How is the average quality percentile gathered from the API?

I'm wondering if anyone has ever used the average quality percentile data point from the API? 

I'm aware of these three quality scores: author quality, content quality, and group author quality (there may be more). Is the average quality percentile simply the three quality scores averaged together? Or does the average quality percentile have something to do with the quality percentile calculation? Or is it something else completely?

I am new to all of this so I appreciate any guidance! 

Parents
  • Are you referring to the User.QualityPercentile?  That is a 1 to 100 percentile ranking to determine where a user falls in the community (when you hover over a user or on their profile you can see this percentage... Top 1%).  That value is based on the User Reputation Score.

    Each content item gets a Content Quality Score that is based on factors like Views, Replies, Votes, Bookmarks, Verified Answers, etc.

    In each group,  each author has a Group Author Quality Score that is the average of all of their content in that group.

    Sitewide, each author has a Author Quality Score that is based on the average of all their content sitewide.

    The User Reputation Score is based on the Author Quality Score plus some other factors (comment count, like count, do they successfully report abuse, have they posted content that was determined to be abusive, etc.).

    You can find additional information on the metrics and adjust how each metric affects the final score in the Scoring section of the administration panel.

  • And one more question please...Is the user reputation score still calculated via the following formula?

    • Participation ([30]%)
    • User activity ([15]%)
    • Content quality ([50]%)
    • Content moderation ([5]%)
    • Abusive content (can reduce score by [75]%)
  • There is an options tab on the User Reputation Score which lets you adjust those values.  The Content Moderation default is 50% not 5%, that may have just been a typo in the old documentation.

Reply Children