Biostar third beta: Only upvotes?

[Michael Dondrup]

Hi Istvan and others,

while I like the overall implementation and wish to show respect for the good work, there is one critical point: 

No down-voting implemented. I have noticed that you has omitted this feature on purpose, but

I would like to see this discussed as it is quite a change, what do the others think about the voting scheme. 

I personally don't like to see the 'dislike' disappear because it seems logical to me to have both directions of either 

consent or dissent available instead of being a 'one-way ticket'. Please share your opinion (thumbs up or down).

Best

Michael

[Deniz Kural]

Quora and StackOverflow have different approaches to down voting,  with different outcomes;  it is more of a matter of which way you would like to nudge community culture by imposing different set of behavioral constraints...   what is your thinking on what's a best fit? 

Best,

Deniz

[Michael Dondrup]

I have had a look at Stackoverflow discussions to see how this is treated there. From these discussion, I found a cite a can totally support and that matches with my gut-feeling of what down-voting is good for: 

In general I like the concept, while I only exercise down-votes in case of severe flaws, like in this cite:

• The OP clearly didn't try to put together a clear question
• The OP is just looking for someone to do their work for them for free
• Someone else has already commented on what the issue is
• The OP clearly didn't do any sort of "self-help", even as basic as a google search, prior to posting
• The post is too chaotic for me to have any reasonable understanding of what the OP is after
• The OP has posted similar crap before (yes, I do look at profiles from time to time to determine my behavior)

http://meta.stackoverflow.com/a/125868

I can support all points in this cite, these are totally valid reasons where there is actually a need for voting down, to set some pressure.

I disagree, that this is too 'punative' unlike Istvan, in fact it is a good to have at least some sort of pressure at hand and also to mark questions that have room for improvement. I therefore highly appreciate the novelty and competitive nature of the stackexchange format over a normal bulletin board or mailing list design.

Being critical about shortcomings in questions and answers of others is crucial, as is the ability to  take criticism. However, if only positive votes (like Like-button in Facebook) are possible we will fall back on an unrealistic view    

of the community where everybody is just 'nice to each other'.

2012/3/16 Deniz Kural <deniz...@gmail.com>

[Daniel Swan]

Briefly, I agree with Michael on this. The downvote is not punitive, it's a metric of the worth of a question.

Regards,

Dan

--
Daniel Swan || dan.swan[at]gmail.com || http://eridanus.net/ || http://twitter.com/d_swan

"Reality is that which, when you stop believing in it, doesn't go away." (Philip K. Dick - How to Build a Universe)

[Sean Davis]

I agree here, also. Downvoting is a useful evil.

Sean

[Istvan Albert]

Hello all,

The infrastructure is all there for downvotes as well, it is just not enabled.

I can easily put it back if that's the consensus.

In brief my main thinking was that perhaps the person that gets
downvoted does not really understand what he/she should be doing in
the first place. A down-vote does not really help them improve on that
in fact it could be discouraging. Having only upvotes would also
perhaps encourage people to vote up more, I consider no votes being a
pretty bad state to being with.

Reddit comes to mind where periodically people complain about
downvotes and attempt educate others that downvote should go to a post
that is badly written and not one that you disagree with - to no
avail.

best,

Istvan

--
Istvan Albert
Associate Professor, Bioinformatics
Pennsylvania State University
http://www.personal.psu.edu/iua1/

[Chris Maloney]

In the current site, the down-vote is registered immediately.  Would it be possible to set it up so that clicking the down-arrow brings up a modal dialog box that explains what down-voting is for, suggests "please add a comment", asks for a reason with a radio-button, and then forces the user to click "confirm" or "cancel"?  If you raise the trouble-bar for downvoting a little bit, maybe that would make capricious downvoting behavior less of a problem, while preserving it's potential for usefulness.

[Michael Dondrup]

I think SO has such a reminder to leave a comment in place, I support this idea.

2012/3/17 Chris Maloney <vold...@gmail.com>

[Egon Willighagen]

On Sat, Mar 17, 2012 at 1:27 AM, Istvan Albert <iu...@psu.edu> wrote:
> Having only upvotes would also
> perhaps encourage people to vote up more, I consider no votes being a
> pretty bad state to being with.

True. But that would actively discourage anything than popular,
because it would make any more specialized bioinformatics question end
up in the same category as bad questions... (yes, metabolomics or
anything with atoms is in that corner of bioinformatics)... I do not
like the sound of that.

Upvoting without downvoting does not make sense to me... it removes an
origin and makes the upvote count relativistic.

If the problem is that people downvote without explaining why they
downvoted that question, then there are surely other mechanisms... SE
already suggests to place a comment after you downvoted sometihng, an
a (bronze) badge for people with 10 downvotes in row without any
comment on the question sounds at first sight an attractive
alternative too...

