# Are we gaining or losing non-novice users?

In any difficult field of endeavor -- ballet, mountaineering, violin-playing, physics -- there are going to be far more people in the world who have devoted a relatively small number of hours to the field than there are who have spent a big chunk of time to it. On physics.SE, non-novice users are in short supply. Is there any way to tell whether the number of currently active, non-novice users on physics.SE is going down or going up? We have certainly lost some non-novice users such as Arnold Neumaier and Ted Bunn.

I suppose a quantifiable way of defining this would be the number of high-rep users. Although rep doesn't really correlate very tightly with competence on an individual basis, and high-rep users may get suspended for being naughty, I would think that the number of high-rep users would be worth looking at on a statistical basis. Of course it won't tell us anything about whether we're retaining new non-novice users, such as ZeroTheHero, who appears to be a non-novice but just showed up today.

• Losing. Sure, ZeroTheHero may have joined, but multiple other good users like Dilaton, Ron Maimon, Arnold Neuaimer, Lawrence B. Crowell, even Deepak Vaid (even though he's an LQGist), etc. have left. – Abhimanyu Pallavi Sudhir Sep 1 '13 at 3:08
• @DImension10AbhimanyuPS you can't claim that we're losing good users just by naming a few of them who have left. I'm sure I could name just as many who have joined. If you have data to support a claim, then by all means, go ahead and make an answer out of it, but otherwise, you can't decide on an answer to the question just because it's the answer you want to be true. – David Z Sep 1 '13 at 3:25
• @DImension10AbhimanyuPS: Although I did succumb to the temptation to give a few examples in the question, I would really just be more interested in a statistical answer. It would be interesting to see a comparison with an SE site such as mathoverflow, which is clearly healthy. SE seems to have some fancy tools for generating statistical reports, so I thought someone might be able to shed some light here. There are separate threads on meta about some of the individual suspensions, and I don't see the point of recapping those here. – Ben Crowell Sep 1 '13 at 3:50
• Non-novice users who continue posting answers and questions. – Deer Hunter Sep 1 '13 at 4:02
• Ben: if you could give me a (partial) list of tags where a high percentage of answers is from non novices (e.g. QFT, string theory, GTR ...), I may be able to whip up a query for this :) – Manishearth Sep 1 '13 at 4:30
• @Manishearth: Cool, sounds interesting! Here are some that I came up with that might work: quantum-field-theory general-relativity string-theory quantum-gravity statistical-mechanics condensed-matter renormalization – Ben Crowell Sep 1 '13 at 4:43
• @Manishearth: I think it would be important to exclude the pop-sci questions... – Abhimanyu Pallavi Sudhir Sep 1 '13 at 5:06
• That's a good start , I'll have a crack at it when I get back to my laptop. No guarantees though; complex SQL makes my head hurt :p – Manishearth Sep 1 '13 at 5:48
• As an aside, my take on this: I do have a habit of skimming through new higher level answers. While I don't understand everything, I usually get an idea of things (from which I can build on) and it's quite a fun experience. Aside from the users we lost in December, I haven't noticed any regular answerers disappearing. I have, however, seen many new faces popping up repeatedly. – Manishearth Sep 1 '13 at 7:51
• @Manishearth Just out of curiosity - what is average age of moderators here? What is their average affiliation? PhD Student Professor? Including previous moderators. – Asphir Dom Sep 1 '13 at 13:46
• Is this where I accuse @Ben of having a theory bias? – dmckee Sep 2 '13 at 2:08
• how would you define non-novice users? What counts? I'm a grad student in physics, am I non-novice? Is an undergrad student non-novice? Also Dilaton is gone!? :( What black magic is this? – Jim Sep 2 '13 at 12:52
• I would definitely consider grad students novices. A grad student may be very bright and knowledgeable but he would lack the experience to be a non-novice. This of course is just my view. – MBN Sep 2 '13 at 13:53
• I would consider grad students non-novices, but of course the whole concept is vague, and the statistical measures we have available through database queries don't give anything but loose proxies for it. – Ben Crowell Sep 2 '13 at 15:09
• Grad students are by definition transitioning from novices to non-novices. Senior grad students in particular should have developed a fairly sophisticated understanding of at least one corner of physics. Sure, that leaves them pretty amateur in many other corners, but then who here is an expert on the whole of physics? I know I'm not. – dmckee Sep 2 '13 at 17:03

Created some queries:

• Number of active users in a tag month wise: This graphs the number of users who have made more than threshhold (this can be edited) posts in the list of tags in a given month, over all months.
• Monthwise streaks in a tag: This was an intermediate query I created which is used in the one above. It tabulates the number of posts per month per user in a given tag.

