Jump to content

Product Analytics/Data products/fawiki metrics summary

From mediawiki.org

Executive summary

[edit]

The Persian (Farsi) Wikipedia blocked IP editing on article pages between October 20, 2021 and April 20, 2022 by community decision. This report was commissioned by the product leads to investigate the impact of the restriction on the project's health. The impacts on each metric are summarized in the below table after comparing the 6 months of restriction to the prior months in the same year, the same months in the prior 3 years, and other similar wiki projects. As Portuguese Wikipedia has enabled a wider restriction of IP editing on all pages, this report also looked into the similarities and differences of the impacts on the two projects.

Similar to Portuguese Wikipedia, the restriction on Persian Wikipedia prevented the edits that would most likely be reverted and reduced administration page protections. Unlike Portuguese Wikipedia, the partial restriction on Persian Wikipedia did not drive new account creation or active logged-in editors, nor did it reduce the administration blocks. The impact on non-reverted content edits (net non-reverted non-bot content edits) is negative but not significant.

In summary, the restriction on article pages effectively reduced vandalism on Persian Wikipedia. However, the anonymous editors were not converted to registered editors after the restriction on Persian/Farsi Wikipedia. It's recommended to re-visit the trends 6 months after the restriction was lifted, to assess whether the short-term impact is persistent or reverted.

Category Metrics Persian (Farsi) Wikipedia Portuguese Wikipedia
Editors Number of new accounts No change Improved
Number of active logged-in editors No change Improved
Retention rate No change No change
Content Number of reverts Improved (number decreased) Improved (number decreased)
Number of net non-reverted edits Declined but not significant Declined but not significant
Administration Number of blocks Improved (number decreased) Improved (number decreased)
Number of protected pages Improved (number decreased) Improved (number decreased)
Number of checkuser checks No change No change

* No change means no change is confirmed to be caused by restriction of IP editing

Introduction

[edit]

The Persian/Farsi Wikipedia (fawiki) community made the decision to temporarily block anonymous editing on article (content) pages for 6 months (see consensus and ticket). Blocking was activated on October 20, 2021, Wednesday in the 42nd week of 2021, and deactivated 6 months later, on April 21, 2022, Thursday in the 16th week of 2022.[1] Per the product leads' request, the Product Analytics team provided quantitative analysis to assess the impact on the project's health. This report summarizes the trends of the metrics regarding editor, content creation, and administration based on the data by April 2022, and compares those metrics to the prior months, prior years, and other wiki projects.

Methods and considerations

[edit]

Metrics selection and definition

[edit]

The analysis focuses on the project metrics that are likely impacted by the restriction. Although the restriction only applied to articles, impacts on article talk pages were also assessed for possible side effects. Therefore, the page-based events, such as edit, revert, and protection, were evaluated on the article content pages (namespace=0) and on the article talk pages (namespace=1) separately, in addition to the overall trend on all namespaces.[2]

Baseline selection

[edit]

Three types of baselines are used in comparison analysis: (1) prior months in the same year, (2) same months in the prior 3 years, and (3) other, similar wikis. This 6-month restriction was not conducted as a controlled experiment, as many variables might disturb the normal patterns and cause changes in the metrics. For example, the pandemic is one of the biggest factors that has changed user patterns in the past two years, and its effects vary from country to country. As such, the comparison baselines should be used as a reference only instead of an assessment criterion.

In the month-over-month comparison, the 6 months in restriction was compared to the period of 6 months prior to the restriction (6Mo ratio). A year of 365/366 days is grouped into the first half of the year, starting from April 21 each year, and the second half of the year, starting from October 20 each year. In this division, the second half of the year usually has 1 or 2 days more than the first half of the year. Therefore, the ratio of 6 months over the prior 6 months in the scenario of no impact and no seasonality is 1.005 ~1.01. To consider the seasonality, the 6Mo ratios of prior years are also referred to.

In the year-over-year comparison, the 6 months in restriction is compared to the same 6-month period of the prior 3 years. 3 years prior (18/19) is pre-pandemic. 2 years prior (19/20) is the year partially in the pandemic. 1 year prior (20/21) is the year fully in the pandemic.

