User talk:SamuelTheGhost/Re-electing admins

WereSpielChequers comments
This is an interesting proposal. I've read several previous proposals for limiting admins terms and this does seem to avoid most of the pitfalls that they have fallen into. I'm not convinced that we need a mandatory term limit like this, but if we were to have one this is close to being workable. But I do have a few comments.  Ϣere Spiel  Chequers 14:59, 19 December 2009 (UTC)
 * 1) Bureaucrats. Rather than measuring the time from their last RFA I would be inclined to do so from their latest successful RFA or RFB, and perhaps even include RFB's where they achieved over 70% but failed to reach the RFB standard.
 * 2) Arbcom members. Rather than simply exempt Arbcom members I would suggest treating their successful election to Arbcom as a re-endorsement and calculate tenure from that date (If Arbcom tenure were reduced from three years to two this would have the effect of exempting sitting Arbs).
 * 3) I'm not sure how you get 730 admins as the longterm minimum number as there is no guarantee that so many admins would agree to serve another term. I think our shortage of admins and their declining numbers is a strong argument against schemes such as this, and would myself be inclined to oppose any such scheme that didn't have a failsafe such as "if the number of active admins falls below 1,000 then no admins will be retired under this scheme until numbers rise again". In practice, assuming that the current drought at RFA continues it would take four years before the admins appointed in 2008 came up for renewal. but all the admins appointed in 2008/2009 would go through this process in less than a year, and it could then take less than a year for the process to catchup. At that point you would perhaps as few as a couple of hundred admins with less tan two years experience, plus however many had been reconfirmed, or elected as crats or arbs in the meantime.
 * 4) Medium term absences. We have a number of admins in the military who cease editing for months at a time when they are on tour. At the least you need provision for admins to submit a reconfirmation RFA early so they can get it out of the way at a time of their choosing, and have that count as that day's reconfirmation RFA.
 * 5) Two years is too short a term, both in terms of the practical issue of making sure we have enough admins, and especialy enough experienced admins; But also because the third year is the least problematic. A term of three and a half years would make much more sense.
 * 6) This scheme omits the option of retraining/refresher training. Wikipedia is a complex and changeable place, an ongoing program of retraining of admins would IMHO be a sensible thing to introduce. If you then had a retirement/reconfirmation a day for admins who have gone over three and a half years since RFA, or an RFB or more than two years since passing a Computer-Based Assessment, you would catch up much quicker, and in my view, have a much better system than you propose.


 * Thanks for your useful comments. My responses:

SamuelTheGhost (talk) 18:10, 19 December 2009 (UTC)
 * 1) I admit I'd not thought about bureaucrats. I agree that dating from their latest successful RFA or RFB would be sensible.
 * 2) It would be simpler to exempt any current holder of a status which is already subject to a re-election process. That means arbcom and stewards I think.
 * 3) There isn't a declining number of admins as such, though there may be a shortage of active ones. Although I think your predictions here are a bit alarmist, of course it would be simple to impose a minimum number of admins below which the re-election process would be suspended. I find the proposed minimum figure of 1000 a bit high, but I wouldn't want to argue about it.
 * 4) As the proposal is currently stated, this is already allowed for. Re-election can be sought at any time.
 * 5) I'd have no objection to a longer time if that's what people think. It wouldn't have any effect for several years, in any case, so there would be plenty of time to re-estimate what was going to happen.
 * 6) The idea of admin training is a good one, but currently it is not mandatory. Introducing mandatory training or retraining needs to be a separate proposal, discussion of which would raise a whole new set of arguments. I'd be rather firmly against incorporating it into this proposal here, which has the advantage of simplicity.
 * I've made the bit of the RFB change that we agree on, I suppose we have to differ on people who achieve the 70-75% that would be an RFA pass but fail to get the 85% needed for an RFB. But I suspect if this went ahead we could see RFB change with some !voters saying good admin - not yet ready for RFB and crats having the option of closing the RFB as not successful, but RFA reconfirmed.
 * Exempting all elected office holders would include oversighters and the new audit committee, I agree that would be a sensible exemption, but I'm not sure that it would be entirely sensible that anyone standing down from or finishing a term as an Arb or checkuser would then immediately face a reconfirmation RFA - hence my suggestion to treat a successful election to a position of trust as the equivalent of a reconfirmation.
 * Sorry you are correct there isn't a declining number of admins, but there is a declining number of active admins. There are currently 862 of us, there were 951 on the 19th Dec 2008, and 1002 on the 3rd of April 2008. If you accept the principle of a minimum number of active admins below which we suspend the reconfirmation process, then the threshold is crucial. In my view we are already below the minimum threshold that I would be comfortable with. Though I suppose we could have a process whereby we suspend reconfirmation for active admins whilst their numbers are below a thousand, but start the process for the 835 inactive ones.
 * Though the term limit would not cut in for years it is in my view essential to decide it at the outset - subsequent changes to such things can be contentious. Currently you propose a two year limit, I believe this would be a very bad idea, partly because of the extremely small number of admins we have appointed in the last two years, partly the need for a proportion of experienced admins but mainly because what data we have indicates that admins in their third year are the least likely to be so contentious as to be desysopped.
 * Glad you agree with the admin retraining idea, I'm afraid we have to disagree on its link with this proposal, as I would be opposed to a reconfirmation system which didn't have refresher training as an option.  Ϣere Spiel  Chequers 13:05, 20 December 2009 (UTC)

Non-arbitrary break

 * Very nice idea - pity it seems to be neglected. I'm very much in support of some sort of reconfirmation process and think this one addresses the concerns of it being too onerous nicely.  However I cannot agree with WereSpielChequers' position that the process should be suspended if the number of active admins drops below a certain level, because it defeats the purpose of letting the community make that call.  I expect that if, for example, we were down to a very low number of active admins like 500, the community would be much less likely to rescind the bit.  There would probably be some editors arguing that very position, and others saying it's a terrible reason to reconfirm User:BorderlineAdmin.  On the flip side, I tend to think that people would relax their standards at RfA if there was a way to remove an underwhelming (but not actually rogue) admin, which would relive some of the pressure.  Putting a minimum number of active admins will artificially protect inadequate admins if the timing is right, and that is an arbitrary factor that defeats the purpose of this proposal.  Trusting the community to set the standards is what this site is based upon, and I sincerely hope our faith in this community is stronger than the USA's faith in capitalism was during 2009!--~TPW 23:52, 10 May 2010 (UTC)