For the first query, for threshhold=2 (So each user needs to have at least 3 posts in the tag list to qualify), I get this graph:

We seem to be on an overall increase, but local decline(since April). Probably due to seasonal fluctuations, note that there was a similar drop in the summer of 2011.

I'll try to make a better query that shows retention later.

• In the meantime I also made this query. It can easily be extended to talk about multiple tags, however I'd prefer to take the top active users in the tag/tag list and show them instead of relying on rep for creating the threshhold. – Manishearth Sep 1 '13 at 9:42
• Using your 2nd query, For string-theory, it seems that 4 of the top 8 users are inactive... data.stackexchange.com/physics/query/edit/133980 – Abhimanyu Pallavi Sudhir Sep 1 '13 at 9:45
• For your 1st, I used a more stringent criteria (st, lqg, cm) and I got quite a decrease: data.stackexchange.com/physics/query/edit/133983#graph; from 6 to 5 – Abhimanyu Pallavi Sudhir Sep 1 '13 at 9:49
• @DImension10AbhimanyuPS (re:comment2) That's an increase. Note that August doesn't count as there's only 3/4 of the data for August in SEDE right now. I don't know how to graph the moving average on SEDE; but that's what we need to look at. – Manishearth Sep 1 '13 at 9:50
• @DIm (re:comment1)The question here is overall amount of non novice users. Individuals come and go; this is normal for any online community (and not a fault of the community). There's a lot of excitement when a site starts which peters out later on. The first few months on Chem.SE were quite active, for example. If you run that query for almost any tag that was around back then, you'll notice that most of the top entries are in 2011. – Manishearth Sep 1 '13 at 9:53
• OK, Looking at this with threshhold 5, it seems like it is currently the same as the begining, but heading towards a minima. Oct 2011 and Jul 2013 were the maximas, I think I realised that. – Abhimanyu Pallavi Sudhir Sep 1 '13 at 9:59
• Do you know how I can exclude a particular tag? . "NOT" doesn't work. – Abhimanyu Pallavi Sudhir Sep 1 '13 at 10:01
• @DImension10AbhimanyuPS Again, ignore the last point. It's a discrete graph, not a continuous one, so moving average is what we need. If there are >3 points that show a steady decline, then that's a decline. Otherwise, not really. – Manishearth Sep 1 '13 at 10:01
• @DImension10AbhimanyuPS replace LIKE with NOT LIKE – Manishearth Sep 1 '13 at 10:01
• If the last point were graphed, it would only show a lower last point, because Dilaton's suspension took place in the last quarter of August. – Abhimanyu Pallavi Sudhir Sep 1 '13 at 10:03
• – Abhimanyu Pallavi Sudhir Sep 1 '13 at 10:04
• @DImension10AbhimanyuPS How in the world can the number of active people decrease by increasing the time over which we are checking? – Manishearth Sep 1 '13 at 10:05
• Oh wait, it isn't average over time... – Abhimanyu Pallavi Sudhir Sep 1 '13 at 10:06
• – Manishearth Sep 1 '13 at 10:18
• Very interesting. For comparison, I tried a similar query on mathoverflow, using a randomly chosen set of tags meant simply to cover a decent fraction of their activity. Surprisingly, they seem to have a clear downward trend over the last few years: data.stackexchange.com/mathoverflow/query/134031/… – Ben Crowell Sep 1 '13 at 14:00

Ill speak for myself, thought I wouldn't call my self non-novice or anything else. I am among the people thinking of leaving the site after being here for about a year. The reason is It stopped being interesting.

While I rarely posted, I used to enjoy reading answers, but off-late nothing in the site seems to interest me(I am student of theoretical physics). I do find good question/answers occasionally(often by the same set of 3-4 people), which is why I continue to check the site and sometimes the meta.

The site is nothing like what it was when I joined a year ago. Perhaps it is a personal view, but it might be shared by a few other people. I would be very much interested in a dedicated TP site. I also have no idea when and why Dilaton got banned, but that was going to happen eventually I guess. There are discussions linked in Dilaton's page about starting a new site, I would definitely be curious to see how it turns up.