Comparison time periods
Year 1st half of year (1HY) 2nd half of year (2HY) Abbreviation
Current year (21/22) 2021-4-21 ~ 2021-10-19 (182 days) 2021-10-20 ~ 2022-04-20 (183 days) * Yr0
1 year prior (20/21) 2020-4-21 ~ 2020-10-19 (182 days) 2020-10-20 ~ 2021-04-20 (183 days) Yr-1
2 years prior (19/20) 2019-4-21 ~ 2019-10-19 (182 days) 2019-10-20 ~ 2020-04-20 (184 days) Yr-2
3 years prior (18/19) 2018-4-21 ~ 2018-10-19 (182 days) 2018-10-20 ~ 2019-04-20 (183 days) Yr-3

* When IP editing was blocked on the article (content) page (namespace=0).

Based on the data from October 2019 to September 2021, 80% of edits on fawiki were submitted from Iran. In the comparison of wiki projects, six wikis were selected from the top ten wikis which were mostly edited in Iran. The six wiki projects are: Arabic Wikipedia (arwiki), Azerbaijani Wikipedia (azwiki), Central Kurdish Wikipedia (ckbwiki), Hebrew Wikipedia (hewiki), Persian Wikisource (fawikisource), Persian Wiktionary (fawiktionary). The general trend of the six wikis was assessed on whether the changes on fawiki were within a normal range. Because no two wikis are the same, the trend on the six wikis is a reference rather than a criterion. It's observed that the Year-over-Year (YoY) or 6Month-over-6Month (6Mo6M) ratios on smaller wikis, such as Central Kurdish Wikipedia, Persian Wikisource, and Persian Wiktionary tend to be more volatile than other medium size wikis. Also, the administration metrics tend to be more volatile than editors and content metrics, likely because administration actions were conducted by a small group of people for irregular needs.

Four other wikis were not selected due to project-specific nuances. English Wikipedia (enwiki) was not selected because it is the biggest wiki project and could be impacted by many other global factors. South Azerbaijani Wikipedia (azbwiki) was excluded due to a poor correlation with fawiki in historical data. Turkish Wikipedia (trwiki) was not considered because trwiki had a level shift event during the comparison period – on 15 January 2020, the block of trwiki along with other language editions of Wikipedia in Turkey was lifted resulting in a metrics shift.[3] Kurdish Wikipedia (kuwiki) was excluded because its metrics have high volatility with an unexplained spike of non-bot edits in the week of 2021-07-28, which is within the comparison period.

Impacts on editors

[edit]

Number of new accounts

[edit]

The number of new accounts is defined as the number of users who register in a given period of time. The hypothesis is that the new accounts would increase when the restriction is enabled. The hypothesis is not supported.

The comparisons to prior months, prior years, and other wikis do not show an increase in new accounts on fawiki during the restriction period. Instead, Figure 1.1. Weekly Trend and Figure 1.3. YoY Comparison shows that fawiki has 18% fewer new accounts in the current year (21/22) compared to 1 year prior (20/21), dropping from 55686 to 45912. This decrease is not unique on fawiki – four other wikis have fewer new accounts in the current year (21/22).

Figure 1.2. 6Mo Comparison shows that the new accounts on fawiki in the second half of the year (2HY) is about the same as the first half of the year (1HY) in the current year (21/22) while 5 out of 6 other wikis have more new accounts in second half of year than first half of year in the current year (21/22). It means that turning off IP editing on article pages did not lead to a significant increase in new accounts. It is different from the trend on Portuguese Wikipedia that new accounts surged after they blocked IP editing on the entire site.[4] The different trends on the two wikis might be due to the different blocking policies or user preference.

