Commons talk:Quality images candidates

From Wikimedia Commons, the free media repository
Jump to navigation Jump to search
This is the talk page for discussing improvements to Commons:Quality images candidates.
Archives: 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30
Older: 2007-01, 2007-02, 2010-01, 2010-02, 2011-01, 2012-01, Commons talk:Quality images

Proposing extra daily nominations for active reviewers

[edit]

I propose that we allow those active in reviewing nominations to nominate up to an additional 5 images per day as long as they are reviewing more images than they are nominating. This would be a way to say thank you for extra review work, and to encourage those that are regularly submitting five daily QI candidates to do more reviewing. We can keep track of this by Commons:Quality images candidates/statistics - basically anyone who is in green there would have the extra nomination slots. This should be self-balancing - it wouldn't cause extra work for other reviewers, since at least an equal number of nominations would be reviewed, and the additional nominations would also count against review counts in the statistics. Thoughts? Thanks. Mike Peel (talk) 19:27, 27 May 2024 (UTC)[reply]

Seems like a good idea, especially if this can be tracked automatically - clearly from the perspective of an overall dearth of reviews, and some reviews with little effort made to provide constructive feedback before opposing promotion, more needs to be done in this regard! SM:!) (talk) 13:08, 28 May 2024 (UTC)[reply]
 Support A good motivation to review more. MB-one (talk) 13:53, 31 May 2024 (UTC)[reply]
I agree, this seems to be a good idea Kritzolina (talk) 15:44, 31 May 2024 (UTC)[reply]
I would appreciate that. --Plozessor (talk) 17:20, 31 May 2024 (UTC)[reply]
While this might be an interesting idea, I suggest retaining a hard limit of 10 images per user and day to avoid flooding of the candidate list by very few highly active users. There is another issue. The number of quality images per day is currently mostly above 100 and the suggestion to weaken the limit of 5 images per user and day could raise the numbers considerably. This would also increase the workload on Commons:Quality images/Recently promoted for the very few people who sort quality images on this page. --Robert Flogaus-Faust (talk) 23:16, 31 May 2024 (UTC)[reply]
Definitely agree with the hard limit, which is part of the original suggestion ("up to an additional 5 images per day"). Plozessor (talk) 07:02, 5 June 2024 (UTC)[reply]
 Oppose I think we should stick to five per day for any user. Also very active users an benefit from being forced to select only their best images. --August Geyler (talk) 20:24, 8 June 2024 (UTC)[reply]

Test run?

[edit]

It seems like this idea has support, how about doing a test run over the couse of a month? Perhaps we should trial it during August and see how it goes? Or is it better to avoid the summer, and pick a month like September? Thanks. Mike Peel (talk) 07:27, 3 July 2024 (UTC)[reply]

 Oppose I think we should not even test this, because every user should gave the same amount of nominations and it is part of the idea of QIC beeing forced to select only the best images. August Geyler (talk) 16:29, 3 July 2024 (UTC)[reply]
 Comment I suppose that there might be an issue to resolve before testing this. Today the script shows 6 nominations and 3 reviews or votes for me. I counted the images with comments or votes from me and my nominated images on the current page and in the last seven archive pages (June 28 to July 4) and I found seven nominations and six votes / commented images. Could you please check again whether your script counts correctly? Thanks --Robert Flogaus-Faust (talk) 21:08, 4 July 2024 (UTC)[reply]
[edit]

We talk a lot about reviewing as an important way of participation in the QI process. But there are also other important ways of participation, mainly maintenance tasks, like adding the recently promoted images to the relvant galleries and categories. These things are less visible, but I would like to mention the super important work of Radomianin, Robert Flogaus-Faust and some others here. As Robert already pointed out, people doing this kind of maintenance are the ones who will see a direct increase of their workload due to any changes of the nomination limits. --Kritzolina (talk) 09:34, 3 July 2024 (UTC)[reply]

Very good points. I wonder if there's repetitive parts of those tasks that could be automated to reduce the workload? Thanks. Mike Peel (talk) 12:35, 3 July 2024 (UTC)[reply]
I think this has been discussed before, and one of the suggestions was that suitable galleries (and categories) could be entered when nominating the images. However, this was rejected as too time-consuming, among other things with the argument that the galleries would be obsolete in the near future anyway, as everything would soon be done automagically via wikidata. --Smial (talk) 14:09, 3 July 2024 (UTC)[reply]
I think galleries are more or less obsolete already - but QIC bot still thinks they are important and stops working once too many files are unsorted. At least to my understanding. Kritzolina (talk) 14:17, 3 July 2024 (UTC)[reply]
As far as I remember, switching off or reprogramming the bot was also not done because it would be too much effort. Perhaps it would help to simply submit deletion requests for all gallery pages, and to delete the "unaccessed QIC" category at the same time? --Smial (talk) 14:31, 3 July 2024 (UTC)[reply]
I do not support this. QICbot might either crash or just continue moving images to Commons:Quality images/Recently promoted, even if nothing happens there ever. AFAIK, QICbot will not stop working if Commons:Quality images/Recently promoted becomes larger, but categorizing the images would become increasingly difficult. Assignment to the gallery pages is also used to update the sample galleries on Commons:Quality images, which is a nice showcase for quality images IMO. If you know how to automatically archive gallery pages (e.g. once a year), this might be a more useful approach. Anyway, I would not risk deletion of any of these galleries unless QICbot is changed so that it does not use them any more. --Robert Flogaus-Faust (talk) 19:56, 3 July 2024 (UTC)[reply]
I can change QICbot, that part shouldn't be a problem, focus on what would help improve the process. Thanks. Mike Peel (talk) 07:25, 4 July 2024 (UTC)[reply]
Of course, that wasn't a serious suggestion on my part, but a kind of sarcasm. I'm quite familiar with the problems with gallery pages that are too large. I bravely categorized them in the past, but gave up because of my unreliable and slow internet connection at home. Nowadays, when I split up and archive gallery pages that have become too big, I occasionally do it at work. I can't say whether it's worth programming a bot for this, because it doesn't take much effort to do it manually. Someone just has to do it. --Smial (talk) 11:04, 4 July 2024 (UTC)[reply]
Thanks for this and thanks for your contributions to image categorization. If you look at the statistics, it is rather unlikely that there would be too many extra images at the moment. So the increased number of quality images might not be a major issue. --Robert Flogaus-Faust (talk) 20:59, 4 July 2024 (UTC)[reply]

