What is meta? ×
Meta Stack Exchange is where users like you discuss bugs, features, and support issues that affect the software powering all 127 Stack Exchange communities.

I'm a bit embarrassed to admit this, but I failed the review honeypot at http://stackoverflow.com/review/first-posts/1126342. (The question is at How to create a link text in a table using c#.)

STOP! Look and listen

I have no technical knowledge of the topics here, but grammar and spelling were OK. I saw that the answer started with a question, but since it also included some "best guess" answers, I figured it was OK. (I did think it was weird that the system said this was an answer by a new user, but the user has more than 750 rep.)

Actually, that answer now seems to have been deleted, so I can't check the review history.

Was the answer really terrible? Or did the system show me a post-edit answer, when the pre-edit answer was the terrible one?

share|improve this question
6  
This wasn't a real answer. This "answer" should have been posted as a comment. If you're unsure about the content because it's not your field, you should just have pressed skip. –  BalusC Dec 4 '12 at 17:47
6  
This actually depends somewhat on what the question was. If this is a best guess attempt at answering the central question it is OK, but if it is a best guess attempt at providing tangential information it is a comment disguised as an answer. –  Asad Dec 4 '12 at 17:48
1  
I just saw the same one; I was reaching for my auto-commenter when I was told it was a fake. –  Martijn Pieters Dec 4 '12 at 17:48
    
The first version of the answer wasn't really an answer. –  Luksprog Dec 4 '12 at 17:50
5  
The answer in the screenshot was the answer I was presented. –  KatieK Dec 4 '12 at 17:51
    
@Asad: You can always visit the question from the review page linked and scroll down. ;) But here's a link. –  animuson Dec 4 '12 at 17:52
    
"I have no technical knowledge of the topics" - did you attempt to upvote? –  gnat Dec 4 '12 at 17:52
    
@gnat - Yes, I tried to upvote. I've gotten previous feedback that it's OK to review out of my areas of expertise if everything else looked ok. (I can't seem to track that link down right now.) –  KatieK Dec 4 '12 at 18:00
17  
You should absolutely not vote posts which are out your area of expertiste! Suppose that the answer was actually wrong, this would only mislead future readers and experts. –  BalusC Dec 4 '12 at 18:07
11  
@KatieK It's perfectly fine to review out of your area of expertise, but the types of review actions you should perform will be limited in such cases. Flagging a post that's obviously a violation of SO guidelines, editing to fix grammar/formatting, etc. can be done regardless of your understanding of the content. Voting on questions also tends to require much less knowledge of the subject material. Voting on answers however, shouldn't be done without understanding the content of the answer. –  Servy Dec 4 '12 at 18:08
    
I'm confused now... the answer was both edited and deleted on 11/29. Why is it showing up the review queue on 12/4 in the first place? –  Rachel Dec 4 '12 at 18:11
    
@Rachel: Because it's a honeypot in order to catch bad reviewers. See also meta.stackexchange.com/questions/152709/… and meta.stackexchange.com/questions/149621/… –  BalusC Dec 4 '12 at 18:12
8  
@Chichiray But is it still so bad that it should be used in there? If this is really such a good example of poor quality, it would have caught me out as well. –  Bart Dec 4 '12 at 18:12
2  
@Bart: I'd just have pressed "Skip". I know nothing about the subject in question either. –  BalusC Dec 4 '12 at 18:15
2  
I view this as a bug. See meta.stackexchange.com/questions/157659/… –  Jeff Atwood Dec 5 '12 at 9:29
show 2 more comments

2 Answers

up vote 15 down vote accepted

To clarify, the original answer looked like this:

Maybe I didn't understand the question, but if you only want the url text, why are you creating an anchor tag? should'nt it be better just to put the url text without the html format?

The reason this was a honeypot is pretty obvious from looking at it. Bad timing I suppose.

I think it is also important to note that reviewing edits is not exactly equivalent to reviewing first posts.

When reviewing edits, you are evaluating validity instead of quality. An edit either follows some rules or it doesn't, and the content of the post being edited is only marginally relevant. On the other hand, when you are reviewing first posts, you are making a judgement about the quality of the post, which means that the content is of the utmost importance.

I would say reviewing edits in areas where you have no technical expertise is fine if you exercise caution, but the same doesn't apply for first posts.

share|improve this answer
3  
So you're saying that "bad items" are still part of the test set, even when they have been improved to within acceptable limits? That smells like a bug then... –  Bart Dec 4 '12 at 18:11
5  
@Bart Maybe they should be pulled from honeypot status immediately after any edit, pending a reevaluation. –  Asad Dec 4 '12 at 18:13
    
My guess is that the way it worked was: 1. Bad post was posted 2. Post was flagged 3. Post was edited to improve it 4. Flag was seen by a moderator who was shown the initial revision, not the latest revision, so he deleted the post. Given that set of actions the last thing that happened to the post may have been that it was deleted by a mod. Without having a link to the post to see the revision history it's hard to say for sure. –  Servy Dec 4 '12 at 18:23
    
@Servy - Here's the question: stackoverflow.com/questions/13621128/… –  KatieK Dec 4 '12 at 18:26
    
