Commons talk:Picture of the Year

From Wikimedia Commons, the free media repository
Jump to navigation Jump to search
Picture of the Year 2023
Voting unavailable
 Share· Control Panel
weAreWorking
1
POTY 2023 || Introduction – Rules – Discussion – Translations – Committee – Help || R1 Candidates – Gallery || R2 Gallery || Results

Categories[edit]

Splitting off discussion of categories. Copying Lego's links from above: current category count, full R1 candidate list. Also last year's discussion.

Splitting categories discussion[edit]

Last year three changes were made to the categories:

  • Split Sports out of People
  • Split Birds taxonomically (passeriformes vs. everything else)
  • Merge Maps into Works on paper

This year, merging maps into works on paper makes clear sense again. I don't think that will be controversial.

We do not have so many sports photos this year, so if we're to divide the people category again, we'll need a different distinction. Things that could be separated: historic portraits, photos from before 1980, posed portraits, etc.

Last year when the idea of splitting birds was suggested, I floated the idea of splitting out passeriformes based purely on the fact that passeriformes account for about half of all bird species, thus probably account for about half of our FPs. There were a couple other proposals, like a split by geography (where the photo was taken) and separating birds in flight. If there's still interest to split the category, I still support splitting passeriformes. From an glance at the nominations, only about a fifth or fewer are in flight, which doesn't accomplish the goal of a relatively neat split. — Rhododendrites talk |  18:14, 23 January 2024 (UTC)[reply]

  • Many people have no idea what a Passeriform is (I often have to check) and I know that some people imagine hummingbirds are Passeriformes. This split would confuse voters. Birds in flight is a sensible split as the photographic/technical challenge is completely different. Geography makes no sense. 20% is what you might expect as it is so much easier to take sharp pictures of stationary birds. Charlesjsharp (talk) 21:40, 23 January 2024 (UTC)[reply]
  • 20% is what you might expect as it is so much easier to take sharp pictures of stationary birds. Yes, but it's also what makes it a poor way to divide one large category into two smaller ones. I don't know that it matters if people don't know what a Passeriform/passerine is, but we can also call them "perching birds" or even go down one taxonomic level to "songbirds" (which constitute 80% of passeriformes). — Rhododendrites talk |  22:10, 23 January 2024 (UTC)[reply]
  • They are. They're just vernacular names for members of the taxonomic categories. A bird in the order Passeriformes is a passerine or a perching bird. A bird in the suborder Passeri is an oscine or a songbird. Passeri comprises about 80% of passerines. — Rhododendrites talk |  13:57, 24 January 2024 (UTC)[reply]
I'm with Charles on this one, I don't really find the taxonomical distinction that useful, and would prefer something more visual like motion vs stationary. My rough count is there are 16 pics of birds in motion, which I think is probably just barely enough for a category. But it leaves us with a birds category of ~100 images, which hasn't, on its own, really solved the problem of a too-large category. Legoktm (talk) 18:11, 25 January 2024 (UTC)[reply]
Regardless of the nature of the separate category, I don't think it's worth considering the addition of a 16-image category, and in general I'd think we should be looking for alternatives if another category has that few (once the initial sorting is done, of course). I'm going to add another possibility to the other subsection. — Rhododendrites talk |  18:34, 25 January 2024 (UTC)[reply]
  • I have a better idea (I think) Rhododendrites. So I start with this assumption: We want each category winner to be chosen from a group of images that have received good numbers of votes in Round 1. So if only one mammal got into Round 2 it would be silly to have a category "Best mammal". Why not start with a bunch of categories like last year, but tweaked, none holding more than, say, 80 images. Then, after Round 1, choose the (say) 200 most popular images and arrange them into 10 categories of between 15 and 25 images. Sorted... Charlesjsharp (talk) 09:06, 24 January 2024 (UTC)[reply]
  • I don't think I understand. What is the role of the categories in the second round? The second round is usually one group with the highest 30 vote-getters overall plus the two highest vote-getters in each category if they weren't among the 30 top (and we pick POTY from that group). — Rhododendrites talk |  13:57, 24 January 2024 (UTC)[reply]

Proportional representation in R2[edit]

Another idea which came up last year, which seems worth a separate subsection, is instead of splitting large categories, to increase the number of photos from those categories that go on to R2. It's an interesting idea, but the main objection I'd have is that smaller categories aren't just about representation -- it's also about user experience. A large gallery gets unwieldy and overwhelming IMO. — Rhododendrites talk |  18:14, 23 January 2024 (UTC)[reply]