Comparison of second half of the year vs first half of the year
New Accounts Persian Wikipedia Other Wiki Projects 6Mo6M Ratio
year 1st half of year 2nd half of year 6Mo6M Ratio Arabic Wikipedia Azerbaijani Wikipedia Central Kurdish Wikipedia Hebrew Wikipedia Persian Wikisource Persian Wiktionary
Current year (21/22) 45790 45912 * 1 1.21 1.12 1.17 0.81 1.03 1.04
1 year prior (20/21) 51448 55686 1.08 1.24 1.3 1.12 2.07 1.09 1.2
2 years prior (19/20) 45107 50795 1.13 1.29 1.39 1.58 1.17 1.15 1.02
3 years prior (18/19) 42160 47189 1.12 1 1.18 1.14 1.02 1.5 1.43

* Restriction period

Figure 1.1. New Accounts Weekly Trend -- The metric decreased
Figure 1.2. New Accounts 6M Comparison -- All wikis usually have more new accounts in 2HY than 1HY. Fawiki does not have more new accounts in Yr-2 2HY.
Figure 1.3. New Accounts YoY Comparison -- Fawiki has fewer new accounts in yr0 compared to Yr-1. Four other wikis have a similar decrease.

Number of active logged-in editors

[edit]

Number of active user editors is defined as the number of logged-in users who made at least one content edit on fawiki in the given period of time. The hypothesis is that the active user editors would increase when the restriction is enabled. The hypothesis is not supported by data.

The comparisons to prior months, prior years, and other wikis do not show an increase in active user editors on fawiki during the restriction period. Figure 2.1. Weekly Trend shows a decline on fawiki in March and April. Figure 2.2. 6Mo Comparison shows that the active user editors on fawiki is 5% less in the second half of the current year (21/22) than the first half of year, lowest 6-month ratio in the past 4 years. 4 other wikis show a similar trend that the current year has the lowest 6M ratio. Figure 2.3. YoY Comparison shows that fawiki decreased compared to 1 year prior (20/21), but increased compared to 3 years prior (18/19). Five other wikis show a similar YoY trend. It means that turning off IP editing on article pages did not lead to a significant increase in active user editors on fawiki. It's different from Portuguese Wikipedia where active user editors increased after the restriction was enabled on the entire site.[5] The different trends on two wikis might be due to the different blocking policies or user preference.

Comparison of second half of the year vs first half of the year
Active Logged-in Editors Persian Wikipedia Other Wiki Projects 6Mo6M Ratio
year 1st half of year 2nd half of year 6Mo6M Ratio Arabic Wikipedia Azerbaijani Wikipedia Central Kurdish Wikipedia Hebrew Wikipedia Persian Wikisource Persian Wiktionary
Current year (21/22) 83310 79525 * 0.95 0.95 1.08 0.88 1 0.99 1.23
1 year prior (20/21) 78538 89573 1.14 1.01 1.31 1.18 1.03 1.71 1.12
2 years prior (19/20) 66827 69569 1.04 1.13 1.23 1.3 1.01 0.95 1.33
3 years prior (18/19) 54926 65154 1.19 0.98 1.12 1 1.07 0.44 1.09

* Restriction period

Figure 2.1. Active Logged-in Editors Weekly Trend -- The metric decreased.
Figure 2.2. Active Logged-in Editors 6M Comparison -- The metric on fawiki decreased in Yr0 2HY. Four other wikis show a similar trend.
Figure 2.3. Active Logged-in Editors YoY Comparison -- The metric on fawiki decreased compared to Yr-1, increased compared to Yr-3. Five other wikis show a similar trend.

Retention rate

[edit]

Retention rate is defined as: out of the (non-bot) users who registered in the week and made at least one edit in their first 30 days, the proportion who also edited during their second 30 days.[6]

Due to the nature of the definition, this metric surfaces two months after user account creation. It's a metric for observing whether user retention rate will change when policy is changed. So far no obvious change in retention rate is observed (see Figure 3.1.), similar to the trend on ptwiki.

Figure 3.1. Retention Rate Weekly Trend -- No significant change

Takeaways

[edit]

New accounts and active user editors did not increase on fawiki during the restriction. The trends on fawiki are different from those on ptwiki, which suggests that blocking IP editing only on article pages did not encourage anonymous editors to create accounts and edit in logged-in mode.

