GAMEPICKNOW

How recommendations are calculated

Last updated: 2026-09-23

Recommendations first apply must-have conditions, then calculate preference scores. Games that fail required conditions such as free-only, Korean-language support, or low-spec filters are excluded. Player-count logic also distinguishes online games that can be played solo from single-player-only games.

When multiple tags are selected

If enough games match every selected tag, exact matches are prioritized. If not, the search expands gradually to games matching the most tags, and result cards show matched and missing tags.

New, popular, and trending

New uses release timing, while popular uses recent concurrent-player scale. Trending is calculated when the latest seven-day average concurrent-player count is higher than the previous seven-day average and both windows have enough observations. Results can change as data updates.