4
$\begingroup$

A hypothesis:

The spectrum of non-physical/-technical topics/proposals via AREA51 is exponentially growing. While this is a minor problem for insulated communities/topics, like programming/cstheory, popular scientific topics like phyiscs/philsophy attract alot hobby-interested laymen (which is good!). BUT, assuming a similar growth rate (probably mainly determined by social sharing functions here and time scales of a year, not montly meta discussion of fluctuating quality here!) for the small number of graduated experts or physics students here and the the much more interested laymen, the signal/noise ratio has to decrease with time, as a bigger number grows exp. faster than a smaller :( simple math.(probably the growth rate for new physics experts on SE is quite lower than average user if you look on progress of theophysics proposal)

With a 1 user - 1 upvote system IMHO this growth will induce serious signal/noise ratio problems here (lets face it, it gets like yahoo answers, digg, most other Q&A sites). Of course one can still regulate and close bad and off-topic questions, but good modern questions attracting experts will just go under in lawmen votes. Not every laymen can understand the implications of a top-notch question, if you explain it literally, the question gets so long, that no one will read it, so will always gain only a small number of expters upvotes. If you look on ArXiv, there are services like scirate.com to overcome this signal/noise ratio problem or citeulike.com. Rating of mainly experts or following expters. Or peer review.

Do you agree with this hypothesis and its consequences? Where are flaws and fallacies? I dont care in this question about the crucial point, if SE owner want a fast growth or higher quality than on other Q&A sites. Thats anyway a matter of their personal compromise we cant argue about. The question here is what minimum level of signal you can establish to keep experts (which make this a more interesting place than other Q&A sites) and what tools/options you give them to filter growing noise. Experts with a phd/master/diploma dont care about their rep count (alot laymen seem imho really addicted/biased to this game). They just leave when it gets a waste of time or you have to give them a insulated proposal like theophysics, cstheory or make migration places like http://area51.stackexchange.com/proposals/4955/popular-natural-science for laymen.

It just looks to me that instead of solving this dilemma artificially by side-proposals it could be solved with SE own rating tools adapted to current state and topic of a community much easier, faster, better and more efficient. Pure democratic 1 user 1 vote works fine with only physicists/programmers(like stackoverflow) or people electing the political direction of a government. But for quality of a topic where serious long education/background knowledge is required & self-teaching doesnt work, it just has to fail. This gap cannot be filled by current set up of SE voting/moderation, esp. with growing laymen number in the long run...

Is there no technical option to implement something like user reputation > 1000 = 2 upvotes per upvoted question, > 10000 = 3 upvotes per upvoted question? Voting strength for answers and downvoting questions keeps 1 for all users. Due to the commitment system you have much more experts here than on other Q&A site, but with ongoing exp. growth its my strong belief you will loose them in the long run on popular scientific SE sites.

Dont worry, i will not hassle any more than this single time, cause it will basically define, if scientific proposals for experts&lawmen on SE have a future for me. Personally i prefer this to experts only proposals, as there is more percolation of topics & users and less divas.

$\endgroup$
  • $\begingroup$ you probably mean "laymen" for "lawmen" $\endgroup$ – Mark Eichenlaub Sep 1 '11 at 1:33
  • $\begingroup$ @Mark thx, ashamed :), i blame it on my mini smartphone keyboard and display... $\endgroup$ – Werner Schmitt Sep 1 '11 at 10:59
  • 1
    $\begingroup$ Related from the mother meta: Should the upvote or downvote by a higher-reputation programmer count more than the same by a lower-reputation programmer?, Weighted Down-Voting based on Reputation, Display a second, alternative, weighted vote count... Variation of this idea have been floating around meta.SO for a long time and the team has never taken action. $\endgroup$ – dmckee Sep 1 '11 at 16:26
  • $\begingroup$ what is the actual question here? I'm unclear $\endgroup$ – Jeff Atwood Sep 2 '11 at 3:34
  • $\begingroup$ @dmckee thx for links, but im not asking for reinvention of voting system like these questions, i only want increased question upvote strength for high reps on physics.SE and similar sites which face a MUCH larger amount of interested laymen than SO. I dont agree with the links, that SO needs another voting/rep system, as it is and STAYS a insulated community of semi- and prof. programmers. $\endgroup$ – Werner Schmitt Sep 2 '11 at 11:08
  • $\begingroup$ @jeff probably you are one of few who can actually answer this: Is there a technical option to increase this upvote strength ON A SINGLE SE SITE? If so, do you think about implenting it because diff. sites face diff. signal/noise ratio problems due to their topic as explained above? I hope we agree that in 1-2 years this place will have a exp. grown ratio of laymen/ educated physicists/engineers and you maybe loose a lot of them as highlighting of interesting questions doesnt work for them anymore with current pure democratic system. Im asking for adaption, not reinvention... $\endgroup$ – Werner Schmitt Sep 2 '11 at 11:19
  • 2
    $\begingroup$ Another problem for all the new stackoverflow sites is that the questioner decides which answer to accept. When by definition the questioner is least qualified to know which answer is correct! For a straight forward software question on SO the questioner can test the code and see if it works for them - but they can't do that on physics. $\endgroup$ – Martin Beckett Sep 15 '11 at 15:29
  • $\begingroup$ @martin i would be fine with this, if they would change upvote weight for high rep. But as no one gives a answer it seems not technical possible to adapt this on single sites. The "SE philosophy" is to vote Q&A you like. Well, for research or by laymen highly populated science sites this works not well enough to keep experts, on SO fine. So probably the roadmap is to propose a expert.SE for every scientific topic to gather and keep experts (bio, chem, psych....) The Q/A quality on these topics will be very low, on philo/astro it already is as to much programmer laymen there, but no students $\endgroup$ – Werner Schmitt Sep 23 '11 at 15:58

You must log in to answer this question.

Browse other questions tagged .