Useful strategy to vote questions and answers

I think an increased number of votes would be valuable for MATLAB Answers and a more or less common strategy for voting would be helpful.

When should we vote questions and answers to improve the use and usability of this forum?

Answer 1

[Andrew Newell]

I think there is no need for a common approach to this, although there might be some good in sharing our views. If we all voted the same way, all answers would have either no votes or a lot of them! But if we think votes are useful, we should vote whenever we read a question or answer and think it is useful. If Paulo wants to boycott the process, that's fine too.

[Paulo Silva]

I won't bother you guys while you vote on every thing that comes up, have fun!

[Jan]

I agree: Let's voting *useful* questions and answers.

Answer 2

[Walter Roberson]

I have not seen anything indicating that the intention was to get a list of the 500 most "important" questions and solutions. On the contrary: the impression I have is that the intention is that this act as a long-term knowledgebase, with the votes acting as guidance towards "best practices". The idea of it being the 500 "most important" is in contradiction to portions of a Mathworks document I was sent.

I myself do not vote very often. With over 4000 Questions and over 6000 Answers and over 9000 Comments, with the rate of new Questions or Answers (or comments) now exceeding 50 overnight (UTC-6 timezone), I don't have time to vote up answers and perhaps even remove my vote from answers as better ones became available. Perhaps if I were concentrating my efforts on particular topics of expertise, but I read (skim) over 90% of everything and think about or actively work on a wide variety of questions. Voting regularly on as much as practical becomes too much overhead for me.

When I do vote, I usually vote for Answers (or Questions) that involve techniques or faculties that would be likely to stretch the knowledge of a "competent" user. "Advanced MATLAB". For example, the perfect Answer to a Question about removing zeros from a vector is still likely to be uninteresting to anyone who knows logical indexing or "nonzeros".

Forums such as this one are bound to attract a large number of beginner questions, and, Yes, a large number of "doit4me" questions. Any consideration of voting strategies needs to take these factors in to account. The "best" structure for those who have not even read Getting Started is going to be different than the best structure for those in search of more complex information, which is going to be different yet from the best structure to guide students through homework assignments.

[Jan]

Thanks for this helpful answer. I think my "most important" and your "best practices" mean exactly the same, therefore I do not see the contradiction.

I understand, that you do not vote answers frequently (you simply cannot vote the most answers, because they are yours). Voting a question, if you find it "interesting" during you answer it, might be helpful for the community.

Most of all I've been attracted by this sentence in Helen's mail: "Are you tired of answering the same question over and over again?" Can we reach this by voting?

[Walter Roberson]

Only 1 in 4 of the Answers are mine ;-)

[Sean de Wolski]

That's a better hitting percentage than our weatherman.

Answer 3

[Paulo Silva]

Another question about votes?! why are you and others begging for votes every day?! I may seem to be rude or impolite but c'mon why the vote paranoia?! there are too many unanswered questions, please stop hunting and begging for votes! thanks

[Jan]

Votes will increase the quality and usability of this database. Imagine a beginner reads the 10 most voted threads to learn efficiently from the problems and solutions of others.

This would be much more useful than reading the all answers of the 10 users with the highest reputation numbers: On one hand these are 3100 answers, and most of them are "Read the FAQ", "Please add the necessary details" and tricky solutions to very rare problems.

While voting can increase the quality of the forum, the reputation number is just an emotional feedback for the authors - and an indicator that the authors have too much spare time. In a thread with several answers, I care for the accepted and voted answers, but the reputation of the authors is meaningless.

As far as I understand, TMW's intention was to get a list of the 500 most "important" Matlab questions and solutions 1. to assist users and 2. to improve their product (which actually equals 1.).

[Paulo Silva]

"but the reputation of the authors is meaningless" yes right, so meaningless that even before the official ranking system some people were very interested in the rank that they made their own code, and there's also the common "does that mean you accept the answer" ;) , even matlab people rank the users by their points.

"As far as I understand, TMW's intention was to get a list of the 500 most "important" Matlab questions and solutions 1. to assist users and 2. to improve their product (which actually equals 1.)."

If that's really true than it should be MathWorks people to ask for us to vote more, not us. It's not a race for points, we are here to help and learn.

Just my 2 cents

[Jan]

@Paulo: Please do not confuse the topics: My question concerns the *votes*, not the *reputation*. The side-effect, that voting increases the funny reputation number, does not reduce the impact of votes.

[Walter Roberson]

Sorry, Jan, you already introduced the topic of reputation when you spoke about reading the answers of the 10 users with the highest reputation. It is now part of the discussion.

[Jan]

@Walter: Of course it is part of the discussion. I just asked Paulo not to confuse the topics "hunting for reputation points" and "improving the quality of the forum by voting". I asked for voting strategies to improve the usability of the forum, mentioned that the reputation number is less helpful, and Paulo speaks about the race for points. These are different topics.

[Walter Roberson]

I have not seen any evidence that this system was set up to be able to select the best 500 questions.

What I _have_ seen is explicit statements from the people responsible for the system that this system is for "Users helping users".

[Jan]

You are absolutely right: This intention was never mentioned by any TMW employee, nor anywhere else. I have used the magic term "500 best questions" as an arbitrary example of what voting might be useful for.

The intention of TMW was to create a knowledge base, which is more useful than CSSM - do you agree? All I want to know is: how can we use voting to increase the quality? And with "quality" I mean the difference between "users helping users" and "long-term knowledegebase" on one side, and "tired of answering the same question over and over again" on the other.