10
$\begingroup$

I signed the petition and committed to participating in this website. This pledge accounts for some of my involvement. However, it seems clear from tracking the site statistics that Physics-Stack-Exchange is ill. (A healthy sign would be exponential, or at least super-linear, growth.) Why?

My guesses:

  1. Advertisement.
  2. Quality.
  3. Nervousness.

--

  1. How many working physicists know about the site? Probably not many, and I believe the answer is tied in with #2.
  2. I understand that the community disfavors simply research-level questions. However, without high-quality answers the site rambles through low-level questions with a high noise-to-signal ratio. This state of affairs will not attract the attention of those who who can make the site better. It's fun to ask about fans and cars and planes and sails, but modeling these things seriously to reveal deeper phenomena is formidable. Instead, we get a mostly qualitative, hand-wavy argument about why things work. That has value, but it is not academic physics. (Maybe that's the goal? If so, I [personally] would discontinue using the site after fulfilling my start-up commitment.)
  3. I think some physicists do know about the site but are unwilling to reveal their uncertainties and insecurities. This may be a physics-culture thing or it may not. There is some similar phenonmenon in Math-Overflow, but at least postdocs seem very open to the Web exchange.

Solutions? I don't know. You?

$\endgroup$
  • 5
    $\begingroup$ Re 3: Noone forces you to reveal your identity, I've seen high-rep users at SO with only nicknames $\endgroup$ – Tobias Kienzler Dec 8 '10 at 10:50
  • 1
    $\begingroup$ Tobias is correct. It is very common, omnipresent even, to use nick names on SE sites, on the web in general even. $\endgroup$ – Noldorin Dec 8 '10 at 18:50
  • 2
    $\begingroup$ @Eric: any example of high quality question you expected? $\endgroup$ – unsym Dec 10 '10 at 10:33
  • $\begingroup$ I didn't see this question, and so I sort of asked my own. The only difference in questions is that I am curious about what happens after the beta period is up if we are not in the "healthy" ranges statistically? will SE close this down? Would you mind adding this to your question, or do you feel the issues are separate?(I will delete my question so as to avoid duplicates) $\endgroup$ – Sean Tilson Dec 11 '10 at 3:04
  • $\begingroup$ I mentioned the site and its "lack of traffic" to a physics grad student friend of mine, and he mentioned that it is probably because there is another more well established forum. Is this the case? should we be "advertising" on that forum? Also, is there any reason why we shouldn't encourage peope committed to TPSE to participate over here? Is the main reason for that site the low level of questions here? $\endgroup$ – Sean Tilson Dec 11 '10 at 3:24
  • $\begingroup$ @Sean (last comment): I assume you're talking about Physics Forums, and yes they are probably keeping some traffic away from us, but I think we should be able to slowly convince people that the SE system is better. Regardless, PF does not allow mention of competing sites, so advertising is not an option. The TPSE people want a higher level of rigor in the discussion (answers, mainly) than we have here, so they probably won't come back, although we could try to work on them. $\endgroup$ – David Z Dec 11 '10 at 8:59
  • $\begingroup$ @Sean: There is no need to care about whether it is close down. The most significant indication of the audience is that it should have a steady number of questions/day. People come here either asking questions, answering questions or finding interesting question and useful answers. No new questions implies that no one come here $\endgroup$ – unsym Dec 11 '10 at 9:06
  • $\begingroup$ @David: Asking homework is very helpful in the PF because of their site policy and the keen helper. It does attract a vast amount of students there. And admitted, homework asking is not good here because of the structure restrict easier communication with the asker. However, the format here do allow a easy edit of answer to give a high quality answer. $\endgroup$ – unsym Dec 11 '10 at 9:11
  • $\begingroup$ @David well then we should have higher standards in the types of answers that are given. I recall a meta question in the early days about what level of questions should be asked and the consensus seemed to be all levels were ok. Now, I think that if we want to attract mroe traffic then we should at least have higher standards about how well questions are asked. MO has a policy like this and it seems to work well. They close down have baked questions until they are ready. Basically I think we need more hard-line moderators! $\endgroup$ – Sean Tilson Dec 11 '10 at 21:44
  • 1
    $\begingroup$ @Sean, yes I had asked that earlier question, too, and I posed the current one in order to revisit the issue. $\endgroup$ – Eric Zaslow Dec 13 '10 at 2:09
  • $\begingroup$ @hwlau: An example? Something like, "What do the Fierz identities mean?" "How can we measure magnetic moments so accurately?" "Was there really a problem with Einstein's perihelion prediction?" "What is a quick derivation of the relation between the sign of the cosmological constant to the sign of the energy of the vacuum?" Anything that's not, "Why does my soup taste better at the start of the meal?" (though any physicist will admit to some interest in such things). $\endgroup$ – Eric Zaslow Dec 13 '10 at 2:15
  • $\begingroup$ @Eric: if it's really some physical phenomenon that makes soup taste better at the beginning of a meal, then I'd actually love to have that question on the site. We spend so much time thinking about abstract concepts and exotic situations that it's nice to see them connect to everyday life sometimes. But then again, I can agree we'd want to limit those kinds of questions so that they don't form a majority of the site. $\endgroup$ – David Z Dec 15 '10 at 3:55
  • $\begingroup$ I'm thinking: in light of blog.stackoverflow.com/2010/10/when-will-my-site-graduate, we're not doing that bad, but we should put some serious effort into site promotion. $\endgroup$ – David Z Dec 15 '10 at 4:19
  • 2
    $\begingroup$ Nice link, @David. And speaking of promotion, it probably won't hurt to directly link this: blog.stackoverflow.com/2010/08/a-recipe-to-promote-your-site $\endgroup$ – Marek Dec 17 '10 at 16:38
  • $\begingroup$ Let me also throw in a link to our chat session planning question. $\endgroup$ – David Z Dec 17 '10 at 20:28
7
$\begingroup$