Egon

--
Dr E.L. Willighagen
Postdoctoral Researcher
Department of Bioinformatics - BiGCaT
Maastricht University (http://www.bigcat.unimaas.nl/)
Homepage: http://egonw.github.com/
LinkedIn: http://se.linkedin.com/in/egonw
Blog: http://chem-bla-ics.blogspot.com/
PubList: http://www.citeulike.org/user/egonw/tag/papers

[Albert Vilella]

I also think the voting and other rules should be as similar as
possible to the current stackoverflow.com site.

[Istvan Albert]

Hi All,

This discussion made me want to look more deeply into this issue and
see what actually happens on the site.

We have a total of 58634 votes distributed over 18324 questions and
answers. Out of these 1597 (2.7%) are down votes. Some posts getting
downvotes or inappropriate therefore are closed or deleted by
moderators. The main value of downvotes is for ranking open posts, at
that point only 1304 (2.2%) downvotes remain.

So the first conclusion that we can draw is that downvotes are very rare.

Now let's look at the most downvoted open question on Biostar:

http://biostar.stackexchange.com/questions/3774/quad-core-i3-i5-and-i7-bioinformatics-tool

This post has the most downvotes 15, its rank is not the lowest since
it also has 9 upvotes.

I hate to say this guys but I do not think that this is a question
that warrants it to be the most downvoted post on Biostar. In fact
(and I swear I did not cherry pick this), it exhibits the exact issue
that I wanted to highlight - downvotes are far more personal than
upvotes. People will start picking sides. This includes me as well. I
for example have upvoted this question, but not because I think that
it is particularly good question but simply because I did not think it
should have a score of -10 or so that I saw when I looked at it.

Note also that the creator of this post has never returned to Biostar
- at least not on this account in question.

I think this above illustrates a few negative aspects of having
downvotes, I would also agree that measuring the benefits is probably
more difficult.

[Michael Dondrup]

Hi,

while this is an interesting observation I don't think this can be used as an

argument for either having or not having down-votes. In my opinion up-voting

without without ability to down-vote is pointless, even though we see

this right exercised with very carefully atm. voting is a substantially contributing factor

to the democratic process to find an optimal answer to a problem, it is most often not 

used to punish but to evaluate the problem and solution. Would you f.e. easily wish to

waive your right to vote for an opposition party in a parliamentary election (given you have it in the first place)

even though you are atm very content with the work of the ruling party? 

Further, if a poster cannot deal with criticism and leaves, I'd say that is a deficiency in they personality

he/she had to overcome in the first place. If they hoever are unable to learn deal with criticism adequately

 (I know quite well personally I need to improve in this aspect as well), then they possibly have nothing 

significant to contribute to the community and it might be better they leave anyway.

Best 

2012/3/17 Istvan Albert <iu...@psu.edu>

[Albert Vilella]

My opinion is that one shouldn't tinker too much with a scoring system
that has been proven to work on many sites, with a variety of user
base profiles, including much bigger user bases, like
stackoverflow.com. Most sites have downvotes and they work *very*
well. Why would they not want to have downvotes?

[Chris Maloney]

Hi, Istvan,

Thanks for taking the time to study this question, it is indeed interesting!

As for your example of the most down-voted question, I agree with Michael that I don't think it is sufficient for arguing one way or the other about down-voting.  First of all, the sample size is too small!  It's really an argument-by-anecdote.  

Next, I took a look at it, and I do happen to think it's a poor question.  I probably wouldn't down-vote it, but I certainly can see why someone would.  You upvoted it -- that's great, that's your perogative.  But by saying that you think this illustrates that "downvotes are far more personal ...", you are second-guessing the others who down-voted this question, and I think that's inappropriate.

You also wrote, at the beginning of your response, "Some posts getting downvotes or inappropriate therefore are closed or deleted by moderators."  I would ask, are the down-votes ever helpful to moderators in making this decision?  I would guess that at least sometimes, they are.  The general population doesn't have the moderator's god-like power to close or delete a question, so that makes it all the more important that they at least have some quantitative way to flag something.

Finally, let's consider the percentage of down-votes among all the votes for open questions -- 2.2%.  Rather than this number showing that it is a rarely-used feature, I would site this as evidence that it is a feature that is not being abused.  It means that users are generally very conservative and careful before applying this tool -- which is a good thing, and argues more for keeping it, in my opinion, than for throwing it away.

Cheers!

[Pablo marin-garcia]

Hi,

I also upvote for downvotes, but forcing the voter to leave a comment.
This will allow the down-voted person to know if the negative point
came from 'ideological dissagreement' or because wrong/illogical/lazzy
question so he can learn from his errors if any.

