RfA inflation is the belief that as time goes on, the standards for an RfA candidate rise. It is a result of a statistical approach to candidate analysis, with the statistics needed gradually rising with time. The most obvious dangers being that of an overall decline in successful RfA, leading to a decline in active administrators. Problems will result if there are an inadequate number of active admins to do the tasks that require their toolset.

This is an essay. It contains the advice or opinions of one or more Wikipedia contributors. This page is not an encyclopedia article, nor is it one of Wikipedia's policies or guidelines, as it has not been thoroughly vetted by the community. Some essays represent widespread norms; others only represent minority viewpoints.

Shortcuts WP:INFLATE WP:RFAINFLATE

This page in a nutshell: RfA inflation is the inflation of standards at RfA, which is a large problem on Wikipedia. The solution is not to judge candidates too much on statistical metrics, but instead on level of trustworthiness.

RfA inflation can take many different forms, but it often applies to statistical analysis of RfA candidates, and the demand for ever increasing positive statistics (often in the form of editcountitis) can be a factor in RfA success.

Here is a list of a few of the forms it can take.

• Overall edit count: this is the most obvious metric. While this provides a useful metric for overall experience, inflation makes this increase as time goes by. More and more experience is needed, reducing the pool of potential candidates.
• Account age is another obvious metric than can increase with time. There is generally less trust with comparatively newer accounts, but inflation would make the time needed for potential admins to be successful longer and longer.
• Namespace balance arguments, while generally to be avoided anyway, can show inflation towards mainspace edits in particular.
• Age is another factor. Young admins are less likely to succeed today than 5 years ago.
• Content creation inflation:
  • Overall content creation metrics, for example the number of mainspace edits. Mainspace edits using semi-automated tools can sometimes be discounted, as they are generally used for vandal fighting, leading to an overall higher count being required.
  • Demand and expectation can also increase with content creation quality, particularly with FA and GA credits being demanded. See also: "must have 10,000 edits, three featured articles...".
  • New article creation may also be demanded, and again the number of articles, or indeed the demand for RfA candidates to be new article creators, can rise with time.
• Experience in administrative areas is another demand that can increase with time. Whilst not as easy to quantify, it can take the form of Articles for deletion and Criteria for speedy deletion statistics with overall level of participation, as well as accuracy rate, being relatively easy to quantify. Again inflation in this area would require both of these statistics to increase with time.
• Pointing to problem edits, while down playing positive contributions, can also be subject to inflation, with higher and higher levels of perfection being asked for.
• Demanding experience in many different areas of Wikipedia can also create problems. Typically candidates specialize their experience into a number of niche areas. Expecting candidates to have experience in most areas of Wikipedia can be demanded, and again subject to inflation with time.

The reasons for RfA inflation are probably as complex and varied as price inflation. But a few theories can be put forward.

• Obviously the age and stature of the project has increased with time. Many editors here have high edit counts, as well as high statistics in other areas. So while several thousand edits might in the past have been seen as being high, today it is commonplace. Editors may feel that other editors with lower edits counts are not "worthy" of adminship on a project of such importance. Standards for articles and so forth have also risen, and this perhaps mirrors RfA standards.
• Prestige of adminship. As the Wikipedia has grown in size and importance, being an admin on Wikipedia has also grown in prestige and importance. Jimbo Wales tried to argue against this in his famous "no big deal", where being an admin should be seen as being of little importance, just editors with enhanced tools. But human nature tends towards progression and success, so adminship being seen as one of them is perhaps inevitable to some extent. If adminship is seen as being prestigious then editors may not like other editors becoming "one-up" because of this.
• Fear, uncertainty and distrust. As the size and importance of Wikipedia has grown, so has a general feeling of alienation within the community. Large communities are less cohesive. The formation of bodies such as ArbCom, as well as many notable cases against prominent content creators may have provoked a siege mentality. Some admins have been desysopped, so fewer people trust admin candidates. One way of counteracting this is to demand higher and higher standards for admin candidates, or only admin candidates who are prominent content creators, without either being particularly necessary. Also, editors who have had difficulties with admins in the past may well seek higher standards.
• The availability of complex statistical tools, such as X!'s Tools, can also produce an increasing emphasis on judging candidates on a statistical basis, rather than on trustworthiness.
• Norming of the Opposes. There is a psychological effect on !voters, with a tendency for those who have lost a !vote to rethink their criteria and fit into the winning side in the future. In simple majority systems this causes regression to the mean, but in systems with a super majority this has a ratchet effect, with losing sides of as much as 70% having their criteria challenged and winning blocking minorities of as little as 30% having their criteria reinforced. Over large numbers of RFAs this should in theory see easily measured arbitrary criteria such as edit count and tenure ratchet upwards, which appears to have happened. One silver lining of this theory is that the process should slow as the number of RFAs falls.

All of this contributes to a loss of potential admins. For example, as can be seen here, the number of unsuccessful RfAs is going down. There were 155 unsuccessful RfAs in 2010, but only 38 in 2015. This decline doesn't affect just the unsuccessful RfAs. In all successful RfAs, there were 75 admins appointed in 2010, but only 21 in 2015. The rate of all RfAs is going down, and this is largely because of higher standards for admins, as discussed above.

The main problem that may result is having too few active admins for the admin workload. While the unbundling of tools has helped to some extent, some tasks can only be done with the admin toolset. While at present there does not seem to be a particular crisis, this could happen in the future, especially if RfA inflation continues unabated.

The solution is not to judge candidates too much on statistical metrics, but instead on levels of trustworthiness and need for the tools. While statistics help to determine overall levels of participation, they do little to help with assessing either of these. A candidate may have low participation in comparison to many of the editors here, but is it really a good way to assess trustworthiness? A reasonable track record can be established from a few thousand edits.

Also, try not to see adminship as a trophy or status symbol or reward for good contributions. Instead it should instead be seen as a necessary toolset for certain activities on Wikipedia. The only two important questions are will the editor use the tools correctly, and will they be likely to use them. If they are both answered in the affirmative then questions whether they "deserve" them become moot. Lessen the statistics, and direct your arguments to these points.

Another good argument is that of being a net positive. Admin candidates are always going to have made mistakes in the past, but these should be countered with their positive contributions. Demanding higher and higher levels of perfection may run counter to the well-being of the project.

Perhaps one of the most inflated requirement is prolific content creation, with the belief that content creation teaches editors how to interact with other editors. But there are arguments against this, as other interactions, such as anti-vandalism work and dispute resolution, are also very valuable at teaching this, which should not be discounted.