Impacts on content

[edit]

Number of edits

[edit]

Number of edits is defined as the total number of edits made on fawiki during the given period of time. Edits that were published by bots, edits that have been reverted, and edits that have been deleted are included among total edits. This metric is for curiosity observation because it is one of the Foundation's core product metrics. To clearly understand the impact of the restriction, the number of reverts and the number of net edits excluding bot edits, reverted edits and edits reverting other edits will be discussed in session Number of reverts and session Net non-revered edits.

Figure 4.1. Weekly Trend shows that the total edits have decreased since the restriction was enabled. Figure 4.2. 6Mo Comparison shows that total edits declined 29% during the restriction. The decrease is from content pages. The number of edits on the content page is 31% less compared to prior 6 months, contributed by the decrease in reverted edits and decrease in net non-bot non-reverted edits. The number of edits on talk pages is similar to prior 6 months.

Comparison of second half of the year vs first half of the year
Edits Persian Wikipedia Other Wiki Projects 6Mo6M Ratio
year 1st half of year 2nd half of year 6Mo6M Ratio Arabic Wikipedia Azerbaijani Wikipedia Central Kurdish Wikipedia Hebrew Wikipedia Persian Wikisource Persian Wiktionary
Current year (21/22) 1575165 1120883 * 0.71 1.03 0.84 1.13 1.12 1.05 1.36
1 year prior (20/21) 1358544 1659739 1.22 0.5 0.73 1.06 0.92 1.96 0.22
2 years prior (19/20) 1480958 1407900 0.95 1.5 0.99 1.05 1.09 0.75 1.57
3 years prior (18/19) 1624038 1256574 0.77 1.13 1.4 1.05 0.96 0.47 1.06

* Restriction period

Figure 4.1. Total Edits Weekly Trend -- The metric decreased.
Figure 4.2. Total Edits 6M Comparison -- The metric on fawiki decreased in Yr0 2HY. The metric is volatile across wikis due to bot edits.
Figure 4.3. Total Edits YoY Comparison -- The metric on fawiki decreased compare to Yr-1, Yr-2 and Yr-3. The metric is volatile across wikis due to bot edits.

Number of reverts

[edit]

Number of reverts is defined as the number of edits reverted within 48 hours of being published in the given period of time. Fewer reverts is considered an improvement. The hypothesis is that the reverts on content pages would be reduced with the restriction. The hypothesis is supported by the data.

The comparisons to prior months, prior years , and other wikis show a significant reduction in content reverts on fawiki during the restriction period. Figure 6.1. Weekly Trend shows that the number of content reverts decreased when the restriction was enabled and increased when the restriction was disabled. Figure 6.2. 6Mo Comparison shows the number of content reverts is 70% less than the prior 6 months. Given in the previous 3 years, the second half of the year always has more reverts than the first half of year, it's a significant decline. Figure 6.3. YoY Comparison shows content reverts decreased 72% compared to 1 year prior (20/21), 59% compared to 2 years prior (19/20), 45% compared to 3 years prior (18/19). Other wikis do not show such a decline. This trend only shows on fawiki.

It means restriction prevents creating content which most likely will be reverted.

Reverts on the talk page only account for 1% of the total reverts. Figure 7.1 Weekly Trend shows that the number of talk reverts does not have a significant change. As usual, the number of talk reverts is slightly higher in the second half of the year than in the first half of the year. (see Figure 7.2.)

Comparison of second half of the year vs first half of the year
Reverted Edits Persian Wikipedia Other Wiki Projects 6Mo6M Ratio
year 1st half of year 2nd half of year 6Mo6M

Ratio

Arabic Wikipedia Azerbaijani Wikipedia Central Kurdish Wikipedia Hebrew Wikipedia Persian Wikisource Persian Wiktionary
Current year (21/22) 186841 60479 * 0.32 0.89 1.32 0.8 1.33 0.68 0.94
1 year prior (20/21) 138073 202461 1.47 0.96 1.91 1.34 1.12 1.51 1.7
2 years prior (19/20) 115745 143937 1.24 1.19 1.04 1.11 1.08 1.25 0.96
3 years prior (18/19) 115351 106696 0.92 1.17 1.4 1.38 1.25 0.86 0.49

