12
\$\begingroup\$

Would it be possible for Code Review to award bounties on good questions that have remained unanswered for more than 6 months? When I was working through zombies I found a few questions I couldn't answer but were good questions. Starting with the questions that received the most up votes and working down to questions with 5 or more up votes?

Examples:

In cases where the user is no longer active on CR, moderators could accept the answer.

Perhaps for high ranking members who have unanswered questions we could suggest they post bounties on their unanswered questions.

The point is that normal users generally won't see these questions and it might be good to bring them back into focus, so that someone can try to answer.

\$\endgroup\$
  • 3
    \$\begingroup\$ Letting moderators do the accepting might be a bridge too far. Bounties have been done before with variable successes, so that might just work enough. \$\endgroup\$ – Mast Jul 24 at 14:16
  • 2
    \$\begingroup\$ Another option might be bringing back RoboSanta or something similar, pointing out questions deserving attention for one reason or another. \$\endgroup\$ – Mast Jul 24 at 14:17
  • 1
    \$\begingroup\$ Bounties aren't always great. Sometimes nobody has anything to say, and sometimes nobody knows the language well enough to say anything. \$\endgroup\$ – 23fc9a62-56de-47fb-97b4-737890 Jul 24 at 14:27
  • 1
    \$\begingroup\$ What about letting the community decide with a vote on zombies.. or is there too much overhead? \$\endgroup\$ – dfhwze Jul 24 at 14:31
  • \$\begingroup\$ “moderators could accept the answer” – all these questions have no answer. Am I misunderstanding something? \$\endgroup\$ – Martin R Jul 24 at 14:32
  • 2
    \$\begingroup\$ In cases where the user is no longer active on CR, moderators could accept the answer - moderators do not have this power AFAIK. \$\endgroup\$ – Simon Forsberg Jul 24 at 14:42
  • 1
    \$\begingroup\$ Either way, for a question to be considered as answered, all it needs is an answer with a positive score. I think it's really important to think of OP's specific concerns when answering though, just to avoid cases when OP asks "Can I improve the performance?" and he gets an answer saying "You should rename your variables to make them more readable". \$\endgroup\$ – Simon Forsberg Jul 24 at 14:43
  • 9
    \$\begingroup\$ I have, in the past, and will continue to offer a bounty from my own rep on any reasonable question that needs attention if someone pongs me about it, or something. \$\endgroup\$ – rolfl Jul 24 at 15:26
  • 2
    \$\begingroup\$ @rolfl I have considered that in the past. I have just blown 300 rep points for 2 of my own questions that never received answers in order to put up or shut up. \$\endgroup\$ – pacmaninbw Jul 24 at 15:30
  • 1
    \$\begingroup\$ I would typically put the minimum 50 rep bounty just to put it in to the bounty page, and get the additional views, @pacmaninbw . \$\endgroup\$ – rolfl Jul 24 at 15:34
  • 1
    \$\begingroup\$ @rolfl my questions are rather long and involved. They got a fair number of votes to begin with but no answers. \$\endgroup\$ – pacmaninbw Jul 24 at 15:37
  • \$\begingroup\$ @pacmaninbw I hope those 300 rep points didn't go to waste :p \$\endgroup\$ – dfhwze Aug 3 at 19:36
14
\$\begingroup\$

I think this would require a large cultural shift to get others to do this too. If you want to post a bounty then feel free to do so. I post bounties for questions every now and then, and I've seen others do the same. But these aren't at the magnitude you probably would want from this cause.


But are bounties really the best option?

What is the cause for all these zombies, and why aren't we focused on fixing that? To me it looks like this is an XY problem. I also feel focusing on zombie cleanup rather than prevention could promote the creation of new zombies. People have a limited amount of time and if they're sniping all the 6 month old zombies who will be answering the new questions?

Personally I don't answer zombies, not because they're bad. Some of them probably would be more fulfilling to answer, but because I'm incentivized by votes. What is the incentive for me to spend hours laboriously answering a question that no-one's going to see? Contrast this to sniping a low-hanging-fruit new post that people are going to up vote and get to HNQ.
I get so many of those up votes that I want for a fraction of the work. It's a clear no-brainer.

And so the only solution to get me to answer zombies seems to be only implementable by Stack Exchange. And I doubt that'll happen. (Say a hot zombie list)


If you want to nuke 1000 of the 6000 zombies we have you can do that by casting a single vote on a lot of answers. This only goes to prove that zombie farming is an unrewarding endeavor.


What's the source of all these zombies? A quick glance at one of the graphs in rolfl's recent post should be all we need to know what the biggest problem is .

enter image description here

Before I look at Python, I'm going to look at some of the other tags. Take C# for example, here we can see that a lot of the top answerers have only a couple of handfuls of answers. To be in the top 10 for the past year you only need 20 answers. Given that the zombie count has gone up by 100 that means if we had 5 more benj2240's then it would have stayed in place. Alternately if dfhwze had joined in January then the zombie count wouldn't have stood a chance. And if either Henrik Hansen or dfhwze stop contributing the zombie count is likely to drastically increase.

enter image description here

If we look at JavaScript we see similar situation as before, but it's even more extreme. The third highest answerer has 33 answers this year, with the tenth having 18 answers. This shows that JavaScript predominantly has low answer count users. Whilst the zombie horde is being held back by Blindman67 and Sᴀᴍ Onᴇᴌᴀ.

enter image description here

And so I think C# has fairly healthy user contributions, but JavaScript looks like it's very susceptible to being hit by a bus.

