22

Prior to July 2015, site graduation was basically decided subjectively by the SE team, based on the site's Area 51 statistics and the capacity of the design team. From what I recall, it was basically a queue-based system, with a site's "position" in the queue being basically decided by the number of "Excellent" Area 51 statistics.

However, in July 2015, the criteria changed: now the main, essential (essentially sole) criterion is that the site must consistently have 10 questions per day on average: sites that fall short of this will still be considered beta sites.

Why was this chosen as the threshold number, and why was this made the essentially sole criterion, rather than considering other statistics? In particular, why are sites with excellent answer and view rates, but with lower questions per day, considered "worse" on that basis alone? Under the old system, those sites with "Excellent" statistics otherwise would graduate.

In particular, I'm looking for statistics that make it clear that sites with fewer than 10 questions per day perform significantly worse (in terms of post quality, moderation ability, etc.) than those with more, and that there is no significant difference when considering other Area 51 statistics; basically, statistics that clearly prove that questions per day is the sole decider of a site's health and that 10 is the right number to set the threshold at.

Update: On August 1, 2019, the 29 sites that had been actively sustaining public beta status for at least seven years launched out of beta despite not having at least 10 questions per day. But this was effectively an exception, not a general rule (as sites that later hit this threshold are still beta today), so my same question still applies: why was the essentially sole criterion for being able to graduate/launch out of beta specifically chosen as 10 questions per day, without considering other statistics regarding a site?

11
  • 2
    I also think your point "... sites that fall short of this aren't even considered for graduation..." will be misconstrued as being flammable, you have no solid evidence that SE doesn't even consider promoting Writing SE. Everyone on the team implies that Writing.SE will graduate, but no one is willing to commit a date. It's all very nebulous and quite unprofessional IMO. Jul 1, 2019 at 13:46
  • 11
    @Mari-LouA we can't even get them to drop the "beta" label, which is a much smaller step than "graduation". There's evidence of that on our meta. And Robert Cartaino deleted his MSE answer where he said he hoped to effect change on the fate of long-running not-really-betas-but-labeled-that-way. Jul 1, 2019 at 15:15
  • 4
    And not just Writing.SE, there are others that are 9+ years old as well such as Board & Card Games.SE. The decision on that site was to wait to see how things go with Writing.SE, but we're of course still waiting. Jul 2, 2019 at 14:33
  • @Thunderforge In fact, if you noticed, I linked to the second-to-last page of Area 51 proposals; those that have been in beta for at least 2,500 days. Many of them have "Excellent" other statistics, but only falter on questions per day. Jul 2, 2019 at 14:35
  • 6
    I don't see anything in today's post that says the the 7 year criteria will be applied going forward.
    – StrongBad
    Aug 2, 2019 at 1:32
  • @StrongBad "The 29+ beta sites that are 7+ years old no longer have the 'beta' label" and "if a site started private beta before July 2012 [seven years before now], that site is no longer 'beta' - it is a full site by our definition" imply that to me. Aug 2, 2019 at 1:49
  • 3
    There's no requirement that a site be 7 years old. I don't really know why it matters that this policy has changed. The change doesn't invalidate your question.
    – Catija
    Aug 2, 2019 at 2:03
  • @Catija Edited the question. Aug 2, 2019 at 2:16
  • 10 surely was chosen because it is a nice round number. 9 or 11 would probably work equally well. The main criterion is probably expected future growth and someone probably observed that above a certain number that future growth was likely to happen. On the other hand it could have been just coincidence or strongly depend on other factors too like the quality of these 10 questions. It may be more like a ballpark number. Apr 30, 2021 at 5:27
  • @Trilarion The question here isn't about the exact number of questions per day, but more so about the fact that the criteria pretty much only consists of "[x]+ questions per day" and doesn't consider anything else (e.g. answer rate, active users, etc.) Apr 30, 2021 at 6:00
  • This rule doesn't work too well for Code Golf SE - we no longer have 10+ questions per day even though we're very active, probably mostly because of how the site works.
    – user
    Apr 30, 2021 at 14:42

0

You must log in to answer this question.

Browse other questions tagged .