 years of age) and 27 months among older players. Younger players then play more intensively that their adult counterparts (26 hours per week among under-20-year-olds, compared to 22 hours among those aged over 26 years). However, the highest intensity of playing MMORPGs was found in the group of those aged between 20 and 22 years, reaching almost 30 hours per week. This implies that it is a very important leisure activity of some contemporary adolescents and emerging adults. When studying the psychology of playing MMORPGs, authors tend to focus their attention on extensive playing or possibly addiction, while putting aside other aspects these games offer for research and study.

Described as post-modern at times, the present offers possibilities to rework one's own identity or create other identities (Turkle, 1995). Working on a specific identity is then one of the specific features of MMORPGs, when the player sets up and communicates with others in the game through the avatar only, functioning as a the representation of the player's identity, which may resemble or differ from the player. So far, the studies have paid limited attention to the interconnections between the player and their game representation, compared to other topics. For instance, Yee (2006), does not include the motive of the “player's character” itself in the categorisation of player's motivations, whereas this may be deduced from the combinations of other ones, in particular from the advancement (advancing through various levels with the avatar) and mechanics (optimising the performance through mastering the game mechanisms). Both motivations may be summarised as motivations to succeed. According to Yee (2006), the motivation of immersion with role-playing components functions as another influence (i.e. creating and playing on behalf of the avatar via a role defined in advance), as well as customization components (working on the avatar's appearance). The motivation to succeed is significantly stronger among younger male players (i.e. adolescents) and among those who play more intensively. According to Griffiths et al. (2004), the intensity of playing (measured in the number of hours per week) is also characteristic of younger, adolescent players.

Some authors (e.g. Allison et al., 2006 and Turkle, 1995) compare the player and avatar's relationship to a transmission field, as roughly defined by psychoanalysis: the avatar is not in exclusive competence of the player, while not being completely separated from them. It lies somewhere near the boarder of external and internal (psychological) reality. From the player's viewpoint, the avatar is a kind of individual overlap owing to which they may experiment with their identities. Wolfendale (2006) then describes directly the player and avatar's relationship as an attachment, i.e. as if with an absent or idealised or unreal person, while the feelings created around this relationship are real.

Some psychodynamic schools point out the focal role of so-called images in one’s psyche (e.g. Hillman, 1997, Kast, 1992), which are certain complexes around which fantasies and emotions resolve and which can be experienced as certain independent parts of one’s psyche. The in-game character can be, in a certain sense, considered an “image” and can thus be an important aspect of a player’s life or psychical development. Understanding what the relationship between a player and his in-game character mean for the player can, to a certain extent, lead to an explanatory framework for e.g. excessive game-play.

This paper focuses on the relationship between the player and their avatar, being the player's representation in the virtual space. The relationship has been analysed from the viewpoint of the participants' age structure (adolescents, emerging adults, and adults), also comparing whether they have a partner, are married or single. The core of this article is thus the development viewpoint of the phenomenon, however we also wanted to point out a possible psychodynamic interpretation.

Procedure and Participants

The study surveyed 532 respondents – MMORPG players. The data collection took place with the help of an online questionnaire. Players were asked to fill it in using an advertisement on several biggest discussion for a about the World of Warcraft and Everquest games, which are two most popular MMORPGs. Over a couple of weeks, the advert was regularly updated in order to make it more visible from discussion fora participants.

On average, participants were 25 years old. The sample was subsequently divided into three age groups: adolescent players (under 19 years old), so-called emerging adults (20 - 26 years old) and adults (27 years and more). The concept of the “emerging adulthood” (Arnett 2000, 2004) was considered suitable in order to grasp and define the differences between adolescence and adulthood, i.e. the third decade of human life, which, as it has been demonstrated, is a specific period of life belonging neither to adolescence, nor adulthood. Table 1 shows the sample's distribution according to age categories and sex.

