User:MargaretRDonald/sandbox/Thoughts on WLE 2021 In AustraliaV2

The outcomes of this competition were reported in the Australian section of Wikimania in August 2021. This report addresses the shortcomings of WLE 2021 in Australia, with recommendations as to how these may be addressed.

Photographs

1. For the effort involved in organising the competition, I felt that the photographs were:
   1. too few
   2. insufficiently various
   3. mostly not of high quality
   4. failed to fill the gaps. See e.g. the following queries:Australian fauna pages which have no image (724), Queensland biota (2385) APNI plants (5758) ( 27 October 2021) and below.
   5. came altogether too frequently with inadequate descriptions.
   6. almost all required enormous work to prove that they satisfied the requirements of the international competition. And this will continue to be the case for the foreseeable future. Nonetheless, however poor an image, it was always valuable to improve and create Wikidata entries for protected areas in order to show that the photograph was taken in a protected area.
2. Culling the approximately 20% images which failed the international requirements spared our prejudges considerably. (Categorisation and Wikidata work will continue to be necessary. We do not have complete lists in either Wikidata or enwiki to permit the automatic uploading of an image to its protected area.) This means that all photographs must be examined, and during this process as many tasks as possible should be carried out:
   1. Categorisation for the purposes of the competition ("needs OTRS", "not from a protected place", "outside the competition", "filling the gaps", "represents NSW", "good description", and so on)
   2. Categorisation to permit photos to be found
   3. Use of the depicts statement (and, if necessary, creation of the corresponding Wikidata item)

I believe we should participate in WLE 2022 but should write our rules to:

1. include those of the international competition, but
2. reward images of places and things not yet found in the Commons or Wikipedia. That is, I would like to see up to three prizes awarded for "filling the gaps" at $100 each.
3. discourage the upload of many multiples of images, with almost no change in focus or framing. (I would suggest that our rules incorporate a message that should this happen then just one image (random or perhaps the first) would continue in the competition with the rest being excluded.
4. offer prizes for the 15 images (or number suggested by the international competition) that are forwarded to the international competition, with $100 for the first 10 and $50 for those whose photos are 11th-15th best. Prizes to be awarded to photographs as was done in 2021.
5. offer a prize for a photo with the best description.
6. judges to be invited to self-nominate via the landing page. Perhaps different sets of judges for different aspects of the competition?

Things which should be done the same

1. Use the German prejury tool. This tool gives a final summary for each photo which shows how many people gave a rating, together with its average rating and the order of the ratings from highest to lowest, and despite some prejudges judging very few pictures, the presentation of the images was sufficiently random to give a relatively uniform voting distribution across the images.
2. Use the Ukrainian judging tool, using the rating scale 1-10. Rating 1-5 did not give sufficient discrimination between images, and using the scales greater than 1-10 was extremely difficult. This tool allows all votes from all judges and judging rounds to be recorded and kept, and hopefully it will be available in a more timely manner in 2022.

Things to be done differently

1. Decisions about prizes, prizewinners and their number to be made prior to the competition.
2. Decisions about all wordings to be made prior to competition (a committee of two).
3. Use a banner which reflects Australia, the aims of the competition, and which is colourful. See e.g.https://de.wikipedia.org/wiki/Wikipedia:Wiki_Loves_Earth_2021/Deutschland Note as neither the Torres Strait Islander flag nor the Aboriginal flag can be used, we should not use the Australian flag. The wording should not imitate the wordings of the European competitions.
4. Allow the competition to run for one month.
5. Allow the judging process as much time as possible.
6. Use the German prejudging tool to cull to about 200.
7. Let the Ukrainian judging tool exclude photographs on size (i.e, not make this part of the task of deciding eligibility, but let it be automatic. This means more work for those using the prejudging tool, less for those (partially) determining eligibility.
8. Then ask the (three) judges to use the thumbs up/thumbs down rating on those two hundred images to choose images which they think should be finalists. Depending on how the judges feel, we would keep images with anything from 1/3 to 3/3 votes.
9. Three judges to allow a majority decision (and the organiser not to be a judge).
10. Use of Wikimedia's big blue button or zoom account to allow judges' meetings to be longer than 40 minutes. The account should permit all users to screenshare.
11. After the thumbs up/thumbs down round, all further rounds would use the 1-10 ranking available on the Ukrainian tool. This gives better spacing of the candidate photos.
12. There should be three judges, to allow a majority decision. The organiser should not be a judge. Five were too many – all five of us did not manage to meet in any of our meetings, and three allows a majority decision. Judges should not be paid. Meetings should be via zoom, since in person meetings would restrict who may participate.

