Topic on Talk:Article feedback/FAQ

30 edits is a very coarse metric

3
WereSpielChequers (talkcontribs)

30 edits is a very coarse metric. Some articles would be changed out of all recognition by far less than 30 edits, whilst a vandal magnet like en:Beaver will often be completely unaltered after 30 revisions, (last 37 revisions to Beaver).

Ideally you'd want some sort of fancy algorithm that looks at the amount of change in the article and/or whether some sort of trusted rater had altered their rating or rated it very differently to the average rating from before the latest changes. But at a minimum I would suggest that edits that juts add interwiki links, reverted edits and their reversions be ignored by this process, and that minor edits be given less weight than other edits.

Siddhartha Ghai (talkcontribs)

Reverted edits shouldn't be counted in calculating the total 30 edits. Also, if a user/anon(IP) happens to rate a vandalised article poorly, and subsequently the article is reverted to its unvandalised state(prior to the rating), the rating should automatically be removed. I think along with giving minor edits lower weightage, weightage should be decided on the basis of the number of characters added/deleted. A user may mark addition of a lot of content as minor and a spell-correction may not be tagged the same.

He7d3r (talkcontribs)
...weightage should be decided on the basis of the number of characters added/deleted.

On the other hand, there is no guarantee that the number of characters added/deleted in wikicode will have any relation to the other measurements of the "size of the page". E.g. a page with only

{{:August 2010 in sports}}

will be very long even if the source has only 26 characters. So the number of characters may not be a good criteria to decide which versions of an article should be considered important.

There is a script on English Wikipedia which provides some other metrics for the "size" of the pages:

Reply to "30 edits is a very coarse metric"