Table 1: Age categories and gender

Age categoryAll participantsMaleFemaleTotal
Adolescents (12-19 years)26.1% 96.9% 3.1% 100%
Emerging adults (20-26 years)36.3% 80.1% 19.9% 100%
Adults (27 and more)36.8% 79.7% 20.3% 100%
Total100%Mean 84.6%Mean 17.6%100%

Due to a significantly lower number of female players in the sample (especially among the youngest containing mere four girls), it was impossible to perform more detailed statistic operations separating the respondent's sex. However, basic statistic differences are provided for the group of 20-or-more-year-olds. Table 2 shows the sample's distribution according to age categories and the fact whether they live alone, in an intimate partnership or are married.

Table 2: Age categories and relationship

Age categorySingleIntimate MarriedTotal
Adolescents (12-19 years)83%17%0%100%
Emerging adults (20-26 years)50.2%40.9%8.3%100%
Adults (27 and more)33.7%28.6%37.8%100%
Total Mean52.9%29.6%17.2%100%

33% participants came from North America, 64.2% from Europe and 2.6% from other parts of the world.

The questionnaire dealt with various aspects of playing; nevertheless, this paper presents the results obtained, in particular, from the battery of twelve questions concerning the relationship between the player and their avatar. Results stemming from a qualitative survey of playing role-play games from the base of this battery. The conceptualisation of the battery is rooted in the viewpoint of depth psychology (Blinka & Šmahel, 2007). The questions tap into various aspects of the relationship between a player and his character, We were inspired especially by the psychodynamic school.

The focal points include: the positive and negative emotional relationship; perceptible blending or separation of the player/avatar, ; the avatar’s superiority (considering the avatar “something more”) or inferiority (regarding the avatar the player’s subordinate); whether the avatar is a subject of daydreaming in complex form (including the game itself, respectively in-game situations) or whether such fantasies are object-oriented and only develop around the avatar itself. Subsequently, these results are compared with individual age categories. The respondents were asked to answer the questions using a five-point scale from 1 – “strongly disagree” to 5 – “strongly agree”.


In table 3, we can see frequencies of agreement with certain items. Certain differences between age categories are easily noticeable. Whilst adolescents and emerging adults are very close in the tendency of not distinguishing themselves from their avatar (respectively they believe they have the same properties) and being like their avatar, adults agree with these items significantly less often. On the other hand, there is no significant difference between individual age groups in fantasies and daydreaming about the game itself and the avatar.

Table 3: Agreements with statements

 AdolescentsEmerging adults Adults
I possess the same skills and abilities as my character does25.0%20.7%10.7%
Sometimes I think just about situations from the game while not gaming.64.6%63.7%57.8%
Sometimes I think just about my character while not gaming.29.6%32.6%30.7%
I would rather be like my character37.4%37.4%19.5%

The thirteen-question battery was subject to a factor analysis, which allows deeper insight into the examined battery of questions. The results identified three dominant factors, explaining in total 54% of variable dispersion.

Table 4: Identified factors and items factor loading

 identification immersion compensation
Initial Eigenvalues Total3.71.71.1
% of Variance 30.614.19.6
Cumulative %30.644.754.3
I possess the same skills and abilities as my character does.0.830.13-0.08
My character skills and abilities are like mine, but somewhat greater.0.820.070.15
My character compensates my own skills and abilities.0.670.110.35
Both me and my character are the same.0.740.18-0.05
Sometimes I think just about my character while not gaming.0.050.830.09
Sometimes I think just about situations from the game while not gaming.-0.030.75-0.04
I think about my character itself rather than about the game situations.0.210.670.16
Sometimes I feel ashamed for my character0.180.420.23
Sometimes I feel proud of my character0.210.460.20
My character posses totally opposed skills and abilities to my own-0.16-0.040.67
I would rather be like my character0.240.260.71
Being like my character would help me in some situations of my real life0.190.330.61

Extraction method: Principal Component Analysis
Rotation method: Varimax with Kaiser Normalization

Factor 1 was called “identification”. The following score reach the reliability level (Cronbach alpha) of 0.80. Comparing the means of this factor for individual age categories (Bonferonni post hoc test), as summarised in Table 5, implies significant differences between adult players and both younger groups of players. No significant differences were then found between adolescents and emerging adults. Specifically, adult players show lower scores in the identification factor, while on the other hand, adolescents and emerging adults have higher scores.