When making these graphs I half expected the Python one to look like the JavaScript graph, with Graipher and Mathias being at the top. However this isn't the case and shows what I think is healthy tag activity. If any of the top ten are hit by a bus then the answers will continue.

It's nice to see that the 10th most active user, l0b0, has provided more answers than the average of the top ten of C#, JavaScript and Java. Where C++ has a nice amount of active answerers, but also has a high bus factor.

Python also gained two new active contributors that got into the top 10 being Reinderien and AlexV.

enter image description here

What's the problem

I don't think I'm a highly active answerer on Code Review in general, even more so this year. But I've contributed an average of 2 answers a week. Which would have been enough to hold back the zombies this year from any of the tags, except Python.

And so I think there's an active answerer shortage. I hope the graph showing so much Python activity, but still tons of zombies poping up is enough to prove that.

Taking another look at rolfl's post we can see some general stats for Code Review. Looking at the active users line we can see we've had the lowest amount of active users since just before 2014.

enter image description here

I don't know how to increase site activity. But I think the focus should be to get more answerers for now to stop the increase of zombies.

Later to help the number of zombies to go down we should incentivize answering them. The only non-rep exhaustive solution I can think of is a chat room/meta post/SEDE query that highlights new zombie kills. This should help reduce the amount of zombies that have answers but no upvoted ones too.

\$\endgroup\$
  • 7
    \$\begingroup\$ Guess I'm going to have to learn Python. \$\endgroup\$ – pacmaninbw Jul 24 at 19:52
  • 3
    \$\begingroup\$ Well, Python... IMHO there seem to be two opposite zombie pools: beginner questions that grow tiring over time especially if they came in large "batches" e.g. from programming challenges. On the other side there are specialised applications that require domain knowledge and thought. Sometimes you take the time to look at them and play with different approaches, write up an answer. and the OPs simply don't seem to care anymore. \$\endgroup\$ – AlexV Jul 24 at 20:59
  • 1
    \$\begingroup\$ @AlexV Yeah I try to keep away from the first one personally. So it'd make sense for it to be a zombie pool. I think those sort of questions are one of the biggest reasons Python has zombies. I'm not too familiar with the second pool, are you referring to posts made by say Graipher or myself that go unanswered? \$\endgroup\$ – Peilonrayz Jul 24 at 21:21
  • 4
    \$\begingroup\$ "What is the incentive for me to spend hours laboriously answering a question that no-one's going to see?" people do see them, because they suddenly appear on the front-page again. Granted it's unlikely to make the HNQ (all the better I say), but it's not a forlorn hope. \$\endgroup\$ – VisualMelon Jul 24 at 23:46
  • \$\begingroup\$ @Peilonrayz: I thought of some numpy questions that I was (more or less) recently looking on. \$\endgroup\$ – AlexV Jul 25 at 5:31
  • 2
    \$\begingroup\$ Glad to see I could be hit by a bus and the Python tag would still get answers :) \$\endgroup\$ – Graipher Jul 25 at 14:56
  • 4
    \$\begingroup\$ FWIW, I regularly look through the "New Answers to Old Questions" list (available to 10k+ users) and reward everything good I find there. So answering old questions gets at least my attention! \$\endgroup\$ – Toby Speight Jul 26 at 7:53
  • \$\begingroup\$ @Graipher Don't be get any funny ideas now... :P \$\endgroup\$ – Peilonrayz Jul 26 at 9:38
  • 1
    \$\begingroup\$ In terms of going after up-votes, I seem to be getting more upvotes per answer for answers to C# zombies than for answers to new Python questions, so it's not quite as simple as you suggest (unless your instinct for hotness is well honed). \$\endgroup\$ – Peter Taylor Jul 26 at 17:16
  • \$\begingroup\$ Have the number of zombies gone down since you posted this answer? More specifically how many of the 1000 zombies with unvoted answers have dropped out? \$\endgroup\$ – pacmaninbw Jul 27 at 13:26
  • 2
    \$\begingroup\$ This only goes to prove that zombie farming is an unrewarding endeavor. the reward is collecting some of the rarer badges: archeologist, revival, necromancer, .. and intrinsically, the reward is that the overall quality of the site increases when more questions get thorough answers. Isn't the ultimate goal to create a vivid wiki for sharing knowledge and experience? \$\endgroup\$ – dfhwze Jul 27 at 19:28
3
\$\begingroup\$

The point is that normal users generally won't see these questions and it might be good to bring them back into focus, so that someone can try to answer.

I support the idea of bounties on zombie questions, but for the opposite reason! :) I have [c++] in my favorite tags, and I typically look at the "unanswered questions" tab when looking for something to answer. So at least "Fluid Simulation with SDL", "Autotools detect C++ language standard support", and "Vergesort — A runs-adaptive layer to enhance sorting algorithms" are very familiar titles to me — I scroll past them every day! I wish someone would answer them so that they'd stop cluttering up my "feed" with the same months-old material every day.

I guess what would solve my peeve more effectively would be a "Hide" button — a way to say "I believe I'm never going to answer this particular unanswered question, so please never show it to me in this tab again (unless maybe there is activity on it)."

\$\endgroup\$
  • 1
    \$\begingroup\$ You can sort the Unanswered page by Newest. \$\endgroup\$ – Peter Taylor Jul 26 at 17:21
  • 2
    \$\begingroup\$ @PeterTaylor yes, when I go looking to close zombies I generally go for the oldest first. In this case I went for the highest votes so that we could at least answer more popular questions. \$\endgroup\$ – pacmaninbw Jul 27 at 13:02

You must log in to answer this question.

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