Issue: More XF trophy criteria

1. How would this be affected when a user edits an article and adds 1000 characters, but then edits another article and removes 1000 characters? How about if they added 1000 characters and removed 1000 characters in the same edit (which the system would diff as 0)?

2. Do you mean:
- A. Average number of views the user made per article (out of all articles)
- B. Average number of views the user made per article (out of articles the user has viewed)
- C. Average number of views other users made per article that this user created (when collaboration protection is on)

VaultWiki doesn't currently track sufficient information in order to calculate any of 2 A,B, or C. If any were ever implemented, it would only be based on new views made after the feature was added.

1. If the user adds 1000 on one article and removes 1000 in another then the total would be 1000.
If the user adds 1000 and removes 1000 in the same article then the total would be 0.

2.
- D. Average number of views other users made per article that this user contributed to.

Also a note with 2 C&D: using an average would actually discourage users from creating new wiki pages, or from contributing on relatively new pages, because those pages would have low starting view counts, which would negatively affect the user's average. How would this affect things if the user already got a trophy, but now the average has decreased too low for that trophy? For D, users might seek to only make barely significant edits to pages that are already popular, because doing so would inflate their average.

By exploiting this, it is possible for User A to take 1 year with hundreds of quality edits to achieve a trophy. But it is possible for User B to take 1 hour using 2 quality edits to achieve the same trophy.

Or, if only views occurring since their edit are counted, then it would encourage users to make 1 edit to a popular page, let it start counting, and not edit anything else (or a 0 would be averaged).

I think a more useful metric might be having an edit (or combined edits) that achieves X views before it is edited by someone else. VaultWiki already tracks enough information to calculate this. A drawback is that this would encourage users to constantly edit over each other, and they may be low quality edits.

Trophies cannot be lost. Once achieved they are eternal.

Its interesting that you mention this dilemma as I have just been discussing this with staff in regards to trophies and views. Measuring author popularity can be done by combining criteria for quantity: amount added combined with views. For example these criteria combines:
- X major edits OR Y minor edits
- contributed to at least X pages.
- average views is at least X

If its possible to measure from the time of edit then this would be nice. Would that not lead to performance issues?
Optimally it would be measured what % of the page was written by the author. Then divide the number of views by the percentage.

We already measure views per edit (on the History tab you can see the number of views each edit received), but we don't keep track of which user made which view.

However, you are right that there might be a performance issue, currently, to include subsequent edits. While we can simply SUM(views) WHERE edit-by-userid = ID to get views while the user's edit was the current edit, to include views since the edit (views of new edits by other users that added on top of it), it would be more complicated, involving temporary tables with millions of rows. It might be more feasible to track views in this way after we start storing edits differently (as individual lines). Then we could have views counters for each line, that would go up after many edits as long as that line was not changed by the edit, and we could tie lines to a specific user who added the line.

In the mean time, only basic views while it was the main edit (which includes any views by viewing the edit directly in History) could be used as a calculation, and there is no way to know if a given view was the same user who wrote the edit vs other users.

And perhaps, in order to prevent discouragement, "average views is at least X" only averages views for edits that are at least a certain age; this way new edits will not negatively impact a user's average until it has a chance to accumulate some views. And others who try to abuse the system would have a delay before it positively impacts their average, giving moderators a chance to notice that an edit should be considered minor and not count at all.

Other possible criteria include:
- achieving a high total view/like/comment/share count for all wiki content the user posted
- achieving a high total view/like/comment/share count for a single content the user posted (out of all content, or out of content of a particular content-type)

Will more criteria for wiki article content be added soon? I am hoping to use it for promotions and trophies.