Jump to content

User:Ivanvector/RFA statistics

From Wikipedia, the free encyclopedia

This page is a compilation of statistics on past requests for adminship from 2008 to 2016, with an update in 2020. Data was compiled from the lists in Wikipedia:Successful requests for adminship and Wikipedia:Unsuccessful adminship candidacies (Chronological). Requests for adminship that have been deleted are not included in these statistics. 2008 was chosen as the cut-off since it is the oldest period for which data was previously recorded on both successful and unsuccessful requests for adminship.

[edit]

Table 1 compiles statistics on the total number of requests for adminship by year, and the rate of successful candidacies.

Year Total RfAs Total not closed early[1] Successful Unsuccessful
2020 25 22 (88%) 17 (77%) of 22 5 (23%) of 22
2019 31 29 (94%) 22 (76%) of 29 8 (24%) of 29
2018 18 14 (78%) 10 (71%) of 14 4 (29%) of 14
2017 40 30 (75%) 21 (70%) of 30 9 (30%) of 30
2016 35 24 (69%) 16 (67%) of 24 8 (33%) of 24
2015 58 38 (66%) 21 (55%) of 38 17 (45%) of 38
2014 62 36 (58%) 22 (61%) of 36 14 (39%) of 36
2013 74 51 (69%) 34 (67%) of 51 17 (33%) of 51
2012 95 58 (61%) 27 (47%) of 58 31 (53%) of 58
2011 140 79 (56%) 52 (66%) of 79 27 (34%) of 79
2010 231 127 (55%) 75 (59%) of 127 52 (41%) of 127
2009 354 212 (60%) 121 (57%) of 212 91 (43%) of 212
2008 591 339 (57%) 202 (60%) of 339 137 (40%) of 339

2016 analysis

[edit]

The rate of overall candidacies has declined drastically, with almost a twenty-fold drop over the 8 year period, and no slowing in decline after either the 2011 or 2015 reforms. The proportion of candidacies that were not closed early has remained fairly constant over the period with a slight drop in 2016, although it's not clear if the drop was due to more suitable candidates being nominated or simply because of a drop in sample size.

The proportion of successful candidacies has also remained fairly stable, with only two years falling outside of a ten-point range (57-67%). The decline in absolute numbers of promotions is related to the drop in total candidacies.

It appears that the reforms in 2011 and 2015 have failed to attract more candidates to run, as indicated by the continuing rapid decline in candidacies. The reforms have also failed to promote more administrators on a proportional basis, as the proportion of successful candidacies has not increased.

2020 analysis

[edit]

The rate of candidacies has not recovered since the 2015 reforms. However, after a slow start in 2016, the proportion of discussions that ran to completion (weren't closed or withdrawn without serious consideration) has increased steadily, reaching a high of 94% in 2019. This is a sign that the 2015 reforms have discouraged unsuitable candidates from seeking adminship, though the slow progression could simply indicate a gradual cultural shift.

The proportion of successful candidacies has likewise increased, but this is due to the reduction in unsuccessful candidacies. The absolute numbers of promotions has remained flat now for nearly a decade.

Table 2: RFA participation

[edit]

Table 2 shows trends in RfA participation. This data excludes candidacies closed early,[1] as well as candidacies withdrawn by the candidate after the full 7 days without waiting for an official assessment, as these candidates gave reasons ranging from not expecting to pass a bureaucrat discussion to being offended by the process and retiring from Wikipedia.

Year Mean RfA
participants
Mean participants
in successful RfAs
Mean participants
in unsuccessful RfAs
2020 184 187 135
2019 214 219 176
2018 237 230 267
2017 197 198 167
2016 199 191 239
2015 150 147 160
2014 111 115 100
2013 120 123 106
2012 119 119 119
2011 110 113 95
2010 115 116 110
2009 99 99 100
2008 87 86 90

2016 analysis

[edit]

An issue identified during the 2015 RFA reform discussion was that the same names tend to appear as voters at RFA, and a stated objective became to attract more voting participation from users who had not previously participated or who could provide new opinions to the process. The numbers of participants remained flat for all of the years up to 2014 but have been increasing steadily starting in 2015, making it seem that this goal has been achieved.

One side effect of increased participation seems to be a marked increase in participation in candidacies which were not successful. It is unclear if this is due to candidates being more widely opposed. This data provides no insight into the supposed "quality" of RFA participation as no such data is collected.

2020 analysis

[edit]

RFA participation has not continued to increase since a marked improvement following the 2015 reforms. Overall participation has been basically stagnant, other than a rapid decline in participation in unsuccessful candidacies becoming evident over the past 3 years.

Table 3: discretionary range

[edit]

The 2015 RFA reform discussion identified the narrow discretionary range as a problem with the process, and suggested that expanding the range would allow more discussions to be evaluated by bureaucrats and minimize the effect of "pile-on" opposition. The discretionary range was expanded by 5% as of the 2015 reform.

Year Total RfAs closed in
2015 range (70-75%)
# successful Total RfAs closed in
2016 range (65-75%)
# successful Lowest pass Highest fail
2020 0 1 1 70% 41%
2019 1 1 1 1 64% 61%
2018 0 1 0 77% 70%
2017 0 1 1 67% 46%
2016 1 1 4 1 75% 68%
2015 2 1 5 1 74% 74%
2014 1 1 4 1 74% 70%
2013 2 1 4 1 71% 75%
2012 2 0 7 0 77% 73%
2011 2 1 5 1 73% 72%
2010 4 2 7 2 71% 75%
2009 2 1 14 2 68% 70%
2008 12 3 26 3 63% 75%

2016 analysis

[edit]

Changing the discretionary range has had no effect at all on overall promotions. Of the 3 candidacies closed within the expanded discretionary range since the 2015 reforms, none have resulted in a promotion. In fact no candidates have been promoted with less than 71% support since 2009, though it is impossible to say if any candidates with lower support would have been promoted by bureaucrat discussion. Likewise no candidacies have failed with more than 75% support.

2020 analysis

[edit]

There still does not seem to have been any effect from expanding the discretionary range, and since 2015 the number of discussions ending with support in the discretionary range have been so rare that the statistics are meaningless. Since 2008, there have been 50 RFAs ending in the expanded discretionary range (65-70%), with three closed as successful.

Since the 2015 reforms, the highest support that a failing RFA has achieved is 70%, and only four RFAs have passed with support below 75%.

Notes

[edit]
  1. ^ a b "Closed early" candidacies are removed from calculations of successful vs. unsuccessful candidacies. Candidacies are considered "closed early" if they are closed under WP:SNOW or WP:NOTNOW, if they are marked "withdrawn by bureaucrat", or if they are withdrawn by the candidate and have fewer than 10 support votes.