• Out of curiosity, what is it about a dedicated TP site (whatever that means) that you think would interest you, that this site doesn't currently have? – David Z Sep 1 '13 at 16:44
• To perhaps be more direct than @DavidZ -- Rather than try to find and start a new community, one could encourage the people who would join it to join this one and it would be better for all involved. – tpg2114 Sep 1 '13 at 19:09
• An (highly subjective) observation on a dedicated TP site: it has already happened once. A non-trivial number of our theorists didn't want to use a site with many novices and formed TheoreticalPhysics.SE and went off to play in their walled garden. When TP.SE petered out for lack of activity and was merged back here many of those users did not return, which seemed strange to me. If I sound a little bitter about this it is because I am: they were warned about overly fragmenting the user base but they didn't want to include us greasy handed types (or even the phenomenologists). – dmckee Sep 2 '13 at 2:13
• A few also quit over Ron's suspension, and I guess I see their point, but then they didn't have to keep cleaning up the man's messes. – dmckee Sep 2 '13 at 2:16
• @dmckee: The ongoing discussion at the TRF is mostly not dedicated just to TP but to All sorts of Cuttting-edge research/BSM Physics. – Abhimanyu Pallavi Sudhir Sep 2 '13 at 5:21
• @DavidZ By TP I mean the roughly level of Landau's books or Susskind's online lectures. I don't say that this site has no TP but the focus of the site has definitely changed due to a change in its user base. – Prathyush Sep 2 '13 at 10:50
• From what little I have seen of this website, I think non-novices generally have better 'offline' resources at their disposal so they have less incentive to ask a question online and wait for an answer. If they do come it will mostly be to answer others' questions. The fact is that a website like this caters much better to the askers who are not well established. You non-novice guys really don't ask too many questions, why should you anyway since you are non-novice. – dj_mummy Sep 2 '13 at 14:37
• @Prathyush Ah, well every once in a while someone complains that we don't exclude questions based on level, but one of the founding principles of the site is that we are not going to do that here. – David Z Sep 2 '13 at 17:36
• @dj_mummy: Interesting. I'm a non-novice; I have a PhD and teach physics for a living at a community college. However, I have zero access to journals except by driving to the nearest four-year school, where I can photocopy them. It seems to me that the internet is much richer in resources for beginners. Of course at all levels, many of us end up posting questions that amount to, "I'm confused about..." or "How do I..." rather than just being requests for information. – Ben Crowell Sep 2 '13 at 18:19
• @dj_mummy: But no matter what, as a teacher you will have far far more exposure and resources at hand. I think we're comparing apples and oranges. As a non-novice in physics, I have good access to recent research that's of interest to me at my level, because the recent stuff is on arxiv. What I have poor access to is anything that's of interest to me at my level and older than ca. 1995 -- that stuff is all paywalled by the evil bloodsuckers at places like Elsevier, even though tax money paid for it. – Ben Crowell Sep 2 '13 at 20:16
• @DavidZ i never said exclude questions, I said ORGANIZE it properly. Make it EASY to use. The main page is a MESS. I repeated that several times in the meta, And im tired of saying it. I will use the site until the few people whose answers I enjoy reading are still there. You have already made your choices by being inactive. This site is no longer conducive for research. – Prathyush Sep 2 '13 at 20:43
• @Prathyush so are you using "dedicated TP site" to mean something other than a site that only allows TP? – David Z Sep 2 '13 at 20:47
• Sure, perhaps you're looking for something other than what the Stack Exchange model was meant to provide. (But bear in mind that just because the site doesn't give you what you want, doesn't mean it's useless to everybody, or even to other researchers.) – David Z Sep 2 '13 at 20:54
• @DavidZ No doubt this site has been very useful to me. which is why I am reluctant to leave, and I might not. But still I again insist that something must be done to organize the site, neatly. – Prathyush Sep 3 '13 at 20:29
• Prathyush, the Stack Exchange engine doesn't really have any concept or "neatly organized" beyond the tagging mechanism. It does have is several tools to tune your view. Set favorite and ignored tags and browse the front page rather than the question view; this makes a big difference for me on Stack Overflow (I don't use it on physics because I'm a mod). Consider using the RSS feeds. There is a Stack Exchange wide "filters" mechanism that I never checked out. The whole the site is unlikely to become "organized" but you can see it in a more personalized fashion. Or take it to the mother meta. – dmckee Sep 7 '13 at 18:25