Wikipedia:Move review/Log/2014 July

2014 July

Yesterday (Beatles song)

Yesterday (Beatles song) (talk|edit|history|logs|links|archive|watch) (RM)

In my judgment, the closer of this requested move, although acting in good faith, did not follow the spirit and intent of WP:RMCI. It appears to me that the closer failed to follow these important Wikipedia guidelines:

Consensus is determined not just by considering the preferences of the participants in a given discussion, but also by evaluating their arguments, assigning due weight accordingly, and giving due consideration to . . . applicable policy, guidelines and naming conventions.

— WP:Requested moves/Closing instructions#Determining consensus

The closer is there to judge the consensus of the community, after discarding irrelevant arguments: those that flatly contradict established policy, those based on personal opinion only, those that are logically fallacious, those that show no understanding of the matter of issue. . . . [The closer] is expected to know policy sufficiently to know what arguments are to be excluded as irrelevant.

— WP:Closing discussions#How to determine the outcome

Most (but not all) of the discussion of this RM revolved around whether the subject article is the WP:PRIMARYTOPIC of the term "Yesterday". Many editors expressed views on this point, but almost all of these were based on arbitrary personal opinion, rather than specific reference to the actual text of the WP:PRIMARYTOPIC policy and supporting empirical evidence. In assessing all views expressed, the closer failed to discount those that were incongruent with Wikipedia policy and/or unsupported by evidence, and instead reached a conclusion based largely on a headcount of votes.

Before opening this move review, I discussed the matter on the closer's User Talk page (here). The closer maintained that the discussion of WP:PRIMARYTOPIC was "editorial" in nature, as opposed to being about how any particular guideline or policy should be applied. I find this claim difficult to understand; how can a discussion of how a WP policy applies to the situation at hand not be about how a WP policy applies to the situation at hand?

In the RM discussion, I presented a carefully-compiled table of pageview statistics covering the immediately preceding 90 days, and including all articles and redirects whose titles were either "Yesterday" by itself, or "Yesterday" followed by a disambiguator. These statistics showed that the article at "Yesterday" (which, at the time of the discussion, was the title of the article about the Beatles song) was by far the most frequently viewed by Wikipedia readers of all the "Yesterday"-related topics, even after accounting for the disambiguation page. This is the very definition of what constitutes a primary topic at Wikipedia:

A topic is primary for a term, with respect to usage, if it is highly likely—much more likely than any other topic, and more likely than all the other topics combined—to be the topic sought when a reader searches for that term.

— WP:PRIMARYTOPIC

However, in the subsequent User Talk discussion, the closer rejected this empirical evidence out of hand, making the rather astonishing claim that "statistics" (by which the closer apparently meant "empirical data") are "rarely . . . unbiased" and therefore of little value. The closer asked, "Where is your control group to show what normal counts are like?" as if the concept of a "control group" was somehow pertinent to the discussion of a move request. The closer then stated that these data "had no [bearing] on my interpretation of the consensus of the discussion", even claiming that the data were so irrelevant that "it wasn't worth mentioning in the close". I find the closer's position baffling, to say the least.

I am concerned that allowing this closing to stand would have broadly negative implications for the application of the WP:PRIMARYTOPIC guideline across Wikipedia.

Lest anyone misjudge, I want to emphasize that I do not object to the outcome of the closing in this case; I believe the article in question is better named with the disambiguator "(Beatles song)" than without it (though for reasons unrelated to WP:PRIMARYTOPIC). Rather, my concern here is with the way the discussion was closed.

In my opinion, the best outcome of this move review would be:

1. to affirm that, although the closer acted in good faith, the discussion was closed improperly;
2. to affirm that the article in question is the WP:PRIMARYTOPIC of the term "Yesterday"; and
3. either of the following:

• to relist the RM so that discussion may proceed regarding the merits of adding the disambiguator "(Beatles song)" to the article title even though it is the primary topic of the term "Yesterday"; or
• to implement the compromise proposal (presented during the original RM discussion) of leaving the subject article at "Yesterday (Beatles song)"; leaving the disambiguation page at "Yesterday (disambiguation)"; and making "Yesterday" a redirect to "Yesterday (Beatles song)", in respect of the fact that the latter is the primary topic of the term "Yesterday".

