almost everyone had 30-day karma ratios lower than their total karma ratios
One factor that contributes to this, but I am not sure how large part of the effect it explains: Downvoted articles disappear after a while, while upvoted articles remain visible. Therefore recent bad articles cost karma, but old bad articles don’t; while both recent and old good articles can get upvotes. It’s a bit similar with comments; when they are collapsed, probably people are less likely to look at them.
Please note that 30-day karma means recent votes on recent contributions—it does not include recent votes on old contributions. And I suspect the “recent votes on old contributions” is biased towards positive, because the old bad contributions disappear.
One factor that contributes to this, but I am not sure how large part of the effect it explains: Downvoted articles disappear after a while, while upvoted articles remain visible. Therefore recent bad articles cost karma, but old bad articles don’t; while both recent and old good articles can get upvotes. It’s a bit similar with comments; when they are collapsed, probably people are less likely to look at them.
Please note that 30-day karma means recent votes on recent contributions—it does not include recent votes on old contributions. And I suspect the “recent votes on old contributions” is biased towards positive, because the old bad contributions disappear.