Hi! I think the discrepancy is because Wikimedia Stats counts all views by
default, whereas Siteviews only counts User views (excluding Bots and
Spiders) by default. Once you exclude Bots and Spiders from Wikimedia
Stats, or include them in Siteviews, the figures match.
On Thu, Feb 15, 2024 at 9:30 AM F. Xavier Dengra i Grau via Wikimedia-l <
wikimedia-l(a)lists.wikimedia.org> wrote:
Bon dia / Hi,
I would like to ask around about the fact that we have very significant
imbalances of official statistics when it comes to the traffic in our wiki
projects.
Let's look at the reading gross value for the English Wikipedia on January
2024. Without applying any kind of filtering to their searching method, the
website Wikimedia Stats shows 12.697.373.117 visits, while the Pageviews
tool from the WikimediaCloud in Toolforge yields 8.864.755.474 visits.
This is a huge figure disparity, and both data repositories are hosted
"under the same roog" and most likely are the two more widely used tools
for this purpose. Am I wrong? For a smaller project, like the Catalan
Wikipedia, there is a 19.5M vs 64.5M inconsistency... Which changes a lot
our conclusions in a tech situation in which we are socially dealing with
Google hiding results in our language from its search engine since a couple
of years ago. What am I missing about these value differences? As an
experienced editor, I have been regularly digging in our available data
tools for several years. But it's difficult not to encounter frequent
problems of comparability, false positives, and reliability. Not only for
my personal pleasure, but when it comes to easily explain our projects'
data to the outside world via referring to a trustworthy portal.
That adds up to the fact that we are not able to see itemised statistics
for some small countries in Wikimedia Stats. We can filter how many
millions of visits does the Dutch Wikipedia get in Belgium or in the
Netherlands, but we cannot see which is the language use of each Wikipedia
in Belgium (% of readers that accesses it in French, Dutch, German,
Walloon, Picard, English, etc). That feature disappeared in 2018 with the
last update of the dismantled WiViVi portal. Altogether, it makes it
impossible to tackle biases or plan actions by chapters, user groups or
even academic policies regarding awareness or revitalization of minoritized
and endangered languages.
I am afraid that, this being my experience as a long-term editor, the ones
of newcomers, journalists, or even scientists may still be much more
uncertain and confusing. Hopefully someone can help to figure out some of
these questions.
Salutacions / Best regards,
Xavier Dengra
_______________________________________________
Wikimedia-l mailing list -- wikimedia-l(a)lists.wikimedia.org, guidelines
at:
https://meta.wikimedia.org/wiki/Mailing_lists/Guidelines and
https://meta.wikimedia.org/wiki/Wikimedia-l
Public archives at
https://lists.wikimedia.org/hyperkitty/list/wikimedia-l@lists.wikimedia.org…
To unsubscribe send an email to wikimedia-l-leave(a)lists.wikimedia.org