Commons talk:Picture of the Year/2013/Workshop
While POTY is still fresh on my mind from the 2012 round, I'd like to kick-start planning for 2013. I propose the following:
|Round||Candidates||N votes||Supplemental description||Commentary welcomed|
|Preliminary||All eligible pictures||∞||No||No|
|Semifinals||Top 25% AND best 5 in each category AND N expert panel choices||5||Editable||No|
|Finals||Top 5% AND best 1 in each category||1||Read-only||Yes|
— C M B J 12:56, 13 February 2013 (UTC)
Expert category judges
- This looks interesting. One idea for more in-depth review I had (this was done in previous years) is having expert cateogiry judges evaluate categories. Mono 23:05, 16 February 2013 (UTC)
- I'd be willing to get behind something like that as long as the expert picks don't undermine community picks. Do you see this as being best implemented in the semifinals, finals, or both? — C M B J 05:26, 17 February 2013 (UTC)
Semifinals or alternative
- Do you really think it worth to spend resources on "Semifinals"? Don't you think it would bore many voters to go through one more Round? Who will check votes and determine the semifinalists and prepare for another round? I personally hesitate to increase one more round.
- I suppose you may have similar effects (with much less resource) if you make a midterm report in a sortable table that contains thumbnail, vote count, category of all the candidates with links to votepages - so that voters can sort the table by votecount or by category&votecount and find which image is likely to be a finalist and which is not. --miya (talk) 05:37, 18 February 2013 (UTC)
- On merit, I do think that it's worth the resources to have a semifinals round now. We have seen more than a three-fold increase in FPs since the inception of POTY. The process has not yet been updated to reflect that trend.
- On user engagement, I do not believe that a semifinals round would necessarily bore users. I attempted to evaluate all 988 pictures in the 2012 contest but inevitably overlooked quite a few good candidates just because they didn't make sense at a glance. When the pictures were resorted by their R1 ranking, I experienced a lot of new content. But if that is still a concern we could always just quietly announce the first or second round and ramp up solicitation as usual for the other two.
- On workloads we have several options, but by far the best would be to try and streamline the process a bit more so that the additional burden is simply mitigated. I would like to see if we can create an algorithm that utilizes our existing tools for improved efficiency or even full automation. — C M B J 06:51, 23 February 2013 (UTC)
Best from each category
- I think you're putting too many images in the finals. Isn't the point of the intermediate round to cut down on the number in the finals? Taking the top 5 from each category would increase the number in the finals instead of decreasing it. To that end, if we have an intermediate round, I would rebalance it to top 25% and top 5 from each category and leave the final at top 5% and top 1 from each category. Similarly, I don't see a reason to increase the number of votes in the final round - 1 makes sense, why change it to 2? – Philosopher Let us reason together. 06:08, 18 February 2013 (UTC)
- To elaborate, the reason for the "top x" from the category in the first place is to ensure that all categories are represented in the final, not to override the decisions made in the previous round(s). – Philosopher Let us reason together. 06:11, 18 February 2013 (UTC)
- The impact of promoting N from each category is less significant than it seems because many of them will double qualify for advancement:
- I'm inclined to agree that shifting the top x focus to semifinals is a reasonable proposition. However, there were several viable candidates below the 25th percentile this year, so I think we really need to implement additional strategies if that's going to be our cutoff. I have updated the table to include Mono's suggestion as a starting point, as well as some of your suggestions. — C M B J 12:59, 20 February 2013 (UTC)
Description and Commentary
It looks a good idea to put supplemental description "Read-only", but who will do that? You will? and are you sure you can really neutral to all the candidate when creating description? Are you going to make it only in English? As to POTY2012 finalists I nominated them to be POTD and made use of POTD templates to be Finalists' description. Nominating around 40 images to be POTD and creating their translation templates took quite a lot of time.If you are going to do it to "Top 33%" of R1 (nearly 300 or so), that would be a great job. Are you sure you can really do that? --miya (talk) 08:29, 26 February 2013 (UTC)
POTY2006 and POTY2007 had a place to comment (though I don't know how). If you can prepare a system for the voters to comment, perhaps it could be a substitution of "Supplemental description".--miya (talk) 08:44, 26 February 2013 (UTC)
- Suppose the descriptions are decentralized and those who support a candidate in the semifinals are invited to create or edit a description. If even five or ten people believe in a candidate's potential, at least one of them will probably make sure it gets described well in their language. This is the most sustainable long-term model.
- And while not exactly my original plan, we could also consider just integrating this process with a picture's regular description, which might doubly benefit the project by crowdsourcing improved and non-English descriptions. We could make the existing descriptions available as an icon with option to mouse-over or click for a pop-up that exists starting in the first round. This could be done with xml output and simple processing for "description en lang-en" etc. — C M B J 04:18, 28 February 2013 (UTC)
- The preliminary and semifinal rounds could also probably afford to show a full description in at least one prominently available view. The finals should show them in every view. — C M B J 09:11, 28 February 2013 (UTC)
Historical Data Year Candidates Finalists 2012 988 44 (4.45%) 2011 599 36 (6.01%) 2010 783 35 (4.47%) 2009 890 38 (4.27%) 2008 501 51 (10.2%) 2007 514 28 (5.45%) 2006 321 11 (3.43%) Average: 5.33%