How to do a what now?
A "wick" is an in-wiki link to a page. For example, here is a wick to the "Protagonist Centered Morality" Trope: Protagonist-Centered Morality. Wicks can also be Pot Holes, where the text linked to the page is not the page's actual name, like so.
Often when a trope gets brought up in the Trope Repair Shop for being repaired or renamed, the issue of its use around the wiki will come up. Maybe someone will claim that the title is confusing editors into thinking the trope is something else, or someone will suggest that the title is lacking in the catchiness department and its proliferation is suffering as a result. To support or oppose such a point, you may want to look into how much, how correctly, and generally how a trope is being used around the wiki. In other words, have a look at how its wicks are behaving. In other other words, do a wick check.
Okay, how do I do it?
You can find a page's list of wicks here (replace "InsertTitleHere" with the actual title). If there's already a thread in the Trope Repair Shop discussing the trope, you can just press the "get usage counts" button above the first post in the thread and it will take you to the same page. Another option is to use the "related" button, the second option in the top left of most pages. It's a list of other pages which have in-wiki links to that page. That is, a list of its wicks.
Now, if you want to make a point about the quantity of the wicks, there's not much work left for you to do. The number is right up there - "InsertTitleHere found in X articles, excluding discussions". What this number means is another issue entirely (is the trope overperforming? Underperforming? How many wicks would you expect this trope to have? How many wicks do similar tropes have?..), but there you have it.
If, on the other hand, what you want to do is make some point about the quality of the wicks- the kind of context they arise in, whether they are using the trope correctly, whether they tend to be vague hand-waving or spot-on - you're going to need to actually go in there and look at them individually. Now, theoretically it would be best if you could just click on wick after wick and see how all of them are doing, but we do not recommend doing that if you value your sanity. Instead, you're going to have to look at a representative sample.
How do I get a representative sample of wicks?
- The number of wicks you check should be either the square root of the total number of wicks or 50, whichever is larger. If there are fewer than 50 wicks, check them all.
- The wicks should be picked randomly from the list.
You can check a smaller number of wicks if you feel you're not up to the task, but unless your results are very conclusive, they will likely be challenged on the basis of your sample not being large enough to be representative. As to randomly picking wicks, this is because going in alphabetical order is more likely to run across the results of some clean-up attempt, and generally taking a wick's attributes into consideration will likely skew the results. There are a few ways to randomize it, and each one has benefits and drawbacks:
- Inserting the wick numbers into a randomizer and selecting specific wicks, based on the numbers, to check. This provides true randomization, but is also time-consuming, especially for longer wick checks with more counting to do.
- Inserting the entire wick list into the randomizer and choosing the wicks that come to the top. This is less time-consuming than the first one, but still requires you to ensure that you didn't generate non-pages such as namespace dividers, and may require you to randomize the list in multiple batches.
- Scrolling up and down the list and stopping at random points. This is perhaps the least time-consuming option, good for collaborative efforts or checks done in bursts of activity. However, it takes some of the randomness away, as the troper needs to still choose which wick to check. Care should be taken here to avoid unintentional bias or skewing.
What should I be looking at when examining the wicks?
There are all sorts of properties of wicks you can check for. Here are a few:
- Is the trope being used correctly? Does it actually apply the way the example claims, and is it correct about what the trope means, or does it fit into Square Peg, Round Trope as a shoehorn?
- Is there a clear, correct explanation of how and why the trope applies? Or perhaps an explanation that gives a false impression of what the trope is, or no explanation at all?
- What type of work is it? Does the medium, region, and/or genre match the trope's intended usage, or is it more broad/narrow than it should be?
- Is the potholing relevant to the example, or is it referencing the trope just because it can?
- Is the trope being used for either Gushing or complaining about the work?
- Is the trope being used as a Verbal Tic or Stock Phrase?
- What namespace is it being used in? Is it correctly sorted as per What Goes Where on the Wiki?
- Is it describing the title, but not the actual trope?
- Is the trope being used as People Sit on Chairs, with examples that feel like meaningless "X exists and does some thing" which has context and plot relevance, but doesn't ultimately convey information?
- If the tropes page is soft-split, are the wicks able to be crosswicked back? Or is that impossible?
If there are significant results in any of those areas, you may be able to use them to illustrate a point you are making. For example, if you claim a trope's name is insular, your case becomes stronger if most of the wicks in the sample have to do with comic books (unless, of course, it is a comic-book-exclusive trope by definition). If you claim that the trope is neutral and does not induce Complaining About Shows You Don't Like, your case becomes stronger if you can show that most wicks are from pages matter-of-factly pointing out that the trope applies without judging the work it applies to.
What does a wick check look like?(For convenience, it's recommended to start each wick check at own Sandbox page)
Opening argument about the problem, predictions about the check.
Wicks checked: Progress/Minimum Required.
[[folder:Correct use per trope page's description (X/Total, %)]]
* CheckedPage: The example with the trope wicked, copied verbatim. '''Checker's personal comment'''
[[folder:Other usage categories (X/Total, %)]]
* (Each category needs a separate folder depending on the alleged issue, like misuse, alternative interpretations of the trope, misplacements, potential splits, and etc)
[[folder:ZCE and Unclear use (X/Total, %)]]
* (Lack of context is also a valid reason to believe there is a systematic problem)
[[folder:Non-tropes, like a wick in an index page (X/Total, %)]]
* (Many prefer to test few extra wicks since this category is rarely useful)
* (Since the check starts with picking random wicks, it's recommended to copy them first and arrange them afterwards)
How do I share my findings?
Post to the existing thread in the Trope Repair Shop, or start a new one if there isn't one. Say you did a wick check. Explain what you were trying to check. For each article where you found a wick, copy and paste the line/sentence it appeared in (this allows other users to double-check your results) and state (preferably in bold) which side of the case it weighs on. Conclude your findings with the relevant statistics.
An alternate method of posting results is to sort the results by verdict so that e.g. all "correct" wicks are organized in one list, all "misuse" wicks are organized in a second list and all the "unclear" wicks where you aren't sure in a third list. This makes the results easier to visualize at a glance, unfortunately it takes a lot more work to do it this way.
This is all still confusing- can I have help?
Sure! Wick Checks can be done collaboratively. Some tropers will leave their sandbox open for however long it takes to complete the check, which in turn allows other interested parties to add their own wicks. If that sounds better to you, feel free to stop by the Wick Check Project thread, where you can do your wick check collaboratively and get feedback on it as well.