Table 5: Comparison of identification factor means and age categories

(I) Age categoryMeanstd. Deviation (J) Age categoryMean Difference (I-J)Sig.
Adolescents9.883.96Emerging adults.82ns
Emerging adults9.053.71Adolescents-.82ns
     Emerging adults-1.31.001

When comparing the means among the types of partnership (the results apply to respondents over 20 years of age only), shown in Table 6, it was revealed that married players show a lower identification with their avatars than players not living in a partnership or living in a partnership, yet unmarried (furthermore, the difference is significant only between married players and other two groups).

From the viewpoint of the player's (not avatar's) sex, this factor did not produce any significant results between men and women.

Table 6: Comparison of identification factor means and relationship (20 years and older)

(I) RelationshipMeanstd. Deviation (J) RelationshipMean Difference (I-J)Sig.

The item frequencies of this factor indicate that the union of a player and his avatar is related to the player’s age:

Factor 2 was called the “immersion”. When creating a score from this factor's items, the reliability level of 0.68 (Cronbach alpha) was achieved. This factor reflects a deeper interconnection between the player and the avatar in a way that the player often thinks of the avatar, feels emotional towards it. This means the player considers important what the avatar does. Comparing the means of the scores of this factor implies that the age is not a sufficient indicator: according to the Bonferonni post hoc test, no significant differences were found among any age category. It may thus be said that there are no differences in the immersion, regardless of age.

On the other hand, partnership proves to be important for this factor. Table 7 shows significant differences occurring between the married and singles (again, only those aged 20 or more years were tested). Surprisingly enough, no differences were found between singles and those having a partner. In general, it may be said that married and adult individuals tend to reach a lower score, while the highest score can be found among lone adolescents and emerging adults.

Table 7: Comparison of immersion factor means and relationship (20 years and older)

(I) RelationshipMeanstd. Deviation (J) RelationshipMean Difference (I-J)Sig.

From the viewpoint of players' sex, men score higher than women as far as the “immersion” is concerned (t(353) = 2.29, p = 0.022), as well as among those aged 20 or more (t(258) = 2.77, p = 0.006), who were tested due to a low number of girls in the youngest group of adolescents.

Factor 3 might be called “compensation” in the sense of complementing, superiorising and idealising the player's ego. The reliability of the factor's score reached only 0.66 (Cronbach alpha). In view of the fact that the factor comprised few items and that there is a relatively higher cross-loading in the items of this factor, we decided not to analyze it. The existence and independence of this factor was only suggested, its confirmation is however subject to further research.

The respondents were also asked to mention whether their avatar is the same sex as them. 96% women play a character of the female sex, while only 77% men play a character of the male sex. It may thus be said that that most feminine avatars in the game is controlled by men (due to a very low women's participation in the game). Furthermore, differences in the identification and immersion factors were also tested from the viewpoint of whether the player acts with an avatar of the same or different sex. However, no significant differences in scores were found. In other words, the players identify themselves on a similar level and there is a similar level of immersion taking place, regardless of the fact whether the avatar is the same or opposite sex as the players themselves. Similarly, no differences in age were found either – the tendency to playing a character of opposite gender is thus the same for male players of all age categories.


The “identification” factor expresses a simple unity between the players and their avatars. This unity, or lack of distinguishing, does not take any emotional form. At the same time, it may mean that the player does not perceive the avatar as something “more” but simply as a game mechanism, i.e. something the purpose of which is to reach something else, such as entertainment, social contact, self-esteem, and self-efficacy through success etc. Probably, this is a component which Yee (2006) describes as achievement motivations. The tendency to reach the highest possible level in the game and to achieve mastery is included in this component. The existing research also implies that it is typical of adolescent players (Yee, 2006; Griffiths et al. 2004). The data obtained in our survey also show that it is younger players who tend to identify with (i.e. not to distinguish from) their avatars, while the younger the respondents were, the stronger the phenomenon.

