Abstract: This paper reports on results of a statistical analysis correlating superhero characteristics such as powers, motivations, weaknesses, and costumes with commercial viability as represented by comic book sales and number of appearances in new media such as cinema and television. Results indicate that features of a character have little impact in the comic book market, and that new media trends support a move away from god-like, untouchable heroes to heroes displaying more human frailties and highly visual super abilities.
The superhero is an ingrained part of popular culture that has seen resurgence in the last decade through the introduction of prominent characters to new audiences through expanded media. There was a time when superheroes were entrenched firmly in the comic book medium. However, advances in technology and changes in distribution channels have changed the superhero industry considerably. With comic books barely achieving a fraction of early sales volumes, publishers are trying to revive their comic book heroes by taking them from the “Silver Age” to the silver screen and other fora. However, with greatly advanced media come great financial risks. In order to secure their investments, the sponsors of these types of ventures must have a solid understanding of the modern superhero audience and what it is looking for in an iconic protagonist.
Heroes and villains express cultural values regarding what society reveres as admirable and fears as deviant (Klapp 1962). Bradford Wright’s (2001) book “Comic Book Nation” outlines the evolution of comic book stories and characters mimicking cultural change; however, this discussion was done retrospectively (as were others). These types of observations tell us that what was popular during the golden age of comic book heroes will surely not be desirable in today’s society. It would be useful if publishers were able to determine what readers are drawn to now so that they might better meet the demands of this generation of reader. Empirically studying the popular appeal of superhero traits can give us insight into our society and trends within its development.
What makes a character popular or prominent as we enter the Aquarian age? How important is boundless virtue or courage in the face of insurmountable odds? How relevant is the softer, vulnerable side of a character? Do readers have preferred super powers? Is an Achilles’ heel necessary? Moreover, what characteristics are enduring? Answers to these sorts of questions should provide the insights that will tell companies which avenues should prove the most lucrative and least risky to pursue in bringing superheroes back to the masses. The aim of this study is to analyse trends in superhero characteristics empirically in order to establish current popular superhero traits. Following standard practice for writing academic articles in the applied social science of marketing, study findings are then used to develop a series of managerial implications for businesses. In this instance, findings may be used to reduce risk in choosing existing heroes to introduce to modern audiences, or to create a new breed of hero that is better able to reach the target readers, viewers and players of the future.
Superheroes made inroads into popular culture in the 1930s at newsstands around America. The “Golden Age” of comics set a precedent of seemingly indestructible superheroes who held secret identities and were dedicated to upholding truth and justice (Bongco, 2000). After struggling to maintain reader interest through the early 1950’s, the “Silver Age” of comics saw revamping and reintroduction of characters thought to be buried deep in the archives at the Hall of Justice, this time bringing some more human qualities to the fore. Now, “when you think of Superman, you most likely think of the Superman that was in vogue when you were a child. There’s been a Superman for every decade since the character was created” (Fingeroth, 2004, p. 20). Even today, one man’s Dark Knight is not necessarily another boy’s Batman. The eighties and nineties saw a bold break from tradition as comic publishers started directly targeting more mature readership by introducing not just human struggles within the characters, but hints of amoral and anti-heroic tendencies as well (Bongco, 2000). Do superhero fans of today prefer the altruistic heroes of yesteryear like Spiderman, the bad-boy good guys like Wolverine, or looking into the darkest depths of the Dark Knight?
The comic book companies that own our superheroes have used in-your-face tactics to increase sales figures before. The Justice League of America and The Avengers were, after all, “the fictional result of the real world commercial desire to gather the company’s most popular heroes into one magazine or animated series” (Fingeroth, 2004, p.104). However, when companies are betting multi-billion dollar budgets on public interest in particular superheroes, they should be sure they are choosing wisely. Only the most popular of the comic book superheroes can make it to the silver screen or video game monitor. Which character is the right one? Which iteration is the right one? Or should we abandon our superheroes of yesteryear and create new ones, like Buffy the Vampire Slayer, exclusively through new media to show us the way?
It is impossible to have a discussion of superheroes without the issue of gender being raised. Superwomen generally are parts of super teams rather than stand alone characters, and often they are in relationships with male counterparts. However, both partners need not be super, and significant others can be used as plot devices in stories. Therefore, it is worth exploring whether being in these sorts of relationships increases a character’s popularity.
