One of the problems of relying on user reviews to rate content is the possibility that some of those reviews are not written in good faith. Recently, Rotten Tomatoes took new measures to manage the impact of the fake reviews sent by Captain Marvel, while Netflix responded to several cases of “bombing review” by eliminating written reviews of its service completely. Over the years, Steam has taken some different steps to solve the problem, but now its latest response is a combination of automatic scanning and human moderation equipment.
In a blog post, he explained the plan: “We are going to identify off-topic review pumps and we will remove them from the Review Score.” In practice, what you have is a tool that monitors reviews in real time to detect an “anomalous” activity that suggests something is happening. Alert a team of moderators, who can review the revisions they will investigate and, if they find that there are a series of “off-topic revisions”, will alert the developer and eliminate those revisions in the manner The game score is calculated, although the revisions they will remain.
Interestingly, players can choose not to receive the filtered scores, in case they prefer the results calculated with the pump activity included. Other details mentioned explained that complaints about problems such as DRM could be considered off topic and that, although their automated tool has been trained in all previous reviews of the store, they will not apply this approach to existing ratings and reviews.
The last time we heard about significant changes in Steam to combat bombing and spam was in 2017, but the team says that this is not the only adjustment that is coming.