Are you really sure that the state of affairs is so bad? What did you compare the site statistics with? Looking at other SE proposals it seems to me that this site is doing quite well for a site 30 days in beta.

Sure, it will take some time to attract more knowledgeable people and maybe higher standards for questions should be probably encouraged (more onto this later), but I think on the whole, the site is fine. But if people like you (who are absolute top on this site, in my opinion) start to leave then the site obviously has a big problem.

Now, back to the level of questions. I think we have decided that the level should be higher. But apparently there are still lots of (or even most?) popular and high school level questions and we are doing nothing to stop that (I am not sure whether that's good or bad, just stating the obvious). But I guess if we would start to drastically close those questions many people would leave and although standards would be higher, there would be just few good physicists left and the site would collapse.

I think a better approach (apart from starting a research level site like TP.SE) is to let these questions stay, let lots of people come to ask them and make the site popular. Because then also good physicist will discover the site (at least I believe so) and the level will slowly rise (hopefully). I honestly don't see any other way than this. In particular, I am not quite sure how to spread the word in a way that would attract established physicists.

$\endgroup$
  • $\begingroup$ +1 for your last paragraph especially. I wonder, is there some way that those of us who are most interested in promoting the site could have a brainstorming session for publicity ideas? It might be something to use the chat room for. If nothing else we could get a sense of how many people are really interested in promotion. $\endgroup$ – David Z Dec 7 '10 at 22:11
  • $\begingroup$ @David, that session is a good idea. We could probably start a question about possible means for that (IRC, IM, whatever...) and let people vote. The chat here would also be fine for this but I guess it's not available on beta sites? Which is strange because at mature sites (e.g. at math.SE) this feature is not used at all and precisely beta stage is where the chat could be the most useful. $\endgroup$ – Marek Dec 7 '10 at 22:26
  • $\begingroup$ @David: if a chat room has no activity for too long, it will be frozen or deleted $\endgroup$ – kennytm Dec 8 '10 at 6:25
  • 2
    $\begingroup$ I think the best way to raise the level of the questions is to ask high level questions. Obviously this is easier said than done. It is curious though that MSE was able to be so much more successful without having a restriction on the level of questions, but maybe that is because there were a bunch of people waiting to use it. $\endgroup$ – Sean Tilson Dec 11 '10 at 3:19
4
$\begingroup$

Dear Eric,

I want to give a little comment on the second part, the quality discussion.
Actually I think that it is no problem to have different kinds of quality on the page as long reputation does not play a major role. Answers and questions of different levels have different work load. You can get a lot of reputation by just answering simple questions and you get almost nothing with stuff that only a minority can actually understand.

