I’ve recently discovered Stack Exchange has a feature called tag warnings which cautions a person with further information about how a tag should be used. Tag warnings look like this:
This example comes from Biology SE’s [species-identification] tag. You can try it out yourself here — click into the tag field then back out to see it show up.
I think that such a feature would be beneficial for our rules-as-written tag, but that’s just me, and I’d like to know what the rest of us think about this. Should we have such a tag warning on the [rules-as-written] tag, and what should it say? (Related material: What, exactly, is the RAW tag for?)
Specifically, I think it’s beneficial because of these statistics:
$$ \begin{array}{|l|l|l|} \hline \text{Questions created...} & \text{in last 16 months} & \text{17-32 months ago} \\\hline \text{RAW tag removed} & 235\;(52.2\% \text{ of total}) & 90\;(15.9\%) \\\text{RAW tag added} & 29\;(6.4\%) & 59\;(10.4\%) \\\text{RAW tag kept} & 186\;(41.3\%) & 416\;(73.6\%) \\\hline \text{Total questions covered} & 450 & 565 \\\hline \end{array} $$
This is a table of questions created during the past 16 months and the 16-month period before that which are connected to the RAW tag in one of three ways. 16 months ago is the time period since the “back to tagging basics” proposal† was suggested and endorsed and changed the way we handle the RAW tag, so these time periods were chosen to compare before/after that meta.
This table expresses these quantities:
- RAW tag removed: Questions that had the RAW tag in the original revision, which do not have it in the current revision. (source 1)
- RAW tag added: Questions that didn’t have the RAW tag in the original revision, but have it now. (source 2)
- RAW tag kept: Questions that had it in the original revision, and the current revision. These could be the same revision if it was never edited. (source 3)
- Total of questions in row 1+2+3 for that column. All percentages in a column are a percentage of this total.
(Further notes: (a) Intervening revisions aren't counted. This means back-and-forth add/remove actions don't affect these statistics. (b) These statistics go exclusively by creation date so recent activity won’t change the period to which a question belongs.)
You may notice what I notice in these stats: before 16 months ago, questions almost always kept the tag. Since the "back to basics" meta† we've seen a dramatic change: in the past 16 months, 52.2% of these questions arrive with the tag and then have it removed — more often than the question gets the tag added or kept in total. Out of the 421 usages of the tag by the OP (sum of tag removed + tag kept), 55.8% usages result in it getting removed. On average a question that gets the tag removed arrives every 2-3 days. I’m hoping that adding a tag warning would reduce these cases of it needing to be removed, and would reduce the maintenance effort involved in this tag.
† "Back to basics" changed the way we handle the RAW tag: we nowadays handle it like any other non-system tag, which means we require questions to make it clear in the body why a tag is present, and if it isn’t clear, we seek clarity from the author and either clarify the post or remove the tag. Prior to that meta (before 16 months ago) the RAW tag was handled differently and it we didn’t have a good way to clarify whether the tag should be on a question or not.