What do we look for in our superheroes? Kevin Smith (2003) has a simple view: “altruism and clothes” (p. unknown). On a more serious note, Fingeroth (2004) asserts that strength of character, positive values, and a determination to protect them start the list. So what separates superheroes from our other heroes? Bongco (2000) offers a costumed, secret identity (p.102). Fingeroth (2004) argues that they have to possess skills and abilities that normal humans do not. Physical, mental and sensory attributes may be heightened. They may have superhuman powers that can range from magical gifts, to scientifically engineered mutations, to high-tech equipment or weaponry combined with an inexplicable ability to walk away from every battle unscathed. But most importantly, they always have to win in the end. By considering a wide range of these characteristics, we hope to be able to determine which combinations of these characteristics are related to commercial success in a modern market.
A database of superheroes was constructed incorporating a wide range of superhero characteristics. The purpose of this article is to determine what characteristics are most popular and commercially viable in today’s market. Marvel and DC Comics make up two thirds of the comic book market; thus it was decided that limiting characters to those in these comics would not be detrimental to sampling frame representativeness. An initial breakdown of character attributes and super abilities, including numerical rankings, was derived from information from fantasy role-playing games written for the Marvel and DC universes, specifically Marvel Super Heroes by TSR, Inc., and DC Heroes by Mayfair Games.
Independent variables included attributes and powers listed in the role playing games, as well as variables for their secret identities, weaknesses, motivations, costumes, significant others, and whether they are human, not human or mutant. The Marvel Super Heroes template and numeric scale was used and DC characters were translated from the DC Heroes system into the Marvel one. Both systems relate their numeric rankings to feats accomplishable in real-world terms (e.g., able to lift 100 tons) and the attributes and powers are similar enough that this translation is formulaic and involves negligible speculation. Costume variables consisted of whether the heroes wear a cape and whether their costumes are primarily spandex. Significant other variables included whether the significant other is aware of their secret identity and whether he or she is used as a plot device; however, because not enough information was readily available about this variable it could not be used in the final analysis. Cape and secret identity were dichotomous (yes/no) variables. Weaknesses, motivation, humanness and spandex were short scales. Weakness scaled from none to debilitating (0-2). An example debilitating weakness is Superman’s vulnerability to Kryptonite. An example intermediate weakness is Iron Man’s alcoholism. Motivation ranged from none-given (0) to vigilante/revenge (1) to sense of duty (2) to desire to serve (3). Humanness extended from not human (0) to mutant (1) to human (2). Spandex ranged from no spandex (0) to sometimes appears in spandex (1) to all spandex all the time (2). Because the majority of superpowers are unique, they were grouped according to common attributes: powers derived from weaponry or equipment, the power to manipulate matter or energy (external to oneself), powers that are inherent to one’s physical body, powers of the mind, and magical powers. The very common powers of being impervious to harm and being able to fly were given separate entries. These variables were regressed on dependent variables representing commercial popularity to see which combinations of characteristics should prove most successful in the current marketplace.
Two dependent variables were used to determine characters’ commercial success. The first was the collective number of movies, television programs and video games in which they have featured as main characters. The sampling frame for this data was the Internet Movie Database (http://www.imdb.com, 2005). The second variable was an estimated comic book sales figure for 2004. It used the monthly Top 300 comic sales lists published by ICV2.com (2004) to calculate the total number of Top 100 comic issues sold in 2004 that featured each of the heroes as central characters (based on title characters for individual titles and feature characters in the case of group titles). While these data can not be entirely accurate, they should provide reasonable estimates of heroes’ market popularity in popular media.