Experiment preparations

[edit]

Given the above discussion, I think we can go ahead with a test run. It would just be for the month of August, and we can re-evaluate it after then based on the results of the experiment. To quantitatively measure the success of this test, we can look at the numbers of nominations vs. reviews at Commons:Quality images candidates/statistics (we now have a baseline going back to March); the overall numbers of images going through the process; and the number of unassessed images. Implementation-wise, how would this notice look:

During August 2024, anyone who is reviewing more photos than they submit can nominate up to 5 additional photos per day. Please check the statistics (updated daily) and only submit extra nominations if your row is highlighted in green. This is an experiment: share feedback here.

Since I'm proposing this experiment, I won't participate in it, and will stay within the regular 5/day nomination limit. Pinging those who have commented in this discussion so far: @Scotch Mist, MB-one, Kritzolina, Plozessor, Robert Flogaus-Faust, Augustgeyler, Юрий Д.К., GRDN711, and Smial: Thoughts? Thanks. Mike Peel (talk) 20:38, 30 July 2024 (UTC)[reply]

 Comment Still unsure that this is the right solution to the perceived problem but let's give it a go for a month. If nothing else, we may understand the issue better, particularly if the August results are significantly different from previous months. --GRDN711 (talk) 01:44, 31 July 2024 (UTC)[reply]
 Support Ok from my side but let's hear other opinions. Plozessor (talk) 04:07, 31 July 2024 (UTC)[reply]
 Oppose When the last test run was suggested we only had an opposing vote and a comment that talked about technical issues. Going from there to "Experiment preparations" doesn't seem appropriate to me. --Kritzolina (talk) 06:09, 31 July 2024 (UTC)[reply]
 Oppose for now. I still do not understand why the script produced the odd results I complained about earlier. There was no answer. --Robert Flogaus-Faust (talk) 07:18, 31 July 2024 (UTC)[reply]
 Oppose Thank you, Mike, for your dedication and your idea to increase the limit. I have already expressed my negative opinion on this matter and, unfortunately, I must do so again. I consider the limit of 5 nominations per day per nominator to be sensible and appropriate. The concept behind this is important to me, as it forces all participants to focus on their best photos and make their own selection. It already happens occasionally that participants nominate five almost identical, technically questionable photos and leave it to the community to find the one picture in the series that meets the criteria. Therefore, I cannot support your idea of a test phase, at least not under these circumstances.
To convince me otherwise, it would take more than the simple and often repeated wish for such a test. Clear criteria would be necessary: What is the test phase supposed to examine? Which parameters should be compared after the test? It is essential for such a test to have a clear and measurable test goal (or goals). Therefore, I suggest you either define a concrete set of parameters and a test goal, or cease your efforts in this matter. I thank you and hope my dissenting opinion does not upset you. --August (talk) 09:12, 31 July 2024 (UTC)[reply]
 Support I am very interested to see, if we can increase reviews this way. Re:Augustgeyler: QI is not about the best of the best (that's FP), but to promote a certain quality standard. MB-one (talk) 13:32, 31 July 2024 (UTC)[reply]
I know. But that wasn't my point. It is about images of good technical quality. To get only images with quality nominated, nominators should be motivated to sort out those not (or less) meeting the criteria. --August (talk) 14:05, 31 July 2024 (UTC)[reply]
Yes, but I have more than 14000 picture in Commons, I find more than 5 "images of good technical quality" among those. While I see all the other points, I do NOT agree with the 'the limit forces participants to focus on their best photos' part. Plozessor (talk) 15:44, 31 July 2024 (UTC)[reply]
The actual rules make it possible to nominate 1.825 images per year. That is a lot. How many Commoners do we have here, who are producing more than this amount of good images per year? August (talk) 15:50, 31 July 2024 (UTC)[reply]
Actually I have uploaded ca. 3400 pictures in 2024 and have 700 in the queue for WLM ;) But of course that includes several that are not QI (but are still are valuable). My personal issue is that I've uploaded images for more than 10 years but only discovered QI last winter ;). But I can live with the 5 per day, no problem for me. I would appreciate the change to 10 but I don't have any problem when it stays like it is. Plozessor (talk) 16:08, 31 July 2024 (UTC)[reply]
  • I wasn't expecting the opposition given the above conversations, but fair enough. Better to wait for September then, and I'll try to address the remaining issues as soon as I can. Thanks. Mike Peel (talk) 20:18, 31 July 2024 (UTC)[reply]
Retrieved from "https://commons.wikimedia.org/w/index.php?title=Commons_talk:Quality_images_candidates&oldid=905485193"