Commons talk:Picture of the Year/Archive 3

Round 1 results sorted

I created a page where all round 1 results are sorted and can be visually seen.

All links are here :

Genre	Category
All genres	All images (1132)
Animals	Arthropods (47) | Passeriformes (44) | Other birds (77) | Mammals (45) | Other animals (64)
Architecture	Constructions and buildings (41) | Religious architecture (108) | Castles and fortifications (27) | Settlements (21) | Other places (129)
Astronomy	Astronomy and outer space (9)
Nature	Natural views (128) | Natural phenomena (30)
Objects	Vehicles and transport (42) | Sculptures (22) | Other objects (69)
Historical	Historical (47)
People and human activity	People and human activity (54) | Food and drink (19)
Plants and fungi	Plants and fungi (69)
Miscellaneous	Non-photographic media (42)

-- Giles Laurent (talk) 01:04, 15 October 2024 (UTC)

I just added at the bottom of all of these pages a "Statistics" section with the number of voters for any category and the number of voters for that category as well as the percentage -- Giles Laurent (talk) 14:42, 15 October 2024 (UTC)

R1 results posted; finalist voting open

Round 1 of the contest has now closed, and the results have been posted to Commons:Picture_of_the_Year/2023/Results/R1. Voting for the finalists is now available at Commons:Picture_of_the_Year/2023/Gallery/Finalists. This year, we had a total of 119,310 votes from 2,530 voters in the first round, a slight increase from 118,086 and 2,407 last year. This is great to see, and means that the new scripts did not have a negative effect on voting. Ingenuity (talk) 00:04, 15 October 2024 (UTC)

Some round 1 raw results seem to be wrong if you compare them with this page. See also this comment. This should be figured out and fixed as soon as possible -- Giles Laurent (talk) 01:11, 15 October 2024 (UTC)

@Giles Laurent: checking it out now. It looks like it might just be a misunderstanding of the 5% rule between our code, but I'll double check. Ingenuity (talk) 01:25, 15 October 2024 (UTC)

It is not a misunderstanding. For example : this file has only 171 votes, which is less than the 269 required votes to be top 30, and is only the 10th image in it's category which is far from the top 5% of that category. Therefore it should not be a finalist. -- Giles Laurent (talk) 01:31, 15 October 2024 (UTC)

Thank you very much to both of you. Also the accepted rule by consensus was "Votes in favor of top 30 overall + top 5% of each category (if that percentage is not already present in top 30 overall and only to reach that number)}", here. -- Basile Morin (talk) 01:43, 15 October 2024 (UTC)

List of files that are not supposed to be finalist for round 2 but that actually appear on the round 2 voting page (should be removed from this json page) : 1, 2, 3, 4, 5 and 6

.

List of files that are finalist for round 2 but that are actually missing on the round 2 voting page (should be added asap to this json page as they are unfairly missing votes right now) : round 2 voting page : 1, 2 and 3

.

Click here for a full list of finalists (only 59 should be finalist and right now there's 62 in the list but 6 should be removed and 3 added) -- Giles Laurent (talk) 02:00, 15 October 2024 (UTC)

@Giles Laurent: figured out what the problem is. A bug with the script meant that, when finding the finalists, it used the images' original categories instead of their current one. A number of images have been recategorized since the contest started, which means the category sizes have changed. It's not a super big deal, and I'm not sure if it's worth the effort to fix at this point (plus re-running the admin script could cause even more bugs for people who have already voted). Ingenuity (talk) 02:06, 15 October 2024 (UTC)

Yes it seems indeed the change of category during contest was not taken into account by the poty script. In any case this json file should be edited manually by removing pictures that are not finalists and by adding pictures that are supposed to be finalists. A list of files missing and that shouldn't be there can be found here and then doing a CTRL+F the images that shouldn't be there can easily be removed and the data of the 3 missing pictures can be found with CTRL+F on this page and then copy pasted -- Giles Laurent (talk) 02:17, 15 October 2024 (UTC)

I've added the three missing pictures. However, since people have already voted for the images mistakenly added, I don't want to potentially cause bugs by removing them. Ingenuity (talk) 02:19, 15 October 2024 (UTC)

Thank you Ingenuity for adding the missing finalists!

Are you sure that removing the mistakenly added finalists from the json finalist file would cause a bug? Once it's done it will not be possible to vote for these pictures anymore and a deletion request might be done for the pages. Also, according to the script page for sorting results of finalists in round 2 and more precisely according to lines 790 to 813 finalists are counted with the "countFinalistVotes()" function. This function can be found at lines 749 to 760 and refers to the data of this json page. Therefore simply removing the mistakenly added finalists shouldn't create any problem I think -- Giles Laurent (talk) 02:48, 15 October 2024 (UTC)

Ingenuity, in my opinion the mistakenly added finalists should not be kept because :

- rightful finalists will miss some of the votes they would have had since it's only 3 votes per user ;

- most of the recategorization was done before beta testing and people voted in round 1 according to the categorization. For example this category was completely deleted (click here for the old version) and for example this file is mistakenly appearing in the round 2 voting page only because the file had that page as first category eventhough that category got deleted before round 1 and beta ;

- if pictures were recategorized it's because they belong to their new category and not to their old one and should be judged according to their exact category else it's unfair for pictures that are actually true top 5% finalists

- simply deleting the mistakenly added files from the json shouldn't create any bug in my opinion (see this comment)

-- Giles Laurent (talk) 03:02, 15 October 2024 (UTC)

I've checked the code again to make sure removing them wouldn't cause any bugs, and it looks like it shouldn't. I've removed them now. People who voted for the images removed may need to clear their cookies before being able to re-vote. Ingenuity (talk) 03:13, 15 October 2024 (UTC)

So absolutely no votes were disqualified? That would be remarkable compared to previous years. Nardog (talk) 03:05, 15 October 2024 (UTC)

I see two possible explanations Nardog :

- 2023 eligibility rules are not the same as previous years eligibility rules. Previously you would need to have 75 edits before January 1st and now you can reach the 75 edits after January 1st and vote as long that the account was created before January 1st of this year. This change originates from here. The script can easily check the current number of actual global contributions on time of voting but it is more complicated to check all global user accounts and edit dates to see how many edits exactly there was at January 1st. So the script would previously not prevent a vote from a user with 75+ edits on time of voting but less than 75 edits on January 1st. Such users would only get their vote removed later. Now this is not an issue anymore since eligibility allows to reach 75 edits during the year. Also, last years it was easy to vote manually which is not the case for this year and this forces people to vote through the script, which makes the script successfully prevents from voting people with an account not old enough and with less than 75 global current edits ;

- or second possibility is that the script didn't work properly. -- Giles Laurent (talk) 12:42, 15 October 2024 (UTC)

Leave a few days between r1 and r2 next year might be better. RoyZuo (talk) 13:05, 15 October 2024 (UTC)

Yes that would allow to have time to check if the script worked well or if some things need manual fixing. The problem is that this requires a change of the poty admin script (and was therefore not possible for this year in the short time given as script was completely rewritten) because right now as soon that you close round 1 it automatically starts round 2 -- Giles Laurent (talk) 13:09, 15 October 2024 (UTC)

In the past there's been 24 hours, which is usually enough. With the new scripts this year it would have been prudent to have a longer gap. AntiCompositeNumber (talk) 23:47, 16 October 2024 (UTC)