Wikipedia talk:Consensus not numbers

From Wikipedia, the free encyclopedia

[edit] Completely wrong

The problem with this idea is that, if voting is abandoned, we abandon the principle that every Wikipedian's opinion carries equal weight. Admins and bureaucrats are not supposed to be decision-makers; they're supposed to be functionaries who carry out the will of the community, even where it conflicts with their own opinion. However, the less we use voting, the more discretionary power we give to admins or bureaucrats who close discussions. Saying "consensus not numbers" is a false dichotomy. "Consensus", in the real world, means most people agree on an issue. It does not mean "hmmm, well, 10 people said Keep and only 5 said Delete, but the latter had a more convincing argument, so I'll delete it anyway". That is miles away from the RL meaning of consensus. This is why (at least on RfAs, and preferably on XfDs as well), voting is a useful tool. What this essay seems to be saying is "Feel free to state your opinion, but it won't necessarily make any difference." That's a dangerous way for Wikipedia to go. Waltontalk 12:22, 5 July 2007 (UTC)

  • Not every wikipedian's opinion carries the same weight. A brand new editor who knows nothing about our policies and principals can not and will not have the same voice as someone who does know our policies and principles. It is well known you are highly in favor of voting. However, voting is very antithetical to a number of our processes here. Observing and evaluating consensus without vote counting does not mean the will of the community is being ignored. --Durin 14:35, 5 July 2007 (UTC)
How do you "evaluate consensus without vote-counting"? What I really don't get is this usage of the term "consensus" on Wikipedia, which differs markedly from the normal meaning of the word. In the real world, a "consensus" would imply that most people agree on an issue; if 7 people argue one way and 6 argue another, then there's no "consensus" however you look at it. I think we need to stop mis-using the term "consensus" and say what we really mean, which is "the closer weighs up the arguments and comes to a decision". That's a valid (albeit rather authoritarian) way for our processes to work, but it has nothing to do with consensus. Waltontalk 16:41, 5 July 2007 (UTC)
  • In the case you noted above, I'm presuming you mean that 7 people are in agreement in a particular way, and 6 are in agreement in another way. True, that's not consensus. But, it's rare that 13 people divide themselves up so neatly. Further, 7 in support and 6 oppose might be measured as no consensus. But, that's arbitrary. Is 8-5 consensus? What about 9-4? 10-3? 11-2? Where do you drawn the line, and more importantly why? Consensus isn't about numbers. Weighing arguments IS about consensus. You might want to read m:User:Durin/Thoughts_on_consensus. --Durin 17:37, 5 July 2007 (UTC)
Hmmm, I find your "thoughts on consensus" quite interesting - I don't particularly agree with it, but I now understand your point of view more clearly. As far as I can see, you make three distinct valid points, which I'll attempt to address:
  • Because of the 70-80% threshold, small numbers of oppose votes carry undue weight and can sink an RfA, even if the opposes are pointless. This is true; however, in the current climate at RfA, a really stupid oppose vote ("Aargh! Not enough portal talk edits!") tends to receive a deluge of criticism from other editors, and such people are often pressurised into changing their vote. Ditto with an oppose vote given without a rationale. As such, the culture of RfA has adapted to compensate for the mathematical flaws in the system.
  • Last-minute opposes can sink an RfA, without giving anyone time to respond to them. This is also true, but it's something of an inevitable flaw. Even without voting, bureaucrats would still take opposition into account, and a last-minute bombshell could still be a threat. So I don't think your preferred solution would solve this problem.
  • Users don't agree on the criteria for RfA, and the overall votes don't necessarily reflect consensus on an individual point. This is harder to answer, but I would say that users have a "right" (well, a conditional privilege, but you know what I mean) to judge the candidate by whatever criteria they think fit. In the example you gave, it's true that some people liked the candidate's answer to Q1, some disliked it, and some thought it irrelevant. But the fact remains that each individual user weighed up the candidate's suitability against their personal criteria, and made a decision. I think we can trust users to do that, rather than trying to hammer out consensus on each point and come to an overall decision.
The reason I don't like your Matt Britt solution is that it makes the process complex and hard to understand, and inevitably reduces the level of accountability. In such a case, the bureaucrats would have to weigh up the candidate against all the criteria, and come to a decision - but we would have no way of holding them to account for that decision. Note also that the Matt Britt experiment was not considered a success by most of the community. In conclusion, while I don't think that voting on RfAs is perfect, I think it's better than any of the alternatives. Waltontalk 09:52, 6 July 2007 (UTC)
  • RfA is unique in this respect. In an AfD, if 7 people want an article kept because it has no problems that couldn't be fixed, and 100 people want it deleted because it doesn't have enough Zs in it, then clearly the 100 people who want it deleted should be ignored, and it should be kept. But if 7 people support a candidate because there are no problems with his record, and 100 people think that lack of Zs in a usernamename is prima facie evidence of power hunger, it doesn't matter. The community has horribly stupid critieria for adminship (and I assume we'd all run away), but if they don't trust someone, it doesn't matter. Admins need the trust of the community, no matter what stupid hoops the community wants them to jump through. -Amarkov moo! 16:59, 11 August 2007 (UTC)
    I think you're completely wrong, here. An infantile reason for opposition would be ignored, no matter how many idiots rolled up on the day to advocate it. In practice, the bureaucrats have felt free to set aside, in substantial numbers, even perfectly reasonable objections because they judged that the objections were not related to a candidate's conduct as an admin (the classic case being the setting aside of some opposition to Danny's second RFA on the basis of actions carried out by Danny during the course of executing his duties as an employee of the Foundation). --Tony Sidaway 15:53, 10 November 2007 (UTC)
  • The notion that every Wikipedian's opinion carries equal weight has never been, and I think never will be, part of Wikipedia policy. --Tony Sidaway 15:42, 10 November 2007 (UTC)