10
$\begingroup$

As I was browsing through Area 51, I managed to look at this site's statistics, and to be honest they looked a mite disturbing. Apparently there are only 2 questions per day and 31 avid users, numbers which seem astonishingly low compared to successful Stack Exchange betas.

On the other hand, I know certain Stack Exchanges move at different paces; Home Improvement, for example, only had about 4 questions per day but was deemed to be in good health and managed to graduate, on the strength of its traffic. Another good example would be the TCS site, which only produces about 7.5 questions per day, but the sense I have is that they have a very strong sense of community.

So ultimately it comes down to asking the users here how well they think things are going: Are you getting your questions answered? Is a vibrant community forming?

From my perspective, things seem quite dead. I noticed for example, in checking the badge pane, that the last "nice answer" badge was awarded sometime in mid-February. And, apparently, if I'm reading this correctly, it would only take 5-up votes for a question-answerer to be vaulted into the top 3 users by rep for the month. Moderators for the site don't seem to be present on some days.

Are my impressions accurate? In particular, I would appreciate an answer from Mr. Robert Cartaino, as I understand as a community manager he keeps track of this sort of thing on the meta sites.

$\endgroup$
9
$\begingroup$

To me, the scope/overlap of this site limits the activity level. If you have a statistical question that is appropriate here, isn't it likely that the same question is appropriate at stats.stackexchange.com? If so, where will it be asked? The same logic applies to a programming question. Will it be asked here or at stackoverflow.com?

Next is the issue of professional versus retail trades. Most of us trade both. Professional in one specific area (for instance, equity trading) and retail in another (for instance, oil trades). Some issues apply to both areas, but in order to supply detailed answers that are consistent with our area of expertise, it's easy to back off from fringe questions.

As far as questions go, it's really hard to come up with something that's specialized enough to fit here, but not too specialized that only a handful of people could possibly answer. For instance, High Frequency Trading questions. How many people here are in a position to actually trade in that format? The questions are interesting, and I am interested in the answers, but is it likely that someone here is in a position to provide a real world answer? Several questions were asked, but nothing really gelled.

I don't have the answers, but I would give the site extra advertising and time to get established.

$\endgroup$
  • 4
    $\begingroup$ The overlap with other SEs is something I've been thinking about as well. Most of what we do is programming and statistics, which are covered by bigger sites. And a lot of what isn't covered by those sites is going to be "secret sauce". That means there's a very narrow set of questions we could possibly have, especially when limited to professional interests only. $\endgroup$ – chrisaycock Mar 10 '11 at 3:56
8
$\begingroup$

It's certainly a valid concern. I think one thing this site really needs at this point is evangelism. We've been promoted on QuantNet and Quant.ly, plus a few non-quant sites. We'll need to keep that up to spread the word.

Of course, one of the keys is that we also need great questions. That's important so that visitors to this site immediately understand what we're about and what we can offer.

$\endgroup$
  • $\begingroup$ Notice on the Quant.ly site that the most recent post was 2 days ago, similar to this site. $\endgroup$ – bill_080 Mar 9 '11 at 19:29
  • $\begingroup$ @bill_080 The most recent post was a few hours ago. Quant.ly uses the Hacker News engine to promote stories. But yes, I agree with your sentiment that quant sites in general are very niche and thus have difficult in getting traction. $\endgroup$ – chrisaycock Mar 9 '11 at 19:34
  • 1
    $\begingroup$ IMO, In order to get good questions, we (the most active members of the site) should start asking questions (even if we know the answer or have a good idea...). $\endgroup$ – Shane Mar 18 '11 at 1:53
  • $\begingroup$ I believe the site is great, the technical infrastructure is a lot better that wilmott wiki for instance. However you may want to extend the audience beyond quant discussions only. General discussions on market finance / risk management would make sense I believe. $\endgroup$ – fabien May 1 '11 at 21:13
  • $\begingroup$ And also you may advertise thanks to CFA / GARP etc ... (there are groups on linked in) $\endgroup$ – fabien May 1 '11 at 21:15
  • $\begingroup$ @fabien This has been discussed before; see this question, for example. We are not going to expand into general finance since there is almost no overlap between that and quant finance. This site was specifically created for quants. If we allowed non-quants on here, it would drive away the quants. A moderator, I have to ensure that the questions on here are interesting to an expert in the field. $\endgroup$ – chrisaycock May 1 '11 at 23:06
  • $\begingroup$ oki, makes sense $\endgroup$ – fabien May 1 '11 at 23:09
6
$\begingroup$

Commenting on chrisaycock's answer - I think that at the moment we not only need great questions. What we need foremost is more questions (current stats are 1.4 per day and dropping). And great questions are not easy to come up with. Thus, where do we draw the line between quantity and quality?

At least that's my impression that the discussion level can be intimidating a bit. Both at asking and answering level. Which is certainly good in the long term (vide MathOverflow) for an established community, but will it not harm the site during the beta stage? Judging by myself, with so many experts in the field around here it's not easy to go from a passive reader to an active participant lest you cover trivial matters or give erroneous answers.

And I suppose we need way more traction for the site to succeed. But probably not at the cost of quality as Owe mentioned. It would be a pity though if this could impact the chances of survival of this site.

$\endgroup$
3
$\begingroup$

The stat that I find most worrying is that there aren't a lot of accepted answers (I say stat, but I don't know the number, just that on the home page I don't see a lot of green for accepted answers). It seems that a lot of questioners either never return or think that the community's answers are wrong. Maybe the latter comes from questioners with a strong prior on what they want the answer to be or too high of expectations about answers the community can provide to broad questions?

I feel like a higher acceptance rate would incentivize more/better answers and a stronger community. But without more repeat questioners, it's really hard to say "Hey, dude, please accept some answers, then we'll tackle your question".

The knock-on effect is that during the community-building phase there are fewer knowledgeable users and the noise level is a lot higher (i.e., there seem to be a lot of "I don't know language X. How can I write an X code to buy low and sell high?").

What have other SO sites done to keep the knowledgeable users coming back? Can/should we go through and accept the most valuable answers for stagnant questions? This seems to undermine what I consider the value of SO sites (and I don't think the knowledgeable users want points, but I think they're like to know that their answer is read and considered). I feel like relative to the other SO sites there may also be the complication that a lot of casual visitors expect to learn how to trade like James Simon instead of having a deeper curiosity.

$\endgroup$
1
$\begingroup$

I think recently the level of the questions was just abysmal - things like "what is persistence" or "how much data do I need for GARCH(1,1)". In this regard I think we need better policing, and maybe sometimes a tap on the fingers of repeat offenders.

$\endgroup$

You must log in to answer this question.

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