But this situation is not so uncommon. It is exactly the difference between fundamental and applied research. You simply cannot compare publishing behaviour of those two groups. So one might ask the question: If this site is ill, what will happen to TP.SE? Or are simply our definitions of "illness of a beta" not comparable between different areas?

Whatever will happen, I wish everybody a nice third advent :)

Robert

$\endgroup$
  • $\begingroup$ +1. Btw, regarding "Whatever will happen, I wish everybody a nice third advent :)" -> I love how you always have this positive attitude. It always cheers me up :-) $\endgroup$ – Marek Dec 13 '10 at 17:23
  • $\begingroup$ @Marek: Thank you very much. You know, what goes around, comes around :) $\endgroup$ – Robert Filter Dec 13 '10 at 20:11
4
$\begingroup$

My 2 cents.

1) Broad advertisement would be a bad idea right now because most professional physicists would take a look or two and not come back because of the low level of questions and answers. It would make sense to selectively recruit a few sympathetic people, perhaps some of the more physically inclined people from MO and anyone else who you think would raise the standards.

2) The quality of questions varies, but what disturbs me more is the low quality of some of the answers, many of which get lots of up votes in spite of being vague, misguided and sometimes plain wrong. I'd love to vote some of them down but don't have enough rep to do so.

3) There are a few legitimate reasons for anonymity, but by and large the reputation of the site will be higher the smaller the number of anonymous users. If you are going to put some effort into asking and answering good questions why not let the world know who you are? It might even help you get a job some day.

$\endgroup$
  • $\begingroup$ Whatever advertisement we do would be targeting students, generally undergrad-level and early grad level, not professional physicists (unless we can raise the level of questions on the site somehow). Anyway, until you get enough rep to downvote, you can still leave a comment on wrong answers explaining why they're wrong and hopefully others will see it and act accordingly. $\endgroup$ – David Z Dec 18 '10 at 5:28
  • 1
    $\begingroup$ P.S. The more good questions and answers you post, the easier it'll be for us to give you enough rep to downvote when you feel it's necessary ;-) $\endgroup$ – David Z Dec 18 '10 at 5:36
  • $\begingroup$ My interpretation of the discussion on advertising, which i have been away from, did not seem like it was aiming itself at undergrad and early grad students. Also, is that the best way to raise the level of the site? but that is a discussion for elsewhere, which i will look for. $\endgroup$ – Sean Tilson Dec 21 '10 at 3:51
3
$\begingroup$

I added this site as a proposed URL into Google and Bing to increase its level of visibility. However, a proper way to do it (and one much more effective) would need administration privileges. Therefore, I encourage the administration to take those steps (if they haven't already). This concern arises from the fact that few searches return this site in a decent ranking (e.g. physics forum, physics Q&A, etc) which is somewhat worrying.

$\endgroup$
  • $\begingroup$ Huh, I thought Google et al would pick up on the site automatically, and that the only thing we could do to increase visibility would be to get linked from many other high-profile sites. Anyway, you should definitely bring this up at our chat session which looks like it'll be on Sunday. $\endgroup$ – David Z Dec 17 '10 at 21:24
  • $\begingroup$ Google ranks this site already searching for something very specific, for example, google.com/… or a title for some questions. Nonetheless, that doesn't help very much when it comes to produce new traffic from search engines. This issue realy makes very difficult to comply with the fifth requirement "Eventually, 90% of a site's traffic should come from search engines". I'd be happy to discuss this in the chat session, although I'm not sure I will be able to participate at that time. $\endgroup$ – Robert Smith Dec 17 '10 at 22:54
  • $\begingroup$ By the way, in order to obtain a fruitful change (in case it has not been put in place yet), we need people from the administration in the chat session. $\endgroup$ – Robert Smith Dec 17 '10 at 23:02
1
$\begingroup$

Currently, there is certainly a tiny amount of working physicist register here. Most of the registered user should be students. I guess there is only a quarter of them are graduate students and post-doc, maybe the same set of those 100 people committed in the TP.SE.

