Valve has introduced a new system on Steam that prioritizes reviews based on how helpful they are for making purchase decisions. In the past, Steam reviews were ranked by the number of "helpful" votes from users, which often led to joke reviews, memes, and ASCII art getting top visibility—content that’s not necessarily informative for potential buyers.

To address this, Valve’s new "helpfulness system" uses machine learning algorithms, user reports, and input from the moderation team to highlight reviews that offer substantial insights. The system automatically enables a "most helpful" toggle that filters out less informative content, making it easier for players to see reviews that genuinely aid in decision-making.

While this new system is now the default, users still have the option to revert to the old sorting method by disabling the new filter in the display settings.

This update follows another recent change by Valve, which will soon prohibit links in Steam store descriptions, effective from early September. The move is intended to prevent store pages from being used as ads for other Steam content and to avoid confusion regarding "prologue" games.

Overall, Valve's latest adjustments reflect its efforts to improve the user experience by curating more relevant content in the Steam store.