On the other hand, the “immersion” factor shows a deeper interconnection of the player and the avatar. This factor connects imagination, daydreaming of the game and avatar with the emotional relationship towards it, which has support in the psychodynamic approach considering imaginations and emotions as “two sides of the same coin” (Hillman, 1997, Kast, 1992). This relationship is however not directly confirmed, since high cross-loading of items inquiring into the emotional relationship to the avatar hinders an unambiguous interpretation of the connection of the emotional relationship to the avatar with dreaming about it.

In this factor, the player distinguishes the avatar, assigning to it an independent status, to some extent. The avatar (and the game) becomes a subject of daydreaming. What is an important act is the fact that the player feels committed to what the avatar does in the game. This factor could be, in part, assigned to the immersion motivation, as described by Yee (2006), i.e. the players play the game focussing on the avatar itself, modifying its appearance and trying to fulfil its role characteristics. In case of this factor, what is likely to be valid is what Allison et al. (2006) claims that the avatar is neither the player, nor an independent entity: it lies somewhere in between. This factor did not show any statistically significant differences in age categories. Nevertheless, a significantly lower score was found among married players. It could be claimed that the avatar, using this perspective, has the status of an almost independent entity, which lacks the sense and place in the psyche provided the player shares their intimacy in a real-life partnership. Wolfendale (2006) implies the existence of a stereotype that the attachment to the avatar is something negative and that it concerns socially isolated individuals (or that it leads to social isolation). In a certain way, it may correspond just to this phenomenon: if the avatar gains some independence, it also means that the player may use it to project his or her psychological contents (or dissociate them) without accepting a commitment on their behalf, which leaves room for experimenting. Nevertheless, the results concerning the connection of playing on behalf of an opposite-sex character do not show that this style of playing may automatically imply experimenting with the player's identity: the avatar of the opposite sex tends to be used as a mere game tool without any influential reflection. But we support previous findings e.g. by Griffiths et al. (2003), that majority of “gender swappers” are male.

The third factor (compensation) is characteristic, and in its most typical position, perceived as and ideal which the players wish to approach. To some extent, this factor is also connected with the identification factor, since the avatar is perceived as an instrument. Subsequently, there is a difference between factor 1 and factor 3 in the extent to which the player reflects their avatars: in the first case, the reflection is low, while in the latter case it is high. This connection can be implied by the “my character compensates my own skills and abilities” item, which should logically belong to the “compensation” factor. However, in the factor analysis it falls under the “identification” factor, although with a high cross-loading towards “compensation”. Unfortunately, a generally high cross-loading disallows an adequate analysis of this factor, and so we only suggest its existence. The distribution of agreement with one of the items of this factor implies that the wish to be like one’s avatar is more often present in younger players (actually same for adolescents and emerging adults), which resembles items in the “identification” factor.

From the developmental point of view, it is interesting to observe the distribution of these factors between the three age categories. Adolescents demonstrated the highest tendency to identification of themselves and their avatars, i.e. they had the highest need to perform well in the game. Some surveys (e.g. Wan & Chiu, 2006, Bessiere et al., 2007) imply a connection between extensive playing of MMORPGs and low self-efficacy and self-esteem. The need for these components of the self is, however, characteristic among adolescents, as well (Macek, 1999; Alsaker & Kroger, 2008), and there is thus a possibility that MMORPGs provide adolescents with space to complete their developmental tasks, typical of this age. Adolescents are focussed on developing the feeling of their own competence, which is strongly linked to the feeling that the individual was an actor and that success resulted from their own skill and capabilities: the perception of one's own performance is thus more important when developing a positive self-assessment than others' opinions. In a way, MMORPGs produce a challenge to their players, both in the form of mastering the technology, and in the form of pitting one's wits with their peers. To some extent, a higher level of aggression among younger players (Griffiths et al., 2004) may be explainable from the viewpoint of this phenomenon, i.e. that they try to achieve quickly a feeling of self-efficacy through connecting a successful avatar and their own self. This tendency, yet weaker, is also visible among emerging adults, while it was significantly lower among adult players. Similarly, the importance of the compensation factor decreased with age. As for experimenting with the avatar (in the “immersion” factor), no differences were found between any age categories; nevertheless, it was significantly lower among married individuals. This might be explained using Arnnett's concept of emerging adulthood. Whereas its bottom age limit (around 19 years) is clearly determined as the period of leaving secondary school, the upper limit is, to a large extent, diffusional and unclear. There is only a subjective determination which may be perceived as clearer, depending on the level at which the individual reflects accepting duties, obligations, or commitments. It is entering into marriage that functions as an important acceptance of commitments, i.e. the moment when the individual accepts the commitment to another person. This moment usually completes the period of emerging adulthood. Our study shows that the age is, at least to some extent, an important factor affecting the style of playing MMORPGs, since there is a significantly lower level of identification among adults than among younger players. Similarly, there is a significantly lower level of immersion among married players, i.e. those who may no longer be classified as emerging adults. It is then believed that among adults, the motivation to play MMORPGs is represented by something else than their avatars (it can be, for instance, the social dimension of the game, etc.). On the other hand, for younger players, the virtual character may be an important motivation factor, especially from the viewpoint of performance and success in the game, as in case of adolescents. Nevertheless, the world of MMORPGs is, in a way, most convenient for the category of emerging adults, since they typically experiment with their identities (Arnett, 2000). Since Sherry Turkle (1995), This has been one of the most important features of online games. Griffiths et al. (2004) also show that players around 22 years of age tend to play most, which implies a connection between the concept of emerging adulthood and playing online games. Still it seems that emerging adults are, in their style of game-play and relationship to their in-game avatars, very close to adolescents rather than defining a specific category.