* Restriction period

Reverted Edits Within 48 Hrs Reverted Content Edits Within 48 Hrs Reverted Talk Edits Within 48 Hrs
Figure 5.1. Weekly Trend -- The overall metric decreased when the restriction was enabled and increased when the restriction was disabled.
Figure 6.1. Weekly Trend -- The content metric decreased when the restriction was enabled and increased when the restriction was disabled.
Figure 7.1. Weekly Trend -- The talk metric has no significant change.
Figure 5.2. 6Mo Comparison -- The overall metric on fawiki decreased in Yr0. It is significant given the historical trends on fawiki and other 6 wikis.
Figure 6.2. 6Mo Comparison -- The content metric decreased on fawiki in Yr0. It is significant given the historical trends on fawiki and other 6 wikis.
Figure 7.2. 6Mo Comparison -- The talk metric on fawiki has no significant change in Yr0.
Figure 5.3. YoY Comparison -- The overall metric on fawiki is lower in Yr0 than Yr-1, Yr-2 and Yr-3. No other wikis had a similar trend.
Figure 6.3. YoY Comparison -- The content metric on fawiki is lower in Yr0 than Yr-1, Yr-2 and Yr-3. No other wikis had a similar trend.
Figure 7.3. YoY Comparison -- The talk metric on fawiki is lower in Yr0 than Yr-1, still higher than Yr-2 and Yr-3.

Net non-reverted edits

[edit]

Net non-reverted edits are defined as the number of edits that were not reverted within 48 hours, excluding bot edits, reverted edits, and edits reverting other edits. They are considered 'good' edits by users.

The first hypothesis is that the net non-reverted edits on all namespaces would decline. It's inconclusive. Figure 8.1. Weekly Trend shows that the net non-reverted edits dropped since October 20, 2021. There was a spike in the week of 2022-03-21, 18221 edits created by one editor within a week (mostly in 2 days). Even though the user is not identified as a bot, their behavior is similar to a bot. So the number of net edits in the second half of the current year (21/22) was adjusted by excluding the 18221 edits. Figure 8.2. 6Mo Comparison shows that after the adjustment net non-reverted edits on fawiki is 17% less in the second half of the year than in the first half of the year of the current year (21/22), dropping from 917617 in the first half of year to 759026 in the second half of the year. It is a common trend on 4 other wikis in the current year (21/22) that the second half of the year has fewer net edits. Figure 8.3. YoY comparison shows that the current year (21/22) is 18% lower than the 1 year prior (20/21), but still 6% higher than 2 years prior (19/20) and 13% higher than 3 years prior (18/19). In summary, though a drop right after the restriction was confirmed, it's inconclusive that the overall net edits decreased during the six months of restriction given other wikis had declined during the same period of time and that the current edits level is higher than 2 years prior (19/20) and 3 years prior (18/19).

The second hypothesis is that the net non-reverted edits on content pages would decline. It is inconclusive. Net edits on content pages contributed 70% of total net edits. The trend on content pages is similar to the trend of net edit on all namespaces. (see Figure 9.1) The net content edits decreased once the restriction was enabled in Oct 2022. The net non-reverted edits on content pages is 24% fewer in the second half of the current year (21/22) than in the first half of the year, dropping from 657228 to 498571. (see Figure 9.2) However, it's inconclusive that the overall net content edits decreased during the restriction given that other wikis had declined during the same period of time and that the current edits level is higher than 2 years prior (19/20) and 3 years prior (18/19). (see Figure 9.3) If assume the organic 6-month ratio in the current year (21/22) on fawiki without restriction is 0.925, the median of the other 6 wikis, it is estimated that fawiki lost additional 109,365 content edits during the restriction, i.e. 17% of total net content edits in the first half of year of the current year (21/22).

