User:Extraordinary Writ/No consensus at DRV and MRV

Sometimes, alas, debates at deletion review and move review do not reach a consensus. As a general rule, "no consensus" means the status quo prevails. But DRV and MRV each have a special rule that applies to this situation:

• At DRV, Wikipedia:Deletion review#Closing reviews says (as of this revision) If the administrator closes the deletion review as no consensus, the outcome should generally be the same as if the decision was endorsed. However...[i]f the decision under appeal was an XfD close, the closer may, at their discretion, relist the page(s) at the relevant XfD.
• At MRV, Wikipedia:Move review#Closing reviews says (as of this revision) If the MRV closer finds that there is no consensus in the move review, then in most cases this has the same effect as Endorse Close and no action is required on the article title. However, in some cases, it may be more appropriate to treat a finding of "no consensus" as equivalent to a "relist"; MRV closers may use their discretion to determine which outcome is more appropriate.

Therefore, while "no consensus" typically means the closure is effectively endorsed by default, closers at both DRV and MRV have the discretion to relist the underlying discussion instead. When should this be done? Neither page provides much guidance, so this essay aims to list (in no particular order) some of the factors that I generally consider when this situation comes up in a DRV or MRV that I'm closing. Obviously nothing here stems directly from our policies and guidelines, but these factors reflect rationales I've heard from other closers and shouldn't be too far out of the mainstream.

• Has the underlying discussion been relisted? How many times? AfDs are regularly relisted twice; third relists are uncommon but not unheard of. RMs are only supposed to be relisted once, but in practice second relists are not rare. Other XfDs have their own norms (e.g. MfDs are rarely relisted at all). I am most likely to relist following a DRV or MRV when the underlying discussion hasn't been relisted at all, and I'm less likely to relist when there have already been several relists. Relisting a discussion that's already been relisted several times often just drags it out even further for no real gain.
• How much participation was there in the underlying discussion? If the XfD or RM was sparsely attended, a relist may do some good, especially if participants in the review discussion decide to add their insights. The classic case is when the DRV or MRV has gotten far longer than the discussion it's supposed to be reviewing: in that situation, kicking it back for further discussion on the merits is often useful. If the underlying discussion was already well attended, though, another relist may not provide much clarity.
• Has new information come up in the DRV/MRV? Relisting allows for additional sources or arguments to be evaluated on their own merits. If no new points have been raised since the original discussion, by contrast, a relist may just let the same participants say the same things for another week.
• Was relisting suggested in the review discussion?. The people who comment at DRV and MRV are generally quite bright indeed, and if it never occurred to any of them to suggest relisting, that's probably because it isn't a very good idea. Stay on the right side of the fine line between discretion and supervoting.
• How close was the DRV/MRV to consensus? There are several flavors of "no consensus". If there was very nearly a consensus to endorse outright, a relist is less likely to be a good idea than if the the discussion was tilting towards overturn or relist.
• What's the chance a relist will make a difference? The other factors get at various aspects of this, but it's worth stating specifically: if it's apparent that further discussion isn't going to change the result, using your discretion to relist is usually a poor idea. Volunteer time is our scarcest resource, and when the outcome is very unlikely to change with yet another week of discussion (or if it's only the keep-vs.-no-consensus distinction that's at issue), just default to endorse.

Oftentimes these factors tilt in more-or-less the same direction, and that makes the closer's choice easy. If there's no clear answer, note that DRV and MRV both say no consensus should "generally" or "in most cases" (respectively) default to endorse, so it's fine to err in favor of letting the closure stand. To be clear, these are just considerations: closers ultimately have to use their discretion, and this essay should be used as a guidepost rather a formula. But I do think writing down the broad outlines of my thinking has some value, if for no other reason than so I can convince people I didn't just flip a coin (or supervote) when I closed the discussion they're complaining about.

References