Overall I like this. I think it adds complexity to the rules and requires a bit of math, but saves us from needing to redo the categories every year just because people took photos of different things (which will always happen!). Last year we had an average of 44 pictures per category, with a standard deviation of ~25. Those are reasonably round numbers, so my proposal would be:
  • Categories with 45 or less pictures advance the top 2.
  • Categories with 46-70 pictures advance the top 3.
  • Categories with 71 or more pictures advance the top 4.
In practice, this would have advanced the following extra images last year:
Legoktm (talk) 18:41, 25 January 2024 (UTC)[reply]
(Edit conflict) I was writing this at the same time as Lego, so this is intended to add on to the top post Adding to this, what if we work backwards and set the number of R2 candidates we want, then calculate what proportion of those should come from each category? It seems like there's opposition to the idea of setting up the POTY structure to influence nominations throughout the year, which, as I understand it, means POTY structure should follow what's nominated throughout the year. For the sake of an example, let's say <30=1, 31-60=2, 61-90=3, >90=4. So, based on pre-sort numbers, R2 would be the top 30 vote-getters overall plus the top 2 arthropods if they weren't in the top 30 overall, the top 1 astronomy if it wasn't in the top 30 overall, the top 3 interiors of religious buildings that weren't in the top 30 overall, etc. I don't know just how much of a difference that would make, practically speaking, since the bigger categories are also typically the ones that attract a lot of votes (i.e. people and birds probably still wouldn't benefit from the increased allotment because there are often that many people and birds in the top 30 overall). Just an idea. — Rhododendrites talk |  18:44, 25 January 2024 (UTC)[reply]
Looking at last year's R1 results shows that only 3 categories had a 3rd place image advance: Waters (31 pics total), Astronomy (23), and Artificially illuminated outdoor spaces (16). Waters also had 4th place advance. I was assuming the same as you, that larger categories wouldn't really benefit from this, but it seems like that's not the case.
Aside from the specific numbers, the main difference I see between our proposals is that I kept the floor at 2 images for every category, while you had small categories only promote 1. My rationale for promoting at least 2 images is mostly because it's the status quo and I think it helps with diversity (this is feelings, would be interesting to look at data for how well #2 images do, and what, if any, correlation R1 votes have with R2 votes). Legoktm (talk) 19:08, 25 January 2024 (UTC)[reply]
@Legoktm: It looks like the scheduled start date is in about a day and a half now, so maybe run with your proposal here? I can be around tonight to help categorize. — Rhododendrites talk |  19:21, 30 January 2024 (UTC)[reply]
@Rhododendrites: those are just placeholder dates, we're certainly not ready to open voting in a day, still need to test/fix the voting gadget, request CentralNotice, send MM, etc. I think we can aim to start in a week? Legoktm (talk) 06:22, 31 January 2024 (UTC)[reply]

First pass at categorization done[edit]

I categorized all 300 uncategorized pictures and moved around a bunch of existing ones.

As it stands, we have one tiny category (infrastructure, with 16) and two large categories (birds with 122 and people with 103).

Please review my work. I'd especially like to highlight that I still don't think we have a clear definition for "panoramic views". I went with "aspect ratio greater than 2:1, not including objects (like shells), not including astronomy, and not including infrastructure". The last decision was really just to preserve what few we have in there.

Also worth noting is that I moved aerial views into the "satellite" category based on the spirit/type of content, and based on past discussions. There are definitely questions of "how aerial", but I feel like it's a fine place to stop.

There are a couple of my own images that were hard to categorize and which I'd like someone else to check. One is the iguana panorama. It's among the widest photos in contention, so using the definition above I kept it in panoramic views. The other is the rainbow with a bird flying by. The focus is the rainbow, but that's hard to categorize (is it water, nature, object, astronomy (?)). Birds are easy to categorize, but it's not a good picture of a bird. Regardless, that's where it is now. — Rhododendrites talk |  20:11, 10 February 2024 (UTC)[reply]

pings: @Legoktm, AntiCompositeNumber, and ZI Jony: — Rhododendrites talk |  20:12, 10 February 2024 (UTC)[reply]
What do y'all think about merging Infrastructure into Constructions and Buildings? Most of those lighthouses I'd call buildings over infrastructure anyway. It looks like People has a significant number of portrait photographs, could be a good category if we wanted to split. AntiCompositeNumber (talk) 03:16, 11 February 2024 (UTC)[reply]

Voting not open?[edit]

On the POTY 2023 page it says: "Round 1 begins on 1 February 2024, 00:00 and ends on 14 February 2024, 23:59:59 [UTC]". Well, I haven't been able to vote and still can't. My account is eligible. --Minilammas (talk) 16:01, 10 February 2024 (UTC)[reply]

@Minilammas: That was a placeholder. In theory the 2023 page shouldn't be linked from anywhere yet. :) @Legoktm: it may be a good idea to replace it with "soon" or something regardless. — Rhododendrites talk |  20:03, 10 February 2024 (UTC)[reply]
So when will it take place, if all goes according to plan? Nardog (talk) 06:43, 11 February 2024 (UTC)[reply]
@Rhododendrites, @Legoktm: is there something we can help? Emha (talk) 16:19, 27 February 2024 (UTC)[reply]
Sorry for being MIA. I'll try to have an update by the weekend. Legoktm (talk) 07:07, 14 March 2024 (UTC)[reply]
@Legoktm, It's been a month. Any update? Shawnqual (talk) 05:08, 12 April 2024 (UTC)[reply]

Waiting for it[edit]

Any date? ★ 21:31, 15 March 2024 (UTC)[reply]

Retrieved from "https://commons.wikimedia.org/w/index.php?title=Commons_talk:Picture_of_the_Year&oldid=867616147"