True. This is a problem since the current net vote count is mutable, while an individual vote, once cast, is not. You could try fitting a much more complicated model that can reproduce this behavior, calibrate it with A/B testing, etc. Or maybe try to prevent it by sorting according to quality, but not actually displaying the metrics.
True. This is a problem since the current net vote count is mutable, while an individual vote, once cast, is not. You could try fitting a much more complicated model that can reproduce this behavior, calibrate it with A/B testing, etc. Or maybe try to prevent it by sorting according to quality, but not actually displaying the metrics.