User:Aaron Schulz/New RfA method

Preamble
This proposal is meant as an alternative to the current RfA process.

What is wrong with RfA now?
See WP:WPOA, WT:RFA, WP:BN. Many, many essays and other opinions have already been written about why RfA is too flawed to continue as it does.

"Yes or No" votes are not that evil when long-standing, experienced, and rational community members come together to make as an informed a decision as possible, with full attempt to clear their minds of any unfair biases. The problem is that there is no such filter or screening used here at all, and we are WP:NOT a democracy; we need not shoot ourselves in the foot or let consensus be perverted by ploys and tactics just to look like a meta-democracy. We need to do what is best for the project.

Here are some of the main reasons why RfA has eroded:
 * 1) No longer representative of the community. We have grown in size considerably. It used to be where a few voters were the whole community. You could literally have a clear, true, consensus on an RfA or AfD vote that represented almost everyone. Now things have changed, and this is no longer a given; only a small fraction of users vote on any given RfA. This can lead to an RFA "crowd".
 * 2) No safeguards. RfA votes have no check over "vote-stacking", "agendas", smart sockpuppeteering (acting as a "new user") or any other ploys, as it is based of hard numbers. I could make 5 subtle sockpuppets a week or two before an RfA, and that can still vote. New user votes, who have only a minimal connection to the community, are counted equally as any other user.
 * 3) Almost no room for discretion. Votes strongly emphasize numbers, leading to strong inevitable contention if a Bureaucrat closes a vote "against" the numbers". In the case of sockpuppets above, if the vote was close, people would either challenge the Bureaucrat for "going against the numbers" or "ignoring users just because they are new". This kind of tension will eventually just boil over.
 * 4) No serious requirements to vote. Having some decent requirements would correlate with more experienced voters. Currently, any new user can oppose a candidate just because they deleted an image or page they uploaded (per policy) or some other non-issue that more experienced editors would be less likely to hold a grudge over.

Objective: optimize selection procedure
The ultimate goal of this proposal is to develop a process that is effective at selecting suitable people for the adminship role while not denying suitable candidates as well. This proposal gives Bureaucrats enough discretion that things like "vote-stacking", unexplained votes or "per X" votes, "revenge" votes by oft blocked users who have been in conflict with the candidate, votes based on personal attacks as reasons (such as using the sole reason: "Don't like this user"), votes that treat adminship simply as a "reward" regardless of competence, votes by very new users, and any other such votes no longer have to be weighed equally as votes with reasons provided. There would no longer be clear support/oppose numbers to offer a point of contention over final decisions.

Alternative: follow up on those selected
As long as elected administrators are assumed to have at least have some level of competence, or their actions and effects are by and large reversible, another approach is to follow up on admins more so that try to get choose of who becomes an admin "perfect". This could involve an easier way to recall admins (rather than arbcom), or making adminship temporary, or requiring new admins to be reconfirmed after a few months. These will not be discussed here, instead see WP:RECALL.

Alternative: force RfA to be a discussion for the first 4 days
This would allow for more informed opinions, exchange of information and ideas, and ultimatily more meaningfull votes. It would reduce drive-by votes that are or often turn out to be baseless. People would be more likely to change opinions based on incorrect factual data as well. Additionally, requirements such as having 250 edits 1 month ago (.de wikipedia) could reduce wildly uninformed votes.

Consensus
Consensus generally refers to the existence of some sort of harmony or accordance of viewpoints and opinions among a group of people that sovereignly represents the people that the decisions they make will affect. See User:Voice_of_All/Consensus.

More discretion
Giving Bureaucrats more discretion is the easiest, most affective, and least contentious way to allow for RfA to become more consensus based, and less prone to violations of consensus by lack of vote reasoning and the use of tricks like vote-stacking, "tit-for-tat" voting groups (i.e. "We will all support each other so we can pass"), or any such other subversions of consensus or the good of the community. Votes simply, per the reasons mentioned on the previous section, cannot accomplish this.

Discussion and preliminary closing
RfA will become an organized, effective discussion forum where the merits of the candidate can be debated.

