On 7/4/05, JAY JG jayjg@hotmail.com wrote:
However consensus only achieves that level of fairness when inaction is less harmful than action.
Which is clearly the case, given that there are no real problems with the existing situation, and 500 admins and growing to take care of admin duties.
There are many axis to consider harm. For example, after making 2000+ edits, failing to abuse procedure, no vandalism, etc.. A user goes up for adminship and is denied. We have 500 admins, why not one more? *Most* people would feel hurt by that.
I'd call that harm.
Obviously we can't just make everyone admins because their feelings would be hurt otherwise... but I think that in cases where a user is likely to abuse the tools after they've been around long enough for 2k edits that it is abundantly clear. In these cases we see unanimous or near unanimous opposition.
In other cases, I think we should give the user a chance to prove themselves as an admin. That the risk of a vandal becoming an admin is infinitesimal at that point, and the the risk of bruising a valuable editors ego is more important.
Exactly. And the people who are denied adminship are generally either not longterm or not trustworthy. Of course there will be exceptions to this, but these are few and far between. If we radically re-vamp existing processes with very low error rates in an attempt to achieve perfection, we are fooling ourselves; no process is perfect, none will ever be, and the likelihood that a new process will achieve fewer errors is low.
I think the people which are unilatterly or near unilaterly denied adminship are potentially not trustworthy. Many of the no consensus users eventually become admins on the second go around, they aren't different people... Usually they just used the time to get to know a few more admins to help their support base. That isn't a bad thing, but it isn't how we should decide adminship. How much good will did we lose from them by denying them once?
Like I said above, I think you are measuring error incorrectly. Our current adminship probably detects all vandals, but it also misidentifies many good potential admins.
To make a more clear example: If the security at the airport simply shot all the passengers we would successfully stop all hijackers pretending to be passengers.
In any system of classification we must weigh both the false positives and the false negative to know the accuracy of the system.
The adminships I cited were not just random failures:
There's no evidence they were "failures" at all.
I meant failures in the sense that they failed adminship. I think it's pretty clear they were not made admins. :)
i.e. There was no consensus.
Yes. It appeared that you understood the point my message started with, that I think that we should accept in the case of no consensus not reject. I'm not sure why you feel the need to point out that there was no consensus when that's what I'm obviously talking about.
By failing to act on these adminships we have done a great disservice to the Wikipedia community.
This is simply hyperbole.
Well it certainly isn't intended to be... I do go on to provide an argument supporting that position.
The community changes and evolves; there are many cases of people who failed at amin nominations the first time, only to be accepted the second time, so there is clearly no systemic issue here. Rather, these individual cases raised specific and individual concerns that had not been adequately dealt with at the time of nomination.
As you point out, many of the no consensus people become admins eventually (the long term trend appears to be all of them). These are perfect examples of where the system has failed once, but worked on the second pass.
How much good will did that first rejection cause us?
No, it turns it into what it should be; a process for ensuring that admins are trusted by the community and created by consensus. There is no duty to take action to fix something that it not broken; on the contrary, it is our duty to ensure that working processes are not damaged by those seeking solutions to non-existent problems which will not improve Wikipedia as an encyclopedia.
or you can say that it is a process they denies a bunch of 'no big deal' permissions to people without any consensus to deny them, and one which insults the integrity of some of our most trustworthy contributors. (Clearly if they eventually pass it they are a trustworthy contributor, so that the process failed them the first time is not a success).
I think you are spending too much effort here defending the preexisting 'working process'. The goal of wikipedia is to be an encyclopedia, not a set of unchanging processes. I do not believe you have demonstrated how it will harm our goal of making an encyclopedia to grant adminship to people without near-consensus opposition.