The third hypothesis is that the net non-reverted edits on the talk page would increase. It is not supported. Figure 10.1. Weekly Trend shows the level of net talk edits is close to the previous level. Figure 10.2. 6Mo Comparison shows the net talk edits in the second half of the year is 14% higher than the first half of the year in the current year (21/22). It is still within the historical range. Figure 10.3. YoY Comparison shows the net talk edits in the current year (21/22) is 13% less than 1 year prior (20/21) and 1% less than 2 years prior (19/20), 6% more than 3 years prior (18/19). In summary, net talk edits did not increase significantly.

Comparison of second half of the year vs first half of the year
Net non-reverted edits Persian Wikipedia Other Wiki Projects 6Mo6M Ratio
year 1st half of year 2nd half of year 6Mo6M

Ratio

Arabic Wikipedia Azerbaijani Wikipedia Central Kurdish Wikipedia Hebrew Wikipedia Persian Wikisource Persian Wiktionary
Current year (21/22) 917617 759026 *(adjusted) 0.83 0.91 0.9 0.64 0.97 1.21 1.4
1 year prior (20/21) 818243 921489 1.13 0.94 1.11 1.4 0.94 1.93 0.93
2 years prior (19/20) 684574 734264 1.07 0.83 1.16 1.17 1.03 0.74 1.63
3 years prior (18/19) 656303 688638 1.05 0.95 1.33 1.01 1.1 0.45 1.1

* Restriction period

Comparison of second half of the year vs first half of the year
Net non-reverted content edits Persian Wikipedia Other Wiki Projects 6Mo6M Ratio
year 1st half of year 2nd half of year 6Mo6M Ratio Arabic Wikipedia Azerbaijani Wikipedia Central Kurdish Wikipedia Hebrew Wikipedia Persian Wikisource Persian Wiktionary
Current year (21/22) 657228 498571 * (adjusted) 0.76 0.89 0.92 0.63 0.93 3.75 1.65
1 year prior (20/21) 553994 611259 1.1 0.94 1.15 1.53 0.95 0.8 0.89
2 years prior (19/20) 452333 497591 1.1 0.85 1.28 1.26 1.04 0.95 2.28
3 years prior (18/19) 419097 467074 1.11 0.93 1.39 0.93 1.07 1.12 0.89

* Restriction period

Net Edits

Excluding Bot Edits and Reverts

Net Content Edits

Excluding Bot Edits and Reverts

Net Talk Edits

Excluding Bot Edits and Reverts

Figure 8.1. Weekly Trend -- The metric decreased since October 20,2021.
Figure 9.1. Weekly Trend -- The content metric decreased since October 20,2021.
Figure 10.1. Weekly Trend -- The talk metric has no significant change.
Figure 8.2. 6Mo Comparison -- The overall metric on fawiki is lower in 2HY than the 1HY in Yr0. It is a common trend on 4 other wikis in Yr0 that 2HY has less net edits than 1HY.
Figure 9.2. 6Mo Comparison -- The content metric on fawiki is lower in 2HY than the 1HY in Yr0. It is a common trend in 4 other wikis that 2HY in Yr0 has less net edits.
Figure 10.2. 6Mo Comparison -- The talk metric on fawiki in 2HY of Yr0 is higher than the 1HY. It is still within the historical range.
Figure 8.3. YoY Comparison -- The overall metric on fawiki in Yr0 is 18% lower than the Yr-1, but still 6% higher than Yr-2 and 13% higher Yr-3.
Figure 9.3. YoY Comparison -- The content metric on fawiki in Yr0 is 18% lower than the Yr-1, but still 0.2% higher than Yr-2 and 7% higher Yr-3.
Figure 10.3. YoY Comparison -- The talk metric on fawiki in Yr0 is 13% lower than Yr-1 and 1% less than Yr-2, 6% more than Yr-3.

Takeaways

[edit]

Restriction prevents content creation which most likely will be reverted. Restriction might prevent creating good content at some level. But the effect is not significant enough to be confirmed.


Impacts on administration

[edit]

Number of blocks

[edit]

Number of blocks is defined as the number of block and reblock actions on accounts or IPs or IP ranges in a given period of time. Fewer blocks is considered an improvement. The hypothesis is that blocks would be reduced when restriction is enabled. This hypothesis is not supported by the data.

