Ingress: Game evaluation heuristics versus player comments. Do they match?
Björn Strååt and Harko Verhagen, from Stockholm University, examined viability of existing heuristics and how they can answer the issues players experience in video games. Researchers listed most frequent design issues and how are they reflected in online user comments. Results indicate that users focus their criticism on narrative, game graphics and sound rather than the issues the heuristics found.
Strååt and Verhagen used two different heuristic types: top-down and bottom-up. Top-down heuristic approach is where data is gathered from theories, domain high-level expert knowledge and other heuristics within the domain. If the data, then again, is derived from field studies, interviews with end users and domain observations, it should be considered to be bottom-up. Used heuristics in this study that are top-down: literature study combined with participant observation, HEP (Heuristic evaluation for playability) HEP-based PLAY (Principles for game Playability). Last used heuristic can be seen as both: 10 heuristics based on video game reviews from the Gamespot.com website where professional game reviewers were acting as end users, which makes this a combined approach. Strååt and Verhagen also gathered data from play session evaluations and reading comments about selected games in Metacritic.com.
Researchers found that low rated game have more heuristic violations than higher rated games, even some problems that are not in the heuristics. This has led to playability problems. Majority of problems were usability issues such as lack or loss of player control, messy user interface and poor visibility of possible actions. Players then again focused their criticism towards storyline, audiovisual experience and sense of immersion. This means that heuristic approaches by domain experts do not match with the players’ criticism well or at all. Players only mentioned certain aspects of the game when it is done badly, such as game control. Existing heuristics lack in the immersion context where errors in immersion can be found, but not evaluate immersion-enhancing elements.