|
Thanks Sasha for the information.
But there is 3 important benefits to forcing people to do so for vote for 1-3 stars:
- First the person has to identify itself. It should help reduce bad vote without an appropriate reason. Perhaps just a bit, I don't know, but it would help reduce inappropriate down vote.
- Second. It would give author an idea of the reason, although it is a silly one, it give him helpful feedback.
- Third. A reply from the author to a bad evaluation comment would give him the opportunity to answer the down voter which would also give valuable information to him and other readers too.
I think it was a mistake to remove that feature.
|
|
|
|
|
All articles got the vote 4 or 5 with that system, as people didn't care to write any reasons for their vote. So a lot of useless articles got the vote 4 instead of 1, 2 or 3, meaning that the voting system didn't represent the real value of the article.
As it happens, there are plenty of ridiculous articles that have been up-voted by less experienced programmers, like code vulnerable to SQL injection attacks, bad design choice etc, but it still had the voting average of 4.8 or around there.
As you might understand you have chosen to open up Pandora's box here
|
|
|
|
|
What I understand from your comment, is that forcing comment to any kind of vote would help improve the quality of the value of votes. It would inevitably also reduce the number of vote but if it is for more significant evaluations, I think it would just give more value to my proposition to put back a system where comments are necessary to vote (at least negative but also for positive).
Thanks,
Eric
|
|
|
|
|
|
|
Also, there is a really effective way to really improve quality of vote (have significant evaluation of articles) and reduce user nonsensical comment.
Suggestion 2:
If the system force comment with vote, track a vote with its comment and the system enable people to vote up or down comments. Then comment quality will improve and their value become more relevant - comment quality related to a vote could be used to calculate the real vote effect. That way the system auto adjust itself and any votes become a lots more relevant and value of the overall website increase by the same way.
|
|
|
|
|
This question has been posted and discussed many times. You cannot force people to leave a sensible comment with their voes. Just accept that there are people who will give you down votes for no reason, but it really does not matter: they do you no harm.
|
|
|
|
|
Have you ever written an article? A vote of 2 hurt me more than a kick in my butt.
My code is node perfect. Some could say it should have been separated into many files, other could say that it requires region but it does what it should do and do it well. It never deserved a 2.
Having such people with such behavior just prevent people like me to publish here or simply stop to take time to publish at all. Our rewards, is votes. Download and views are find but votes is almost everything!!!
|
|
|
|
|
Eric Ouellet wrote: Have you ever written an article? Yes.
Eric Ouellet wrote: A vote of 2 hurt me more than a kick in my butt. Why? It means almost nothing, just the opinion (flawed) of 1 out of 11,000,000+ members of CodeProject.
Eric Ouellet wrote: votes is almost everything! If that is what you think then you are probably at the wrong place (physically and emotionally).
|
|
|
|
|
I don't know about you, but my reward is the knowledge that I might have helped someone. That's worth far more than votes.
|
|
|
|
|
I don't know if this idea has been discussed before. Sounds like it could do the trick but the implementation would probably give the CP-staff some headaches.
If the brain were so simple we could understand it, we would be so simple we couldn't. — Lyall Watson
|
|
|
|
|
Yes, you are right, it is not simple. But I think they need to do something. Not necessarily my idea but anything to minimize bad behaviors.
|
|
|
|
|
So you have people voting on comments on votes?
Perfect in theory. Perfect if we have lots and lots and lots of votes to help provide statistically relevant data. However this, unfortunately isn't the case.
The whole reason I scrapped comment-on-vote was because the data we were getting was no longer statistically valuable.
cheers
Chris Maunder
|
|
|
|
|
This hole discussion reminds me of something like: make the CodeProject like a university, the professors will decide what it's worth etc. And you, obviously, would have to take lots of exams with perfect grades to be a professor
|
|
|
|
|
No. Not at all. Don't turn my discussion as a joke. I'm serious.
The best way to manage a website is to program it to be auto managed by its member.
But you have to put system of reward/awards and rules that would make users being able to manage it them self according to their trustfulness.
|
|
|
|
|
The hole point of a voting system is to accurately evaluate its usefulness and how good the solution really is.
