In a recent question “Randomized algorithm that “looks” deterministic?” by arnab, he posted an answer directly communicated by Shafi Goldwasser with the following note:

Please do not upvote this answer, as it's not mine.

(This is not the first time I have seen this kind of note on cstheory.stackexchange.com, although I do not remember when and where I saw a similar note before. I use arnab’s post as an example.)

It is perfectly fine (and necessary for honesty) to state that the answer is provided by someone else. But the “please do not upvote this answer” part makes me hmmm.

I guess that he put this note because he considered that he should not earn reputation points by posting an answer provided by someone else. Although I understand his humbleness, I do not think that it is a good idea to say “do not upvote” for several reasons:

  1. In my opinion, this answer is good and it should be listed in a higher place on the page than score 0. In general, a good answer should be listed in an earlier place. That is how Stack Exchange sites work.
  2. Avoiding upvoting a good answer just because of reputation points seems missing the point, because reputation points are not of primary importance anyway.
  3. If we have this note as a norm, I am afraid that in the future, not stating “do not upvote” in an answer provided by someone else may be viewed as inconsiderate. This seems like a wrong standard to me.

I hope that people will not mind if an asker or someone else earns some rep points by posting a good answer communicated by someone else.

What is your opinion?

(As an aside, an answer can be marked as community wiki to avoid reputation gain (or loss). But note that the primary purpose of marking a post as community wiki is to let other users edit the post, so that is a slight abuse of the system.)

Added note about community wiki. As I wrote above, I am aware that a poster can avoid gaining (or losing) reputation points by marking a post as community wiki (CW). However, this has its own drawback if the poster does not intend to make editing easier. Moreover, this does not solve the item 3 above; I do not think that we should give a poster a pressure to mark an answer provided by someone else as CW.

| |
  • 3
    $\begingroup$ Actually, I think using community wiki to mark such answers is a suitable use. Then it can be up-voted as readers see fit, without gaining any unwarranted reputation. $\endgroup$ – Dave Clarke Nov 4 '10 at 21:14
  • $\begingroup$ @Dave: Thanks for the comment. The point of the “aside” is that there is a disadvantage of marking an answer as CW and that it is understandable if someone does not want to mark the answer as CW. By the way, I am not completely convinced that reputation points for an answer provided by someone else are really “unwarranted.” $\endgroup$ – Tsuyoshi Ito Nov 4 '10 at 21:22
  • 5
    $\begingroup$ Making it CW sounds fine to me too. One way to justify it is that the other person who provided the answer can later come to cstheory and edit it if he/she wishes, since it is CW. $\endgroup$ – Robin Kothari Nov 4 '10 at 21:36
  • $\begingroup$ Yes, I put the note because I didn't want to earn reputation points from it. But the CW option does sound like it is the better option! So, I edited the answer to make it CW. $\endgroup$ – arnab Nov 4 '10 at 22:51
  • 3
    $\begingroup$ I agree with Tsuyoshi, there should be no pressure on the OP to make a post a CW in these situations, if OP prefers to make it CW it is their choice. It is probably better to just state that this answer is based on communication with X, and make it CW, no need to state that this was the reason to make it CW. Regarding the first point in the added note, I don't think that others will modify them (even if someone modifies it, the OP gets notified by the system and it is easy to role back). $\endgroup$ – Kaveh Nov 5 '10 at 3:31

I have done this before, but would have used the community wiki solution instead if I had realized that checking community wiki prevents me from getting reputation for the answer. Maybe we should have a faq about how to use this trick in situations where we want to allow answers to be properly ranked by their votes but we don't want to collect the reputation? Because it seems a reasonable use of the community wiki flag to me.

| |
  • $\begingroup$ that makes a lot of sense. I'll try to add an entry tonight $\endgroup$ – Suresh Venkat Nov 4 '10 at 22:00
  • $\begingroup$ It would be nice to collect the common uses of CW in one place: 1. don't want to earn reputation for some reason (like this case), 2. allow others to edit, 3. ranking list. Does anyone know other common uses of CW (with the recent modifications to CW made by SE overloads)? $\endgroup$ – Kaveh Nov 5 '10 at 0:08

Maybe it will seem harsh, but in my opinion it is not to the person who asks/answers a question to decide about its reputation, but rather to the community. So, if one has the information that the answer is not from the person who wrote it, it is up to her to decide whether she upvotes or not.

Let be pragmatic, if someone can bring useful answers from reputable researchers, he deserves the upvotes.

| |
  • $\begingroup$ I agree with you that “it is not to the person who asks/answers a question to decide about its reputation, but rather to the community” (and it is blatantly honest to say so :) ). However, I guess that some people want to upvote an answer without giving reputation points gain (because they consider that an answer provided by someone else should not give rep points to the poster). (more) $\endgroup$ – Tsuyoshi Ito Nov 5 '10 at 18:25
  • $\begingroup$ (cont’d) With the current system, voters can choose whether they upvote or not, but they cannot choose whether their votes increase the rep points of the poster or not. Therefore, even though I agree that rep points should be up to the community, it is impossible to put this into practice. $\endgroup$ – Tsuyoshi Ito Nov 5 '10 at 18:28

You must log in to answer this question.

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