A rough estimate of all working physicist should be around 50k (see number of physics author and the employment). It is obvious that there is much room for improvements while it is comparing with the 800 people registered here. The grow of the registered user seems to be slowed down, so some form of advertisement may help this situation.

In regard of the quality of questions, it should be fine if part of them are easier than the research level questions. Sometimes askers do want a deep answer for an easy question. You also need new questions to attract people coming back. It should be noticed that more questions itself is a good way to advertise because you can easily found the website in the google. I know the SO simply because it often appears in my search of programming topics.

For high quality questions, it is likely to be asked by knowledgeable people in the particular field. Hence, the number of high quality questions is always much less than the number of low quality questions. So, why don't you try to ask difficult questions here to attract them. It may surprise you that there is person that can answer those questions. Those questions may need weeks of reading and consulting your supervisor.

There are no single good ways for physicists to ask questions in the web and this site has a good system. So, hopefully, it should attract them slowly if we can keep the current quality.

Edit: Regarding your last question, I have been surprised to see Peter Shor (shor algorithm of prime factorization) participating the discussion in the TCS.SE.

Edit 2: It also seems that people rarely asking questions about papers, which should be count toward research level question.

$\endgroup$
  • 1
    $\begingroup$ Regarding Edit 2: I'd ask some questions about papers but I have a feeling that currently there is no one to answer them (or they don't care... which is the same). Actually, it's not just a feeling, no one answered my question about Ising model. And note that my question was definitely below research level (or at least I believe so). So I don't think it would be any good asking too advanced questions right now. Maybe just from time to time. But I'd like most of the questions to be undergraduate and graduate level so that there is a reasonable amount of people to provide answers. $\endgroup$ – Marek Dec 10 '10 at 18:27
  • $\begingroup$ I understand that there is no one who can currently answer your questions, but you should ask them all the same. Right now, the stats for the site indicate that "we are in need" of more questions (as well as more overall users). That is one reason, but what if someone is visiting the site and dismisses it because they don't see in any interesting high level questions? It seems to sound like that is one of the reasons people think there are not a lot of working physicists on the site. So please, because we all know that if you ask it, they will come. $\endgroup$ – Sean Tilson Dec 11 '10 at 3:23
  • $\begingroup$ @Marek: I would consider your question at least research level because you ask people to compare different methods. There is no need (of course you can read it yourself) to know more than one onsager solution if they are not working on this field. $\endgroup$ – unsym Dec 11 '10 at 8:49
  • $\begingroup$ I have noticed that there is a small audience group in this site. It is the group of senior undergraduate studying physics that may ask technical questions that can be interesting sometimes. This is very obvious when you compared this site with the SO. Maybe we can make a poll here for this information. $\endgroup$ – unsym Dec 11 '10 at 9:00
  • $\begingroup$ @Sean: fair enough. It's true that right know the statistics is about 98% answered questions so it wouldn't hurt to lower that to even 90% and also have lots of great and hard unanswered questions that people can find later. $\endgroup$ – Marek Dec 11 '10 at 11:45
  • $\begingroup$ @hwlau: hm, I have a different idea about research as in working on things no one has worked on before (or at least you don't know about that). My question just asked for comparison that I am pretty sure lots of people who work in the field should be pretty familiar with. $\endgroup$ – Marek Dec 11 '10 at 11:47
  • $\begingroup$ @hwlau: yeah, poll might be a nice idea. But I wonder how many people would answer (because there aren't that many people visiting meta) and whether it would give us relevant statistics. $\endgroup$ – Marek Dec 11 '10 at 11:48
  • $\begingroup$ @Marek: Actually, I have the same idea of research as you. However, the real world does not work like this. The difficult thing being worked by you now may have been worked out by others or your supervisor before. It might be the details and the different approach you done so you can publish a paper. There are also easy thing that no one work before, especially the start of a new research field, but those count as research in many people's opinion. $\endgroup$ – unsym Dec 11 '10 at 13:05

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .