Area 51 says:
90% answered is a healthy beta, 80% answered needs some work. In the beta
We are far from 90% . How do we tackle that?
$\begingroup$
$\endgroup$
-
2$\begingroup$ No, the main thing we need to work on is site visits and daily users. If we don't post any new questions than this will kill drawing in new people, and existing users will also leave. $\endgroup$ – Cody Aldaz Apr 30 '20 at 6:32
-
$\begingroup$ That makes sense. $\endgroup$ – Thomas♦ Apr 30 '20 at 6:45
-
$\begingroup$ I don't know if we want to leave the percentage listed in the question the way it was at the time when the question was first asked, or if we want to keep updating it once in a while, but right now it's at 72% :) $\endgroup$ – Nike Dattani May 1 '20 at 18:22
$\begingroup$
$\endgroup$
The statistics and ranges on Area 51 were set almost a decade ago, under a very different idea of how launching sites would work (among other things, the idea was that every site would reach Stack Overflow levels of popularity and traffic). The actual ranges that matter for these numbers:
- Questions per day: For most sites, as long as this stays above 1 or 2, things are fine. Most sites will hit the graduation target of ten questions per day between two and five years after launch.
- Percent answered: This is mostly an indication of the sort of questions the site attracts. Worldbuilding graduated with 100% answered, while Law graduated with 77%.
- Avid users: The targets listed don't matter. What matters is that by the time the site hits 10 questions per day, there's a group of regular users with at least 3000 reputation to perform routine content curation (closing duplicate and off-topic questions, editing posts for formatting and style, etc.) and at least a few users with 10,000 reputation to perform moderator-like tasks such as deleting problematic posts.
- Answer ratio: Like "percent answered", this depends on the nature of the site. 2.5 answers per question may be good for a site like Stack Overflow, where there's usually more than one way to do something, while 1.5 would be reasonable for something like Law, where there's usually only one correct answer.
- Visits per day: Largely irrelevant. For most sites, this tracks fairly closely with questions per day.
-
$\begingroup$ I appreciate this answer. However with the increasingly challenging (almost impossible) criteria to survive Area51 (maybe 90% of the proposals were killed when they made the 1-year commitment limit, then 99% were killed when they made the 4-month commitment limit), and the almost immediate closure of almost all the new proposals being made if they don't rapidly demonstrate that they are serious, then the answer that they are "not sure" whether or not new sites will be allowed: meta.stackexchange.com/a/344244/391772, are you still completely sure that we will be okay with only $\endgroup$ – Nike Dattani May 1 '20 at 4:58
-
$\begingroup$ 77% answered questions? I have the impression that they only want the best of the best proposals surviving now. $\endgroup$ – Nike Dattani May 1 '20 at 4:59
-
$\begingroup$ @NikeDattani, completely sure. At about the same time they tightened the criteria for creating new sites, they loosened the criteria for keeping a site open. Once a site makes it out of private beta, the only reason it'll be closed is if the community is no longer capable of keeping it free of problematic content. See, for example, Hardware Recommendations, which is struggling to keep the answer rate above 50% but is still open, and Augur, which was only closed after a call for moderators went unanswered (and unread) for more than a month. $\endgroup$ – Mark May 1 '20 at 22:05
-
$\begingroup$ That is reassuring. I am fascinated that the call for moderators on Augur went unread for more than a month! I do worry that 1/3 proposals fail in private beta. I understand Artificial Intelligence was shut down because there was too many non-expert type questions, and other were shut down because of not enough activity. Could you give us some feedback about our activity level: area51.stackexchange.com/proposals/122958/materials-modeling, and perhaps question & answer quality? $\endgroup$ – Nike Dattani May 1 '20 at 22:55
-
2$\begingroup$ @NikeDattani, compared to failed private betas I've been in, I'm not seeing any of the early warning signs of trouble (eg. half the questions being posted by a single user, or large numbers of questions that could be answered by typing them into Google). The activity level right now is about 70% that of UAVs or Law, comparable to that of Constructed Languages, and way ahead of Bioinformatics. Assuming activity levels follow the usual patterns for private beta, things should be fine. $\endgroup$ – Mark May 1 '20 at 23:31
-