There are 75 heroes in our database; some individual data points are missing for some characters. This sample is small for the number of independent variables in our analysis. Therefore, the first step was to conduct a factor analysis to group individual measures together to create a smaller, but still representative, variable list. A principal components analysis with a varimax rotation was used. It yielded 9 factor components. Factor loadings above 0.400 may be seen in Table 1. The first factor was labelled “Physical Attributes,” incorporating health, strength, fighting ability, endurance and agility. Wearing a cape also loaded onto this factor, suggesting that particular fashion item is reserved for only the most powerful of superheroes. Apparently the new cape-less trend started by top superhero designer Edna “E” Mode has not yet caught on with the big boys. The second factor, “Sensory Attributes,” included karma, intuition and psyche, which may represent such characteristics as “spidey senses.” The third factor included resources, reasoning ability, popularity and motivation. This combination suggests someone who is smart, rich, resourceful, charming and philanthropic, and was thus labelled “Appeal.” The fourth factor seemed to consist of the most stereotypical traits of a superhero: being bullet proof or indestructible, the ability to fly, and a desire to help protect mankind. As such, it was decided to call this factor “Golden Age” attributes. Factor five consisted of powers derived from weapons and/or equipment and the power to manipulate matter and/or energy, and was thus labelled “External Powers.” The sixth factor was labelled “Secret Identity.” Powers associated with the body loaded negatively with secret identity, suggesting that characters without superhuman defences are more likely to require secret identities to protect themselves. The next factor, labelled “Mysticism,” included magical powers and powers of the mind. These powers were also associated with capes, which are still used by such modern media magical superheroes as Harry Potter. At first glance, factor eight appeared to be a bit of a mosaic of collected variance; however upon further inspection it is quite enlightening. This factor encompasses a weakness combined with powers related to the character’s body and high endurance. This factor captures the Achilles heel of the character, along with the struggle to overcome it; thus it was assigned the label of “Tenacity.” The ninth factor demonstrates the humanness of the character. Amusingly, wearing spandex is positively related to being human. It must just fit better under a wider variety of everyday street clothes.
These factors were used as independent variables in two regression models relating the factors first to appearances in movies, TV shows and video games (new media), and then to comic book sales volumes (as outlined in the method section). Neither of the dependent variables was normally distributed, so natural log transformations were conducted on each of them before analyses were conducted.
When the superhero factors were regressed on the new media variable, residual analysis uncovered a non-random error distribution that suggested weighted least squares was required. Thus, a weighted least squares regression analysis was conducted using unstandardized residuals as the weighting factor. The model returned an adjusted R² value of 0.77, meaning that 77% of the variance in new media appearances for superheroes could be explained by the nine superhero factors (Table 2). Four factors had significant effects on how often a character appeared in new media. Sensory abilities, external powers and humanness had positive relationships with new media appearances. It is easy to see why external powers would be desired in today’s visually focused media. Interestingly, the Golden Age factor had a negative relationship with new media appearances, with Golden Age characteristics making a character 35% less likely to appear on screen. Some of our favourite superheroes have been recreated and revamped over the years. An example of a Golden Age character shedding these stereotypical powers in new media would be the television program Smallville, in which Superman can’t fly and has not yet devoted his life to saving mankind. This example is also an apt one for demonstrating the appeal of the other two factors, humanness and sensory attributes. Directional results indicate that audiences want characters to be less untouchable and more human and to have an innate ability to understand their fellow man. They want to empathize in some way with the protagonist in the story. These results could describe the hardened street smart Batman or the gentler Tobey Maguire version of Spiderman, both of whom are due to make reappearances at the box office, and both of whom will be sporting spandex. Alternatively, these results could be used as a skeleton around which to build new characters to debut in video format.
A second regression analysis was conducted using comic book sales figures estimates as the dependent variable. The mass audience for movies is a much broader one than the comic book niche market, and as such is looking for more sophisticated superhero stories. Thus, what these two audiences are looking for in superhero characteristics may also be very different. As in the first analysis, the 9 superhero factors were regressed on the transformed dependent variable, in this case Top 100 comic book sales. Linear regression results did not require weighting. Results can be seen in Table 3. Only one of the measured superhero factors, tenacity, played a significant role in determining comic book sales. The ability to overcome weakness in the face of insurmountable odds is something that comic book fans have come to expect from their superheroes over the decades. When writers gave their characters flaws in the Silver Age it was meant to make them look even better when they beat the bad guy (Lang and Trimble, 1988). Apparently the formula still works. These results indicate that changing superhero characters, or developing new ones, will have little effect on sales.
This stagnancy with respect to characters may be caused by a general shift among comic book fans away from emphasizing characters and more towards purchasing issues based on their artists (Brown, 1997). Alternatively, this is a small market that is steeped in tradition. Fans involve themselves heavily in these products and are quick to point out inconsistencies. They may simply not be open to change, meaning that the superheroes themselves are hardly considered at this point, so long as they are consistent. In short, this small market is not where growth will happen and is not representative of a broader audience so pursuing alternative characters in this medium could prove to be a waste of resources.
