A Home For Every Hero

Frequently Asked Questions

• What data sources were used? 

  To measure the audience values that form the foundation of the map, we used surveys that explore attitudes on race, gender, place, and class, along with core values. Using these results, we created four audiences and a suite of predictive models that helped us assign audience scores to the nationally representative, opt-in media consumption panels based on demography, geography, and lifestyle features (e.g., age, gender, race, zip code).

  Media consumption panels gave us visibility into the minute-by-minute media behaviors—of over 300,000 people in the U.S., who opted in and are compensated for their participation—across desktop, mobile, tablet, and TV. This allowed us to enrich our audience profiles with the media artifacts—videos, news articles, TV shows, Tweets, Pinterest pins—that actual audience members actually engaged with.

  We measured audience interest in games using internet searches conducted by more than 250,000 panelists included in a research panel maintained by Comscore. We generated a list of relevant games using three methods: 1) top grossing games from 2021, 2) search terms including the word “play," and 3) a priori recommendations from our staff’s gamers. We retrieved more than 6,000,000 searches relevant to 212 games.

• How do you estimate membership in the audience zones?

  We estimate membership in each of the 8 audience zones using big values surveys and predictive models. We have a good guess about audience membership for every person in the anonymous panel who has supplied demography, lifestyle characteristics, and geographic information.

• How did you identify what you call “distinctive media”?

  We analyzed all the game mentions to find those that were most distinctively associated with each of the 8 audience zones. Importantly, distinctive engagement doesn’t mean that a particular audience consumes the most of a game—the If You Say So audience games more than other audiences, so they play more of lots of individual games than anyone else. Distinctive engagement means an audience over-indexes on a game relative to how much they play games overall. Focusing on distinctiveness helps to identify the games that speak uniquely to each audience.

• Did you supplement data about gaming with anything else? 

  Since the goal of this project is to understand all the narrative components of each audience’s storyscape, we supplemented the games data with data about movies and TV series that each audience was particularly interested in. In the past, we have explored this using data from individual platforms like YouTube or TikTok, but here we didn’t want to constrain the medium or the platform, so we used user searches again and retrieved everything (more than 2,000,000 queries) that mentioned “show,” “series,” or “season.” This is only a sample of the media that interests fans, obviously, but it yielded distinctive titles like Haikyuu, Yellowstone, and The Ellen Degeneres Show for the 8 audience zones that supplemented our narrative analysis.

• How did you create the images that appear on the site?

  For each distinctive game and media property, we collected the setting, tone, themes, core conflict, protagonist portrayal, and antagonist portrayal. For games, we also collected information about core in-game rewards. We analyzed all these story components to find the common elements within each audience that set that audience’s storyscape apart from those of the other audiences.

  To help create a visual sense of the storyscape for each audience, we used Midjourney, an artificial intelligence image generator, to create examples of settings, heroes, and antagonists which shared the characteristics of games and stories appealing to the audience and were “inspired by” those narratives. See more about this process in the If You Say So case study above.

  It’s important to acknowledge that outputs from AI tools like Midjourney represent averages, and may inherently prioritize big and popular types of media due to the nature of the data sets it draws from. This means that certain images or portrayals that represent prevailing visual trends in media are often featured, like female characters with exaggerated and unrealistic body features as in People Power + If You Say So, because the generated images capture major themes in patterns across storyscapes.

• How were the specific media examples chosen for inclusion on the media consumption map?

  The figures/characters featured in the map of media consumption were based on actual findings. The inclusion of specific content was driven by quantitative data analysis of actual user search patterns and interests.

  The representation of race and gender in the media consumption map is the result of user searches and interests, all of which influence the prominence of specific media and its characters. The data in this map show the current landscape of media as it currently is, and how audiences are actually consuming media, based on actual user behavior and analyses.