@KatieK The revision history confirms my previous comment. It was deleted after the latest revision. –  Servy Dec 4 '12 at 18:27
1  
@Servy I'm not sure it is delete worthy as it stands though. Regardless of whether it is right or wrong, it contains an attempt at solving the problem. –  Asad Dec 4 '12 at 18:29
    
@Asad My point is simply that Bart's comment about it being edited after it was deleted clearly isn't the case here. Whether or not it should be deleted is something I haven't commented on. –  Servy Dec 4 '12 at 18:31
    
@Servy Yeah, I was just putting that out there. –  Asad Dec 4 '12 at 18:33
3  
So looking at the answer it's clearly of poor quality, and I would downvote it if it weren't deleted, but it does appear to attempt to answer the question enough that it shouldn't be deleted. It's not answering the question as asked, but it is solving the problem that the OP is looking to solve (kinda). –  Servy Dec 4 '12 at 18:34
4  
@Servy At no point did I say it was edited after being deleted. I'm saying it was apparently edited after being selected for the set of honey pot questions. And it is still in there when that might no longer be entirely justified. –  Bart Dec 4 '12 at 18:45
2  
@Bart All indications are that it was not edited after it was selected as a honeypot item. The issue is that it was deleted by a mod when it shouldn't have been, which indicated to the honeypot post selector that it was a viable option when it really wasn't. –  Servy Dec 4 '12 at 18:48
    
@Servy Hmm, okay, might be as well. I'll have to look into this whole honeypot business. It's been a long time since I've used /review seriously anyway. :S –  Bart Dec 4 '12 at 18:49
5  
@Bart While I haven't seen any official post stating so, all indications I've seen thus far indicate that honeypot posts are simply chosen from posts deleted by a moderator as a result of a flag. Sometimes it picks good examples of what not to do, sometimes it picks borderline posts that, maybe should have been deleted, but are in a grey enough area that some people justifiably get it wrong, and some posts just shouldn't have been deleted in the first place. This isn't the first honeypot post that I felt really shouldn't have been deleted in the first place. –  Servy Dec 4 '12 at 18:52
3  
@Servy Ah, like that. Would explain it indeed. It all seems rather careless though for a set of questions used to train the users in reviewing. I feel a Meta question coming up, but I can't be bothered to write it now. Maybe tomorrow. :) Thanks for the info though. –  Bart Dec 4 '12 at 18:57
add comment

I have no technical knowledge of the topics... I tried to upvote.

In cases like that, don't upvote. It's

that

simple. Test item taught you just that. "Learn to love that Skip button."


  • As explained in comments, reviewer should act judiciously in posts out of their area of expertise:

    It's perfectly fine to review out of your area of expertise, but the types of review actions you should perform will be limited in such cases. Flagging a post that's obviously a violation of SO guidelines, editing to fix grammar/formatting, etc. can be done regardless of your understanding of the content. Voting on questions also tends to require much less knowledge of the subject material. Voting on answers however, shouldn't be done without understanding the content of the answer.


This accident made me wish there was a dedicated kind of , specifically targeted to address "ignorant upvotes" issue: Present a unambigously good answer but one taken from totally different question to verify that user doesn't upvote.

share|improve this answer
4  
This answer is incorrect. Asad's answer is correct. –  Jeff Atwood Dec 5 '12 at 9:28
    
    
@gnat Almost all of those posts pertain to approval of invalid edits or upvotes to answers that violate basic rules, not to lack of judgement in assessing the content of posts. No one is complaining that someone upvoted an answer that uses eval to parse JSON, they're complaining about approved vandal edits and upvotes to lone link answers and spam –  Asad Dec 5 '12 at 11:49
    
@Asad since bounty remarks are not added to the question history, I'll just quote for further references the current one in "upvote everything..." discussion - "Trigger of today was stackoverflow.com/a/13669048 An answer which has absolutely nothing to do with the concrete question magically got 3 upvotes"... –  gnat Dec 5 '12 at 12:01
    
...don't get me wrong it's good that you discovered bug in test item. But I would much prefer this bug to left unnoticed if (if) the reason was that OP skipped post that was unclear to them. Heck I myself likely let couple bugs like that slide through by skipping and... you know I am happy with that –  gnat Dec 5 '12 at 12:09
1  
@gnat Fair enough, I missed the bounty remarks. –  Asad Dec 5 '12 at 12:22
    
for the reference, here is a discussion related purely to the fact that Careless review upvotes are actively contributing to the quality problem, not correcting it. "The default answer sort is by score. The crap is rising to the top. This sends bad messages on so many levels that it isn't even funny..." Are you reading @JeffAtwood? –  gnat Dec 20 '12 at 12:32
1  
@gnat I am but bear in mind I do not work at stack exchange any more. I did email shog my concerns (I agree with that post) and his recent answers here are worth reading if you haven't. It never hurts to email the team@ address at the bottom of the page for things you are seriously concerned about either. –  Jeff Atwood Dec 20 '12 at 15:38
    
@JeffAtwood I see - fair enough –  gnat Dec 20 '12 at 15:49
    
...for the reference, I have not only read Shog's recent answer but even cast a bounty on it few hours before I mentioned that discussion here. Not to mention that my bounty has been on that question for a few days now. And not to mention that right after giving first bounty to Shog, I started a second one, to reward Geoff Daglas' answer there :) –  gnat Dec 20 '12 at 20:45
add comment

You must log in to answer this question.

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