$\begingroup$ (And the Augur call for moderators wasn't entirely unread, but 17 views is far below the usual level of attention. Even the site closure announcement only got 144 views.) $\endgroup$ – Mark May 1 '20 at 23:39
-
$\begingroup$ that is very reassuring, and I'm glad to hear this. I'm surprised Bioinformatics had a bad time, because that is a huge field (however they were competing with the "BioStars" site which is a "clone" of SE that had been running for years already). I followed the UAV/Drones site very closely and did notice that they had about 80 questions in the first day or two, but when I compare screenshots of the A51 stats, we at least have more visits/day than them :) I find that almost "anyone" can own a drone and ask UAV questions, but our questions are research-level and far more specialized: $\endgroup$ – Nike Dattani May 1 '20 at 23:51
-
$\begingroup$ Therefore the pool of people that can easily participate is (almost) limited to people that are experts in the topic, or beginner academics (ex. grad students) that are learning it. Then there is the issue that it might take more time and thought to ask a highly specialized question, than to ask for example if a drone bigger than a certain size is legal to fly in some country, so this almost makes it hard for us to have a high number of questions/day. $\endgroup$ – Nike Dattani May 1 '20 at 23:54
$\begingroup$
$\endgroup$
No. We should not concern us with these statistics right now (if ever). It is more important to figure out which questions are on topic, and which are not, because this will influence the way the site works the most. For this we need questions, plenty of them. If, by the end of the year our answer ratio isn't where we want it, we need to figure out a way to deal with that. But let's burn the bridge once we've passed it.
Right now, this is a small, but determined community, as of now 32 people have obtained precognitive. We should focus, although I have no idea how to do that, on getting the remaining commitors to the boat.
$\begingroup$
$\endgroup$
We can focus on answering the un-answered questions, which might be difficult, because these are expert-level questions which require a lot of time and careful thought, so they won't just be answered overnight.
Alternatively we can focus on raising awareness about this site, now that it exists! Then as we get 100s of expert users joining, all of the statistics can improve organically :)
$\begingroup$
$\endgroup$
I would just like to give an update now that about 10 days have passed since the question was asked.
When the question was asked, 61% of the questions had been answered, and we had still only finished 1 full day of Private Beta. Before the second full day of Private Beta went finished, we were up to 66%. By the end of the 4th day we were at 69%, and we've been at at least 70% on each day since then (sorry that the second image is a bit bigger than the rest, it's because I took the screenshot using landscape instead of portrait, on my phone):
This increase from 61% to 73% was despite everyone here saying not to worry about the answer percentage: None of us "tried" to purposely bring these numbers up (especially me: I told myself I'd spend my energy more on asking questions than answering them, and I told all my friends to do the same too, but somehow I ended up answering 9 questions and only asking 6, because to my surprise, many questions came up that I didn't foresee coming, which I genuinely wanted to answer!).
I do feel it would be nice to have 90% of the questions answered, but Mark and Martin know Stack Exchange much more than me, and they are suggesting that we shouldn't be too fussed about it. At present, people are not advertising the site as much as they would if the site were Public, so it is expected that our answer percentage will not be as high now as it will after a week in Public Beta.
Also, keep in mind this is an SE on an expert topic, much different from other science SEs like Chem.SE and Physics.SE which cover general subjects that are taught in elementary school, high school, and undergraduate-level university. We won't get homework questions from students, instead we get research-level questions from researchers, which take a bit more time to solve, which means answers will lag behind questions, so we will never get close to 100%, as long as questions keep being asked and answers keep taking a couple weeks to arrive (which is perfectly okay).
More than 60% of the committers listed themselves as Experts or Academics:
Compared to only 11% on the most recent site to graduate out of Private beta (Drones.SE):
So the questions are mainly coming from academics or experts, and can take more time to get answers, so a 73% answer percentage is actually very good in my opinion, and if my hopes come true it will get even better once we enter Public Beta.