Author acknowledge the support of the Czech Ministry of Education, Youth and Sports (MSM0021622406) and the Faculty of Social Studies, Masaryk University in Brno.


Allison, S.E., Walde, L.von, Shockley, T., O´Gabard, G. (2006). The developement of self in the era of the internet and role-playing games. The American Journal of Psychiatry163, 381-385.

Alsaker, F.D., Kroger, J. (2008). Self concept, self-esteem and identity. In: Jackson, S, Goossens L. (Eds). Handbook of adolescent development (pp. 90-117). New York: Psychology press.

Arnett, J.J. (2000). Emerging adulthood: A Theory of Development from the Late Teens through the Twenties. American Psychologist, 55, 469 – 480.

Arnett, J.J. (2004). Emerging Adulthood: The Winding Road from Late Teens through the Twenties. Oxford: Oxford University Press.

Bessiere, K., Seay, F., Kiesler, S. (2007). The Ideal Elf: Identity Exploration in World of Warcraft. Cyberpsychology and behavior, 10, 530-535.

Blinka, L., Smahel, D. (2007). Role-playing" hry v kontextu analytické psychologie [Role-playing games in the context of analytical psychology]. Československá psychologie, 51, 169-182.

Blizzard entertainment (2008). World of Warcraft reaches new milestone: 10 million subscribers. Retrieved March 15, 2008, from

Entertainment software association (2007). Essential facts about the computer and video games industry: 2007 sales, demographics and usage data. Retrieved April 10, 2008, from

Griffits, M.D., Davies, M.N.O, Chappell, D. (2003). Breaking the stereotype: The case of Online Gaming. Cyberpsychology and behaviour 6, 81-91.

Griffiths, M.D., Davies, M.N.O., Chappell, D. (2004). Online computer gaming: a comparison of adolescent and adult gamers. Journal of adolescence 27, 87-96.

Hillman, J. (1997). Re-visioning psychology. New York: HarperPerennial.

Kast, V. (1992). The dynamics of symbols: Fundamentals of Jungian psychotherapy. New York: Fromm International.

Macek, P. (1999) Adolescence. Praha: Portál.

Turkle, S. (1995). Life on the screen: identity in the age of the Internet. New York: Simon & Schuster.

Wan, C., Chiou, W. (2006). Why are adolescents addicted to online gaming? An interview study in Taiwan. CyberPsychology & Behavior, 9, 762-766.

Wolfendale, J. (2006) My Avatar, my self: virtual harm and attachment. In: Polčák, R., Škop, M., Šmahel, D. (Eds.): Cyberspace 2005 (pp. 305-310). Brno: Masaryk University.

Yee, N. (2006) Motivation for play in online games. Cyberpsychology and behavior 9, 772-775.