p1
edit- Credit distribution will be moved to the archiving phase of the bot. This will:
- (a) prevent nominators receiving credits for hooks which are subsequently pulled;
- (b) ensure that the bot archives the final version of the update, including any corrections, rather than the original version, which may contain hooks which were subsequently corrected or removed.
p2
editA new system will be implemented to encourage a better standard of reviewing at DYK. Under this system, users will, in effect, have the number of nominations they are permitted to add to DYK limited to the total number of successful reviews they have already completed. A bot will keep track of the number of successful reviews that each user has accumulated, and will prevent users with no review credits from adding new nominations.
Please note that while the rough draft of how this system would operate, given below, might look complicated, it should in fact be easy to participate in given that most of it will be done by a bot.
The QPQ requirement as currently established will be abolished. In its place, a new credit template, {{DYKreviewer}}, will be created, in the following format: {{DYKreviewer|username|errorstatus}}.
The DYKreviewer template will function as follows:
- The template will be automatically added to every DYK nomination.
- When a reviewer has reviewed and approved one or more of the proposed hooks in a nomination, they will add their username to the username field of the template. As with the GAN process, only one user may be credited as a reviewer [per nominated article?], though anybody can comment on a nomination. The template or templates are transferred to prep along with the other credit templates at the set building stage.
- If the hook subsequently makes it through to the archiving phase of DYK without being removed, or without substantial corrections, the reviewer will receive a review credit.
- The bot will maintain a record of the total number of review credits accumulated by each user. For each new nomination (DYKmake) made by a user, the bot will subtract one review credit.
- Users will not be able to claim DYKmake credits in excess of the number of review credits they have already accumulated. In effect, users will not be able to nominate an article to DYK unless their accumulated review credits are above zero. If someone attempts to nominate an article which includes a DYKmake credit for a user with no review credits, the bot will not allow that nomination to be posted.
- Users could be awarded a small number of review credits by the bot on first encounter, to ensure a smooth transition from the old system.
The errorstatus field of the reviewer template will operate as follows:
- If a user finds what they consider to be a substantial factual error in a hook that is in prep, the queues or on the main page that requires the hook to be corrected or removed, the user may, at their discretion, fill the "errorstatus" field of the template with the letter "e".
- When the bot finds a review credit with a non-zero errorstatus field, it will not award a review credit for that review.
- If it is determined by subsequent discussion that the hook should not have been removed or that the error was trivial, the errorstatus field will be cleared and the reviewer receive their credit automatically if the set is yet to be archived, or manually if it has.
- A fast and simple review process, yet to be determined, will also be established to enable users to appeal if they think they have been unfairly deprived of a review credit. Users who repeatedly make frivolous appeals may be banned for a time from the appeals process. Users found to have repeatedly set the errorstatus field on frivolous grounds may be banned for a time from doing so.
The overall effect of this approach will be as follows: firstly, it essentially automates QPQ monitoring, as a user must have already reviewed an article to post a nomination. Secondly, it is no longer sufficient to have merely reviewed an article to receive a QPQ credit; rather, one must have successfully reviewed an article, as defined by the fact that the approved hook made it to the archiving phase of the bot without being removed or substantially corrected. This should provide a strong incentive for users to thoroughly review nominations, because if they make a substantial error in doing so that results in the hook being pulled or altered, they will not receive a review credit and thus have one less nomination at their disposal.
Note that a more rigorous approach could also be added transparently using the above system if it proved insufficient to discourage sloppy reviewing. One possible approach:
- Instead of merely failing to provide a review credit for an erroneous review, the bot could add a negative credit for a bad review. Thus, reviewers would have to ensure that they had done more good reviews than erroneous reviews in order to keep participating at DYK. Negative credits could be capped so that a user's total never drops below zero, ensuring that users would always be as good as their next good review, or they could be uncapped, meaning that reviewers would need to have historically done more good than erroneous reviews to continue participating. The in-between permutations are practically limitless.
Also:
- discourages nominations of iffy notability.