--
   - Pablo Marin-Garcia

[Istvan Albert]

A few more numbers and a plot to think about over the weekend:

Registered users: 4166
Active users (users that posted at least once): 2940
Users that voted up at least once: 1446 (49% of active users)
Users that voted down at least once: 218 (7% of active users)
Users that bookmarked at least once: 773
Users that bookmarked but never voted: 135
Users that only downvote: 0

And finally a histogram of vote counts per user:

http://imgur.com/73VWc

horizontally you can see the number of votes that have been cast,
vertically how many such users exist. x-axis is scaled
logarithmically, upper panel for downvotes, lower for upvotes.

Highest votecounts are not easy to read off so here they are 156 are
the most downvotes by a person and 3135 are the most upvotes by a
person.

best,

Istvan

[Giovanni Marco Dall'Olio]

Hello,
You can also take into account that downvoting costs some reputation points. This is a smart idea that the stackexchange creators have had, because it forces people to think before downvoting. The penalty for downvoting is just 1 or 2 points, which is a balanced number, not too big so people won't be scared of putting a down vote, and not too small so people are forced to think before doing so.
The idea to force people to add a comment when commenting is nice. I think it is already implemented in stackexchange, although in that case it is only a suggestion, and it is not mandatory.

Cheers,
Gio

--
Giovanni Dall'Olio, phd student
IBE, Institut de Biologia Evolutiva, CEXS-UPF (Barcelona, Spain)

My blog on bioinformatics: http://bioinfoblog.it

[Casey Bergman]

I agree downvotes should be kept, with the same negative disincentive for downvoting as the current system.

I don't agree with requiring a comment for a downvote, since this would additionally discourage downvoting and oftentimes the reason for downvoting is obvious. If this is required, why then would we not require commenting for upvotes as well?

As an alternative to prevent inappropriate use of downvotes, I suggest listing the avatars of each person who upvotes/downvotes for a question/answer (with a hover-over for user names, like a wordpress post), so there is transparency, accountability and more user-specific context for all voting. This should make it less likely for callous/retributive downvoting and allow users to interpret the score of a post by who cast votes for/against it.

Best,
Casey

[Istvan Albert]

Hello All,

I see that you guys prefer to have down-votes.

While I have some doubts regarding their overall utility: few people
use them, more importantly it is not clear how those receiving them
interpret and react to them.

Some of you brought up SE as an proof that these work, upon closer
inspection it seems SE also tries to steer people away from downvoting
- urges people to try to get the author to improve the answer instead.

That being said I'll add back the downvotes since it seems to be the
consensus. I do however want to change a bit how scoring works.
Discussions like this and disagreements usually force people to
properly verbalize concepts and that in turn help clarify their goals
and expectations - and in this case my own.

I think scoring questions alone is wrong. No one reads Q&A site for
the questions, we read them for the answers. A bad question can have a
phenomenal answer that makes the thread worth reading. Therefore a
question's score should be the sum of all upvotes that are associated
with answers/comments in the thread that was started by the question.
This will allow us to rank threads by the value of the entire content
in them rather than just the voting on the question. Many great
answers a buried under not so great questions.

Of course the author's reputation will only be increased by votes on
posts they have themselves made.

Downvotes will not carry reputation penalty neither for the voter nor
for the target. This just leads to all kinds of complexities to
protect from abuse. I find the current model where a voter reputation
is reduced just because they choose to downvote as borderline absurd.
If we allow downvotes let's not discourage people from using them.

With these two tweaks I hope to find a compromise.

[David Quigley]

I'm confused. If I ask a question that no one up-votes on its
intrinsic merits, but someone gives a fantastic answer that is up-
voted 20 times, does the mediocre question have a "20" next to it when
someone is browsing questions? I think I understand you to mean that
the answering person gets credit for 20 up-votes and I get no credit
for my question.

[Istvan Albert]

On Thu, Mar 22, 2012 at 9:51 AM, David Quigley
<david.a...@gmail.com> wrote:

> I'm confused. If I ask a question that no one up-votes on its
> intrinsic merits, but someone gives a fantastic answer that is up-
> voted 20 times, does the mediocre question have a "20" next to it when
> someone is browsing questions? I think I understand you to mean that
> the answering person gets credit for 20 up-votes and I get no credit
> for my question.

Yes that was my thinking.

To have a way to show which questions have excellent content
associated with them.

On the question itself we could display the score as the sum 0 + 20
rather than just 20

best,

Istvan

[Giovanni Marco Dall'Olio]

Maybe we can use a different color, or a symbol, to highlight the questions that have a good answer.

[Istvan Albert]

On Thu, Mar 22, 2012 at 10:23 AM, Giovanni Marco Dall'Olio
<dallo...@gmail.com> wrote:
> Maybe we can use a different color, or a symbol, to highlight the questions
> that have a good answer.

In the meantime have implemented and deployed this behavior see the test site:

http://test.biostars.org/

It works like this: when you look at a listing of questions the number
of votes corresponds to the number of total votes for the entire
thread.

Once you click on the question you will see the votes broken down by posts

(also downvotes are back as well)