Apart from shedding light on current cultural values, these findings can be applied commercially. While results suggested that the comic book market is not strongly character driven, results indicated that characters have a major impact on on-screen popularity. Television and movie audiences seem to want to see heroes that are intuitive human beings that have powers that are external to themselves, whether they stem from equipment or the ability to manipulate matter or energy outside of themselves. They are no longer interested in god-like characters that can fly and can’t be killed. This combination includes mutant characters. Results can be applied to choosing existing characters to introduce into these media, developing new characters specifically for those media, or altering existing characters to better suit the wants of the audiences viewing those media. For example, Iron Man fits these criteria, making him a perfect choice for an upcoming silver screen venture. Findings also suggest that Professor X would be a more popular character if he were to develop the ability to move things with his mind. Alternatively, an entirely new breed of superhero could be introduced to the world. A league of new and improved, culturally customized superheroes to rise from the wake of Buffy the Vampire Slayer and Harry Potter, never having been restricted by the walls of a comic book panel.
There are several issues that require the reader to interpret these results with caution. First, for reasons of practicality, we use quite a small sample of the most well established characters created by the two biggest comic book publishers for our analysis. Therefore, innumerable other superheroes developed by them, smaller companies and independents, as well as those that did not originate in comics, but in some other medium such as television or movies, are not accounted for. Next, much of our analysis was based on comic book sales figures, the overwhelming majority of which were published by Marvel and DC; however, these sales figures were used as a surrogate measure of character popularity. It must be recognized that collectors now often consider the comic book artists rather than the characters that they create when purchasing issues, as well as purchasing multiple copies of those they believe will be valuable (Brown, 1997). Thus, character popularity may not be as highly correlated with comic book sales figures as we assume in our analysis. Similarly, Marvel, DC and the majority of the new media depicting those superheroes come from North America. As such, our results may not be generalizable across cultures. Finally, audience demographics for Marvel Comics, DC Comics and the various other superhero media discussed here are proprietary, so we are unable to consider how the audiences for these media are similar or how they differ in our analysis. Instead we are making an assumption that there are differences and inferring that comic book producers have been successful in their bids to increase the superhero audience through the use of new media channels.
This study could be expanded to include a more representative sample of heroes from a wider range of publishers and media if a consistent system for quantifying many of the superhero traits could be agreed upon. Findings would be much more robust and reliable if this were achieved. Demographic or cross-cultural comparisons of preferred superhero traits and media could also be introduced. Villains should also be included in any future analyses. These additions would make the findings more inherently interesting because “comic book evildoers…help define the comic book hero” (Fingeroth, 2004, p.15). Should further sociological analysis be of interest, a longitudinal study of these characters’ traits, their associated artists and their evolutions could be compared empirically to the observations provided to date and used to attempt to predict current popular trends and their changing sociological implications for now and into the future. Little empirical work exists in the realm of superheroes as they relate to popular culture, leaving endless opportunity to test the many sociological theories being advanced about them.
Bongco, M. 2000. Reading Comics: Language, Culture, and the Concept of the Superhero in Comic Books. New York, NY: Garland Publishing.
Brown, J.A. 1997. Comic Book Fandom and Cultural Capital. Journal of Popular Culture 30 (4): 13-34.
Fingeroth, D. 2004. Superman on the Couch: What Superheroes Really Tell Us about Ourselves and Our Society. New York, NY: Continuum.
Grinfeld, M. J. 1997. Superheroes Impart Life’s Lessons. Psychiatric Times 14.
Klapp, O. E. 1962. Heroes, Villains, and Fools: The Changing American Character. Englewood Cliffs, NJ: Prentice Hall.
Lang, J.S., and Trimble, P. 1988. Whatever Happened to the Man of Tomorrow? An Examination of the American Monomyth and the Comic Book Superhero. Journal of Popular Culture 22 (3): 157-173.
Smith, K. 2003. The Superhero. Rolling Stone, May 15.
Wright, B. W. 2001. Comic Book Nation: The Transformation of Youth Culture in America. Baltimore, MD: Johns Hopkins University Press.
Bio: Lisa Watson is an Assistant Professor in the Faculty of Business Administration, University of Regina, Saskatchewan. Her research interests include consumer psychology, affect and decision making, and electronic commerce. Phil Stocks is an Assistant Professor of Computer Science in the Faculty of Information Technology, Bond University, Queensland, and a defender of justice in his spare time. His research interests are in the areas of programming languages, programming language theory, and artificial life. Email: firstname.lastname@example.org