Here is how it will be set up:
 * On the front page, there will be a Candidate statement, Questions for the Candidate, and a Peer review section. The talk page, not here, will be the place for any comments that users would like to add.
 * Edit count stats can go on the front page, as the edit break down is a good factor for determining that an edit is way too new or way too inexperienced, though that’s about the extent of their usefulness.
 * On the talk page, there would be an Opinions (one per user) section and a Discussion section.
 * Users add their one opinion statement to the Opinions section to express their thoughts on the candidate. Each such opinion will start with "Comment by X" or "Statement by X". No threads will be allowed in the Opinions, save the respective users adding on to their own thoughts.
 * Responses to opinions go in the discussion section (this keeps the page clean).

Bureaucrats will be empowered to consider all of the statements on the front and talk page and all of the above factors of consensus to make an outcome decision on the RfA after about 7 days. They need to keep the existence of consensus as the focus of their considerations. One Bureaucrat will close the RfA from further comments and post the outcome. The nominee or nominator cannot close, endorse, or dissent on the RfA.

Suggestion: More good questions needed. Things like "content over community or vice-versa", "IAR", "OFFICE", "wiki over encyclopedia or vice-versa" and what not.

The finalizing of RfAs
After a bureaucrat closes an RfA for commenting, it then waits for a minimum of 3 days for other Bureaucrats to review it. This should help to minimize rogue decisions that may be against the will or interests of the community.

A new section on the RfA front page will be used, as mentioned above, called Peer review.
 * Any Bureaucrat that is not the nominator or nominee can add Dissent or Endorse, with a reason provided in the Peer review section.
 * If, after the 3 days, at least 2 Bureaucrats have not endorsed and 2 Bureaucrats have not dissented on the RfA, then the RfA stays open until either happens.
 * After 3 days from closing from comments, once 2 Bureaucrats object, and a simple majority object, the RfA is promotion is reversed before being finalized.
 * After 3 days from closing from comments, once 2 Bureaucrats endorse, and a simple majority do not object, the RfA is promotion is finalized as is and rights can be set.

Note: Bureaucrats should be expected to being following RfAs for the 7 days, so the 3 day time period should not be particularly limiting in any way.

Discussion and preliminary closing
RfB will become an organized, effective discussion forum where the merits of the candidate can be debated.

Here is how it will be set up:
 * On the front page, there will be a Candidate statement, Questions for the Candidate, and a Peer review section. The talk page, not here, will be the place for any comments that user's would like to add.
 * Edit count stats can go on the front page, as the edit break down is a good factor for determining that an edit is way too new or way too inexperienced, though that’s about the extent of their usefulness.
 * On the talk page, there would be an Opinions (one per user) section and a Discussion section.
 * User's add their one opinion statement to the Opinions section to express their thoughts on the candidate. Each such opinion will start with "Comment by X" or "Statement by X". No threads will be allowed in the Opinions, save the respective users adding on to their own thoughts.
 * Responses to opinions go in the discussion section (this keeps the page clean).

Bureaucrats will be empowered to consider all of the statements on the front and talk page and all of the above factors of consensus to make an outcome decision on the RfB after about 7 days. They need to keep the existence of consensus as the focus of their considerations. One Bureaucrat will close the RfB from further comments and post the outcome. The nominee or nominator cannot close, endorse, or dissent on the RfB.

Suggestion: Several standard consensus questions needed.

The finalizing of RfBs
Same as RfA.

Additional Oversight
The above procedure required Bureaucrats to watch over each other for RfA decisions, and for the ArbCom members to watch over RfBs. By simply letting Bureaucrats control who gets to be an admin and even more serious, who gets to be a bureaucrat, you create an inescapable power vacuum. Additional, more community involved, measures should also be required.

Term limits
Bureaucrat will have to re-RfB 3*365 days (3 years) after their last RfB promotion or give up the flag. For those that are already Bureaucrats, if this proposal were enacted, they would have 365 days before they had to re-RfB.

Arbitration Committee
One change to the Arbitration Committee by this proposal would be that a user cannot simultaneously be a Bureaucrat and on ArbCom. While they are welcomed to run for one of those positions, while having the other, they would have to drop the old one if they succeeded. Bureaucrat elected to ArbCom would have their flag removed by Stewards.

The ArbCom's role is simply to check the closing Bureaucrat to prevent the appearance of a self-supporting "Bureaucrat cabal".