TabbycatDebate / tabbycat

Debating tournament tabulation software for British Parliamentary and a variety of two-team parliamentary formats
https://tabbycat.readthedocs.io/
GNU Affero General Public License v3.0
242 stars 806 forks source link

If we only expect FeedbackFromTeams on the orallist, multiple submissions should unconfirmed each other #473

Closed philipbelesky closed 3 years ago

philipbelesky commented 7 years ago

I assume it is a relatively rare issue that a team would (through malice, or incompetence) submit on all of their panellists and also the chair. However if we expect feedback only on the oralist, whenever a submission comes through for an adjudicator from a team it should probably unconfirm their previous feedbacks for adjs on that same panel.

This both removes those 'bad' pieces of feedback from the feedback scores, but also ensures that the team record pages and expected feedback pages will give a better guide of what is actually going on.

czlee commented 7 years ago

I see, so if they submit on the (not-rolled) chair and then a panellist, only the latter is confirmed causing it to look like there's no submission on the chair and an unexpected submission on the panellist? (This is the effect it would have.)

philipbelesky commented 7 years ago

Yea, that's exactly the result I'd like. The thinking here is that it would hopefully solve two problems at once: the need to 'clean' unexpected feedback from the database and signal to teams that they are doing something wrong. Even if they don't realise what they are doing wrong, hopefully the tracking page at least prompts them to resubmit feedback on the chair one more time and in doing so unconfirmed their previous attempts.

czlee commented 7 years ago

Cool. This actually interferes with the current mechanism a little; we might have to implement it at a different layer. Will think on it.