Figure 11.1. Weekly Trend shows that the number of blocks reduced in October and November, 2021, but spiked twice in February 2022 and March 2022. It's different from ptwiki which showed a significant decrease in blocks.[7] The blocks were added mainly on 2 days by a bot admin, with comment {پروکسی باز} (en: "open proxy"). It is confirmed with a community member that the bot was copying over blocks made on other projects. Excluding blocks by bot admins, the number of blocks is 57% fewer than 6 months prior and 54% fewer than 1 year prior.

Figure 11.1. Blocks Weekly Trend -- The metric decreased in October and November, 2021, but spiked twice in February 2022 and March 2022.
Blocks Persian Wikipedia
year Apr 21 - Oct 19 Oct 20 - Apr 20 6Mo6M Ratio
Current year (21/22) 12774 66987 * 5.24
1 year prior (20/21) 6953 11433 1.64
2 years prior (19/20) 8629 5741 0.67
3 years prior (18/19) 3555 15257 4.29
Blocks excluding bot admins Persian Wikipedia
year Apr 21 - Oct 19 Oct 20 - Apr 20 6Mo6M Ratio
Current year (21/22) 9504 4125 * 0.43
1 year prior (20/21) 4995 8968 1.8
2 years prior (19/20) 7623 4719 0.62
3 years prior (18/19) 3555 15257 4.29

* Restriction period

Protected pages

[edit]

Number of protected pages is defined as the number of unique pages protected in a given period of time. Having fewer protected pages is considered an improvement. The hypothesis is that the number of protected pages would reduce during restriction. This hypothesis is supported.

Figure 12.1 Weekly Trend shows that the metric spiked in October 2021. 5867 pages were protected by one user. Per a community member's request, those protections were excluded from the total of protected pages in the first half of the current year (21/22). After the adjustment, compared to the prior 6 months, the number of protected pages decreased 32%. The reduction in protected pages is from content pages, as the protections on talk pages only account for 1% of the total.

Figure 12.1. Protected Pages Weekly Trend -- the metric spiked in October 2021.
Comparison of second half of the year vs first half of the year
Protected pages Persian Wikipedia Other Wiki Projects 6Mo6M Ratio
year 1st half of year 2nd half of year 6Mo6M Ratio Arabic Wikipedia Azerbaijani Wikipedia Central Kurdish Wikipedia Hebrew Wikipedia Persian Wikisource Persian Wiktionary
Current year (21/22) 3487 (adjusted) 2372 * 0.68 0.95 0.97 1.09 1 - 1.5
1 year prior (20/21) 2874 2964 1.03 0.93 2.8 0.26 1.11 0 0.78
2 years prior (19/20) 3542 2739 0.77 1.42 0.72 0.41 1.37 0 6.5
3 years prior (18/19) 1452 1828 1.26 0.59 1.47 0.92 0.95 0 2

* Restriction period

Checkuser checks

[edit]

Number of Checkuser checks is defined as the number of checkuser requests in a given period of time. The metric is for curiosity observation. Figure 13.1 Weekly Trend shows the number of checkuser checks has high volatility. It varied within its volatility range without an obvious trend.

Figure 13.1. Checkuser Requests Weekly Trend -- The metric has no significant change.


Takeaways

[edit]

The restriction does not impact administrative actions significantly. Restriction on content pages reduced page protections but did not show an impact on blocks and checkuser requests.

Takeaways

[edit]
  • The restriction on Persian Wikipedia article pages prevented content creation which would likely be reverted. But the non-reverted edits slightly declined.
  • The restriction on Persian Wikipedia article pages did not drive the growth of new editors and active editors.
  • The restriction on Persian Wikipedia article pages did not reduce administration actions such as blocking and checkuser requests but reduced the number of page protections.
  • Different restriction policies or user preferences lead to different trends. Portuguese Wikipedia blocked IP editing on the entire site while Persian Wikipedia only blocked article editing. New editor accounts and active logged-in editors improved on Portuguese Wikipedia while they have not changed on Persian Wikipedia.

References

[edit]