Currently, users submit new priorities without knowing the specific effects they will have. This is defensible on the grounds that the specific priority values are "inferred" from subjective preference and so do not need to be known in advance. Nonetheless, this is an unexpected and potentially frustrating experience.
Users should encounter a "confirmation" screen as part of the priority flow, in which the results of their new preferences will be shown. Users can "submit" these preference to lock in the effect.
In addition, updates should be semantically consistent, such that a "prioritize" action should never result in deprioritization behavior, for example if a user "prioritizes" at 0.7 a preference which is currently stored as 1.0.
Currently, users submit new priorities without knowing the specific effects they will have. This is defensible on the grounds that the specific priority values are "inferred" from subjective preference and so do not need to be known in advance. Nonetheless, this is an unexpected and potentially frustrating experience.
Users should encounter a "confirmation" screen as part of the priority flow, in which the results of their new preferences will be shown. Users can "submit" these preference to lock in the effect.
In addition, updates should be semantically consistent, such that a "prioritize" action should never result in deprioritization behavior, for example if a user "prioritizes" at 0.7 a preference which is currently stored as 1.0.