Clearly the voting system have to be somewhat time limited, as a great Pascal article is not so useful as a good C# article, or perhaps it was a workaround for a problem that has been fixed. So the vote cast is no longer valid? It was good, but now is bad.
Then there is the problem on who gets to say what a good article is. If it's math, someone with a degree will be able to point out mistakes that someone without a degree can't. Majority is not always right logic.
So the votes from a high rep member is more worth than a low one, but that might not be accurate either, as there is no guarantee that the high rep member knows how good it really is. You do, after all, get high rep for everything that is produced (that people like), like you are a virtual polymath.
Then there is the friend or family vote, the good old corrupted vote. I like you, therefore your article is good, or worse you are my boss so I always like your stuff.
So, I wasn't mocking the idea. It's just that it has been discussed before, and the problem is always:
Some votes are genuine
Some votes should be ignored
How to decide?
|
|
|
|
|
Hi Chris,
I don't understand how you can say it is a wrong idea because it is the main reason why StackOverflow works so well. Also statistics become relevant after times, more times = better stats. And Stats could be use to improve them self. I highly think that is the way to go to improve the quality and usability of this website. My article about WrapPanelFill is fine and could be really useful to many people, i'm sure. But a stupid a**hole voted 2 for it without any comment and he was the second to vote. Result is that the article is now evaluated to 3. I personally never take a look at article with evaluation of 3 because I think it is probably not valuable. There should be a way to prevent that to happen, or at least to minimize it to a minimum.
|
|
|
|
|
SO works well because they have dedicated moderators who close poor questions with almost zero tolerance.
Eric Ouellet wrote: Also statistics become relevant after times, more times = better stats
Not if the input is biased, as it was. More time = more biased results = worse stats.
Believe me: we tried it. For a long time. It was a hard, hard decision to turn off comment-on-vote but it wasn't done on a whim.
Eric Ouellet wrote: But a stupid a**hole voted 2 for it without any comment and he was the second to vote. Result is that the article is now evaluated to 3. I
Ans so your article needs time. More time with unbiased results means a more accurate vote for your article (And it seems your articles are now over 4.8 average - so well done!)
Eric Ouellet wrote: There should be a way to prevent that to happen, or at least to minimize it to a minimum.
Take a look at the Code Project Rating and Reputation FAQ[^]
cheers
Chris Maunder
|
|
|
|
|
Thanks for having taken the time to answer me. I appreciate.
|
|
|
|
|
Hi talked to 3 persons about what happen to my tips.
They all agree with me on 4 points:
- Peoples voting 1,2 or 3 should leave comment (with a minimum of character length) in order to ensure to give feedback to the author in order to let him know the reason and give him the opportunity to improve its article.
- Due to anonymous nature of voting (now without comment) clearly make it easier to vote anything.
- A system that let vote down so easily let room for possibility to discredit good article in its early life and lead to improper perception of usefulness.
- Peoples that receive bad vote will stop writing. In general, it is fine. But a good potential writer fooled by wrong vote in its first articles could decide to stop writing and it could lead to many good articles being lost (if wrongly evaluated by some users).
Although there is probably good reasons to not force people to comment when voting, you absolutely not convinced me that it is reasonable to let people vote down without any comment. I definitively find that removing this restriction is a very hard to understand decision.
Thanks,
Eric
modified 21-May-15 17:08pm.
|
|
|
|
|
|
The article was deleted; why I don't know. But the download is gone and there's no way to get it any more.
If the brain were so simple we could understand it, we would be so simple we couldn't. — Lyall Watson
|
|
|
|
|
I don't know how Griffs muckup happened. But I'm always a bit scared to remove the wrong message when I'm.on a tablet. The precision on a tablet is a bit lacking and the confirmation is on a new page.
So could we please have the button moved from the subject bar so that you need to have the message open before removing it.
|
|
|
|
|
It's a bit close on my tablet with the new layout, simply because in Portrait mode it sits over the message subject. I don't really have a problem in Landscape (but the previous version worked better on a tablet in "full site" / Landscape IMHO)
Bad command or file name. Bad, bad command! Sit! Stay! Staaaay...
|
|
|
|
|
I missed my button when deleting spam in the lounge, and...um...I deleted your sticky post Chris...
Please restore it!
Bad command or file name. Bad, bad command! Sit! Stay! Staaaay...
|
|
|
|