– – –
Respectfully submitted,

— Jaydiem (talk) 00:24, 4 July 2014 (UTC)

• Endorse close It would have been nice to some mention of the statistics in the close, but I still feel the close is valid. Given that many of the other comments still argued against primary topic even with the page view information I can see the closer as discounting it. Slightly off topic as this drifts towards rearguing, concerning the statistics. First page view statistics is that they are inherently biased in that they are highly influenced by existing links to the pages. Since the song was at the title Yesterday a lot of existing links would have lead directly there. Also ~65% is not really what I would consider primary topic. A primary topic needs to be much more likely the intended target than all the other topics combined, I actually think that those page views stats could be used as an argument that the song is not the WP:PRIMARYTOPIC. PaleAqua (talk) 02:04, 4 July 2014 (UTC)

• Response: Respectfully, I would like to call your attention to the explicit conditions stated in the WP:PRIMARYTOPIC policy. To be primary, a topic must be:

(1) more likely than all the other topics combined, and

(2) much more likely than any other [single] topic.

As I noted in the RM discussion, a topic obtaining even a simple majority of pageviews is, by definition, being viewed "more... than all the other topics combined". The data table documents that the subject article received 65.8% (roughly two-thirds) of the relevant pageviews during the 90-day sampling period—well above a simple majority. As for the second criterion above, the second-most-viewed topic received less than an 8% viewshare. It seems clear that 65% is "much more likely" than 8%. Therefore the WP:PRIMARYTOPIC definition is clearly satisfied. In light of this empirical evidence, to rule that the subject article is not the primary topic is tantamount to declaring that the current definition promulgated in the WP:PRIMARYTOPIC policy is invalid and should be disregarded throughout Wikipedia.

⋮

In the post-closing discussion on the closer's User Talk page, the closer objected that the statistics I presented did not account for the possibility that readers who were seeking a "Yesterday"-related topic other than the Beatles song may have inadvertently visited that article first, then clicked to the disambiguation page, and finally clicked to their intended topic. But this objection does not withstand scrutiny: Even if we stipulate that every single one of the 3,061 views of the disambiguation page shown in the table was an instance of this scenario, and we therefore subtracted that number from the pageviews of the article about the Beatles song, that article's viewshare would be reduced by a mere 1.2 percentage points—a negligible amount that is inconsequential to any conclusions to be drawn from the data.

⋮

Lastly, I must also reiterate that none of the other arguments presented in the RM discussion in regard to the primary-topic question were based on the actual WP:PRIMARYTOPIC policy definition, or on relevant empirical data, and therefore those arguments should have been disregarded by the closer. — Jaydiem (talk) 04:40, 4 July 2014 (UTC)

• Questionable close. Irrespective of the quality of the close, I'd like to have enough time to pass to have some page view data now that the move has occurred; more than enough time should pass for search engines to update. Then decide to reverse the close iff <sic> the stats support a reverse, as I'm willing to bet they will.

The stats shouldn't have been dismissed. They should have been weighted, but were NOT conclusive. We may find that Yesterday (Beatles song) gets far less traffic now.

I think that the poor quality of the DAB page also suggests the song is the primary topic; FCS, what is this nonsense? It's the day immediately before today, NOT the 24 hour day period before the Present. FS!

--{{U|Elvey}} ^(t•c) 03:05, 4 July 2014 (UTC)

• Endorse closure. With due respect to User:Jaydiem for good intentions and basically sound reasoning here, I feel this RM is a bit too legalistic in nature. The results of consensus discussions on Wikipedia are of limited precendential value; hence, Jaydiem's contention that "allowing this closing to stand would have broadly negative implications for the application of the WP:PRIMARYTOPIC guideline across Wikipedia" is misguided. No single close has "broadly negative implications," because closures are generally narrowly limited to the article in question. If the right result is reached, there is usually no reason to fret over the particulars of the reasoning used to reach it.

Additionally, Jaydiem's argument fails to the other important means by which PRIMARYTOPIC status may be declared: "A topic is primary for a term, with respect to long-term significance, if it has substantially greater enduring notability and educational value than any other topic associated with that term." Essentially, proponents of the move argued that Yesterday (time) was the PRIMARYTOPIC on this ground, irrespective of pageview statistics. By discounting statistics, the closer affirmed the basis of this argument. Perhaps the closer was inartful and too harsh in his talk page elaboration; however, the key conclusion -- that, in this case, the consensus discussion decided that the "long-term significance" ground outweighed the "usage" ground -- was consistent with policy and practice. It is common for RM discussions to decide, in any given case, that one basis for PRIMARYTOPIC status is more relevant or germane than the other. The arguments that Jaydiem would have preferred to see discounted -- those he refers to as "personal opinion" -- were in fact mostly expressions of the "long-term significance" type, a basis for PRIMARYTOPIC status that is inherently less empirical (but no less valid) than the usage ground. Xoloz (talk) 15:46, 4 July 2014 (UTC)