dAtA-TRoN
commented
7 months ago @fast4x I've tagged you here because I've reopened this proposal, and I wasn't sure if GitHub notifies you about these things. Please take a look at it. I consider it to be a very important topic. I had closed this proposal before because I wasn't quite sure how relevant it is. I have now delved deeper into the statistics in RiMusic and found that the current implementation is not particularly meaningful. Hence, the resumption.
fast4x
5wHN28Dg
Description
I would like to request the implementation of two features to enhance the accuracy of song listening statistics.
I suggest incorporating an option to define the duration threshold for considering a song as 'listened to.' Currently, there appears to be no such feature in place, which poses an issue. Upon reviewing my statistics, I observed several songs listed that I'm certain I didn't listen to for an adequate duration. Determining what qualifies as an adequate duration varies for each individual. Hence, it would be advantageous to allow users to customize when a song is marked as listened to, perhaps after one-quarter, half, or three-quarters of its duration.
I propose that statistics should exclusively reflect deliberate actions, disregarding random song suggestions or radio playback (if this isn't already the case).
Suggested Solution
Implement an option in the settings to adjust the time threshold for when a song is considered listened to. Provide users with choices such as one-quarter, half, or three-quarters of the song's duration.
Ensure that statistics are only influenced by conscious decisions. Conscious decisions refer to intentional actions such as selecting a specific song, album, or personally curated playlist. This excludes statistical calculations influenced by random song suggestions or radio playback (again: if this is not already the case).
Important for this: Even when quickly skipping through songs, the previously set threshold should still be upheld. Only when the cumulative duration of all listened song segments exceeds the specified threshold should the respective track be considered statistically relevant. For example, if someone only enjoys the intro of a song and repeatedly skips back after the first 20 seconds, it wouldn't yet meet the statistical relevance threshold, which might be set to a quarter of the song's total duration. In this scenario, additional segments of the track would need to be listened to for the cumulative duration of all heard segments to reach at least a quarter of the respective song. Thus, the statistic isn't focused on tracking the time spent on a specific part of a song but rather on the cumulative duration of all heard segments within a song. This approach ensures that the statistics provide a holistic view of each individual song. This behavior is significant because enjoying a specific section of a song doesn't necessarily translate to liking the entire song.
Example Graphics:
Here is the sum of all listened song parts in one chunk:
The threshold was not reached = irrelevant for the statistics
Again, the sum of all listened song parts in one chunk:
The threshold was reached (exceeded) = relevant for the statistics
Additional Context
The proposed features are intended to enhance the accuracy and relevance of RiMusic's statistics, offering users deeper insights into their listening habits.
The current implementation appears to be suboptimal, as the statistics mimic the listing behavior of a listening history. This approach has a flaw. While a listening history tracks every song, regardless of how briefly played or quickly skipped, the statistics should provide a more precise reflection of actively and consciously listened-to content, thereby boosting their relevance.
Therefore, the implementation of these thresholds is of immense importance to ensure the statistics possess any relevance whatsoever.