Talk:PAG

Polyalkalene glycol

pag = POLYALKALENE GLYCOL used as a quenchant dilutet i water. —Preceding unsigned comment added by 83.91.53.173 (talk) 12:44, 25 October 2008 (UTC)

on merging Pag or Pags

https://wikinav.toolforge.org/?language=en&title=Pag indicates a substantial portion of readers want the acronym, but a larger contingent actually want a non-acronym, something called "Pag".

https://wikinav.toolforge.org/?language=en&title=Pag indicates the most substantial portion of readers want Polyalkylene glycol, which was placed at only #7 in the list and didn't even have a topical redirect up to now, so the improvements to navigation there should be obvious, and adding in more stuff above it definitely doesn't sound like it would be helpful.

Sadly we don't have enough traffic at Pags last month to render WikiNav, and the clickstream archive doesn't have anything for the last few months.

So that's why I don't see much indication that merging the lists would produce better navigation outcomes. @Shhhnotsoloud do you have some other insights about this? --Joy (talk) 16:14, 23 March 2024 (UTC)

Thanks @Joy:. I've started a merge proposal for further discussion. Shhhnotsoloud (talk) 16:23, 23 March 2024 (UTC)

Merge proposal

I propose merging the disambiguation pages Pag and Pags into PAG. Pags itself now has 2 entries, and in normal circumstances a plural should redirect to a singular. Pag should be merged into PAG, a normal course of action for disambiguation pages, see WP:DABCOMBINE: "A single disambiguation page may be used to disambiguate a number of similar terms. Sets of terms which are commonly so combined include ... Terms that differ only in capitalization". The problem with not doing so is ambiguous uses: abbreviation "PAG" may be entered as "pag"; language code "pag" may be entered as "PAG"; where do you put "PAg"? I tried doing this by WP:BOLD action, but was reverted by @Joy:: the result was here [1]. See also Joy's useful analysis in the section immediately above. Shhhnotsoloud (talk) 16:22, 23 March 2024 (UTC)

Yeah, we know from many examples that people often type in (all-uppercase) acronyms in all-lowercase. At the same time, that's a pretty obvious distinction for readers, i.e. they are usually aware that they were just being lazy and don't have particular trouble identifying that. IOW it seems that it's more like "Right, I want the all-uppercase variant of this, I'll just click that" and less like "Wait, where am I, why are they not showing me what I wanted?!".

I've never been keen on applying WP:DABCOMBINE in cases where we don't see a clear singular/plural or see distinct patterns of reader traffic - making multiple groups of readers wade through combined and longer navigation lists seems objectively worse than making them click through shorter navigation lists.

As some may know already, I've questioned a number of our disambiguation-related guidelines recently at WT:D because they often seem to be based on various opinions of editors, not hard data on what is useful and matching actual reader behavior (which goes against the spirit of WP:RF, and in turn the spirit of some of our rigorous rules like WP:V, WP:OR, ...).

Of particular interest here, perhaps, have been discussions about how even traffic patterns at some well-known plurals like echoes and rumours just don't match those at the respective singulars.

So I wouldn't like us to assume that the average English reader identifies pags so strongly with being the plural of the word pag (and in turn strongly identifying that with just another way of writing PAG) that they'd be anywhere close to astonished - unless of course we can find some evidence for that. --Joy (talk) 18:08, 23 March 2024 (UTC)

Glossing over PAg (should we redirect that to Professional agrologist?) the meanings of Pag are distinct from the meanings of PAG, so there is a case for keeping them on separate pages. However, WP:DABCOMBINE gives the guideline that Terms that differ only in capitalization ... should almost always share a disambiguation page. The resulting page is not too long and I see no special feature that would recommmend an IAR case. It may be worth discussing a change to WP:DABCOMBINE more widely but, if we follow its current advice, the two dabs should be merged.

Although most English plurals are formed by adding s, pags is not a plural of pag because none of the meanings listed on Pag is a countable noun. (Someone who visited Croatia and Iran might see two Pags, but that's a stretch and the combination is not an encyclopedic topic.) That suggests that we should keep Pags separate, keeping the See also entries between the two pages. Certes (talk) 08:35, 3 April 2024 (UTC)

@Certes WP:DABCOMBINE is a guideline, not a policy. Using it here despite your better judgement would be a travesty of WP:NOTBURO, which is in fact a policy. --Joy (talk) 19:21, 3 April 2024 (UTC)

Looking into the history of that guideline wording, I found Wikipedia talk:WikiProject Disambiguation/Archive 42#Invariably? where we see how this sort of a thing came about from a discussion between a handful of people. This is yet another matter where we just have a few editors arguing based on assertions, instead of any sort of a data-based examination of what sort of navigation actually fits most readers best.

Speaking of which, I should correct what I wrote above - [WikiNav] indicates a substantial portion of readers want the acronym, but a larger contingent actually want a non-acronym, something called "Pag" - we see that within the 138 views in the month of February '24, our current clickstream system could identify 28 clickstreams towards the latter, and 12 towards the former. These numbers are so close to the anonymization thresholds that trying to compare them is probably useless (as it could well lead to grave errors). Furthermore, even if we take the numbers at face value, it's just ~20% and ~8% of readers whose navigation behavior was identifiable - so >71% of readers that month are unaccounted for, which means we don't know much at all.

So I'd say we should experiment with both ways of doing things and see if we can measure any improvements in navigation. --Joy (talk) 19:58, 3 April 2024 (UTC)

As I suggested above, I don't think this is an exceptional case, but I do think there's merit in reconsidering DABCOMBINE in general before applying a possibly changed version of it to Pag. Certes (talk) 08:42, 4 April 2024 (UTC)

BTW in the meantime I noticed another exception to this guideline, TOP and Top. Then I went looking for more, and found a huge number of cases where the lowercase and uppercase differed, but also found these interesting cases in the three-letter space:

• Wim (set index, but still navigation) and WIM
• Val and VAL
• Vad and VAD
• Use and USE
• Ulf (set index, but still navigation) and ULF
• Uel and UEL
• Udi and UDI
• Tuv and TUV
• Tut and TUT
• Tso and TSO
• Ted (set index, but still navigation) and TED
• Swe and SWE
• Soe and SOE
• Soa and SOA
• Sme did not even have a hatnote to SME
• Sel and SEL
• Sde (set index, but still navigation) and SDE
• Rut and RUT
• Roa and ROA
• Rnd and RND
• Ris and RIS
• Rie (set index, but still navigation) and RIE
• Ran and RAN
• Rae and RAE
• Rab and RAB
• Pup and PUP
• Puk and PUK
• Pla and PLA

That's where I stopped arbitrarily, because I was tired and it seemed like already a sufficient sample to illustrate this issue.

Now, some of these may be warranted, some may not be, but my point is that there's a lot of variety, and I don't think we should consider this guideline as operationally consensual as it sounds, for lack of a better description. --Joy (talk) 21:14, 21 April 2024 (UTC)