"Filling the gaps": attempting to measure some of the gaps in Australian biota

id used	subset	# Q items	# articles	# articles with no image	% articles	% image
AFD-id	Total	16388	4860	1867	30%	62%
AFD-id (3/12/2021)	Total	17747	4981	1930	28%	61%
	Arachnids	4561	936	524	21%	44%
	Annelids	3124	259	131	8%	49%
	Beetles	1174	14	5	1%	64%
	Beetles (3/12/2021)	4084	214	106	5%	50%
	Sponges	1916	221	120	12%	46%
	Birds	1172	1043	13	89%	99%
	Mammals	544	504	120	93%	76%
	Reptiles	499	474	23	95%	95%
	Amphibians	267	159	105	60%	34%
AusLichen		8262	548	385	7%	30%
APNI		84275	19728	5767	23%	71%
APNI (30/11/2021)		84275	19812	5814	24%	71%
QLD Biota		33115	12890	2387	39%	81%
SPRAT-id		1737	1126	515	65%	54%

A crude check on biota articles with and without images (November 2021) shows Australian insects as very poorly represented in Wikipedia, with just 1% of Curculionidae having article pages. Australian reptiles and mammals mostly have pages (95%, 93%), with reptiles mostly having images (95%), and mammals having images in 76% of articles. Spiders still need many pages written with just 21% of our spiders having pages and of those only 44% have images. Amphibians have 60% coverage in terms of pages, but just 34% of those pages have images.

The queries used to derive these could be used to show missing articles and missing images, in WLE 2022's attempt to fill the gaps. (And hopefully by 2022 a higher proportion of Australian Faunal Directory ids (AFD-ids) will be attached to their corresponding Q items. (The number of such ids should be considerably greater than the order of APNI ids.) (I am currently working on uploading the AFD-ids for valid names of genera and species to permit us to know just which Australian animals lack both pages and images.)

Comments on counts

1. APNI is a "names" database. Hence many of the Q items with APNI-ids are not accepted names, and thus would not (should not) have enwiki pages. If we consider the total count of APNI ids to be inflated by a factor of 4, then 94% of Australian plant pages have been created, and 71% of those have images.
2. Images have been counted as images attached to a Q item. There are numerous plant pages for which an image has not been attached to the Wikidata item. However, most people working on newly created biota pages do this as part of their processing to make elements within a new page as widely disseminated as possible. Thus the count of images within articles is potentially an undercount.
3. Despite these counting problems, the numbers are indicative of major gaps in illustrating Australian biota.

Judging "filling the gaps"

I would like to democratise the judging and adopt the method of judging the Photo of the year, where in the final round any Wikipedian is invited to cast three votes. This would require a banner (ideally across the world) to get enough judges to give a decision. But inviting participation in this way would also be a way of inviting Wikipedians across the world to place the photos within their wikipedias as well as voting.

Some image usage comparisons

State representation
State	# contributors	# images	Proportion
SA	34	56	4%
TAS	21	50	4%
NT	31	138	10%
VIC	40	212	15%
QLD	40	206	15%
NSW (inc ACT)	87	259	18%
WA	50	498	35%

In categorising images by State as part of the background work, the hope is that people might be encouraged to compete across States in seeking to represent their State via images.

Viewing and usage statistics for 2021-07 for the Catgories:Images from Wiki Loves Earth (year) in Australia
Year		files	viewed	used	%files used	pages	# wikis	file views	fileviews/upload	file views /page uploaded
2016		4067	334	/385	9.47%	1402	93	1020871	251	728
2017		1503	109	/122	8.12%	247	52	245177	163	993
2018		987	75	/85	8.61%	122	33	68886	70	565
2019		698	49	/63	8.88%	33	29	88993	127	2697
2021		1532	112	/138	9.01%	278	32	25841	17	93

Not surprisingly, the images from 2021 still lag in visibility, but do show comparable usage.

Year	Country	files	viewed	used	%files used	pages	# wikis	file views	fileviews/upload	fileviews /page
2021	Armenia	6125	47	48	0.78%	39	15	67079	11	1720
	Spain	1323	19	20	1.51%	69	31	52239	39	757
	Portugal	482	9	9	1.87%	7	2	24600	51	3514
	Indonesia	680	13	16	2.35%	21	10	4600	7	219
	Italy	1748	41	42	2.40%	52	12	38148	22	734
	Benin	949	24	24	2.53%	14	3	12606	13	900
	Azerbaijan	231	5	6	2.60%	10	7	519	2	52
	Nepal	585	20	20	3.42%	16	5	52513	90	3282
	Ireland	362	9	14	3.87%	16	4	1025	3	64
	Russia	6216	336	338	5.44%	191	20	1455949	234	7623
	Rwanda	311	18	18	5.79%	24	6	23925	77	997
	Germany	13734	746	922	6.71%	703	17	207559	15	295
	Ukraine	11285	784	790	7.00%	401	4	68945	6	172
	Sweden	1825	281	329	18.03%	405	3	14585	8	36
	Austria	457	117	131	28.67%	148	46	82817	181	560
	Philippines	1012	296	303	29.94%	333	32	122324	121	367