Dpreview Voting System should be change 1 to 3 Stars.

win13

Well-known member
Messages
154
Reaction score
79
Location
Kanchanaburi, TH
Hi Folks,

Having read about so much negative voting patterns of late lets just cut the c*ap and do it the Michelin Stars way with taking out the 0 and just using 1 to 3 stars only.

If you wish to use the 0 = No Vote.

What do you Folks think about this method of voting?

Win

--
Kan Win
 
Last edited:
Hi Folks,

Having read about so much negative voting patterns of late lets just cut the c*ap and do it the Michelin Stars way with taking out the 0 and just using 1 to 3 stars only.

If you wish to use the 0 = No Vote.

What do you Folks think about this method of voting?
It doesn't end negative voting, it just changes the range of values. The easiest fix to the problem of low votes is to eliminate low votes. One of the voting schemes that DPR has said they're considering giving hosts an option to use is some form of top 5 or top 10 system, in which voters only vote for their favorites. My preferred implementation of this is also the simplest: let voters select the pictures they like up to a specified maximum number; no further differentiation is required (that is, you don't rate the pictures--you've either marked it as one you like or you didn't); the picture that gets the most "likes" is the winner. This doesn't eliminate dishonesty in voting, of course, but at least good pictures won't receive undeserved low votes.
 
Hi Folks,

Having read about so much negative voting patterns of late lets just cut the c*ap and do it the Michelin Stars way with taking out the 0 and just using 1 to 3 stars only.

If you wish to use the 0 = No Vote.

What do you Folks think about this method of voting?
It doesn't end negative voting, it just changes the range of values. The easiest fix to the problem of low votes is to eliminate low votes. One of the voting schemes that DPR has said they're considering giving hosts an option to use is some form of top 5 or top 10 system, in which voters only vote for their favorites. My preferred implementation of this is also the simplest: let voters select the pictures they like up to a specified maximum number; no further differentiation is required (that is, you don't rate the pictures--you've either marked it as one you like or you didn't); the picture that gets the most "likes" is the winner. This doesn't eliminate dishonesty in voting, of course, but at least good pictures won't receive undeserved low votes.
To vote for your top 5 pictures, or even just mark your top 5 pictures, I like that idea.

regards

JahnG
 
Last edited:
Hi Folks,

Having read about so much negative voting patterns of late lets just cut the c*ap and do it the Michelin Stars way with taking out the 0 and just using 1 to 3 stars only.

If you wish to use the 0 = No Vote.

What do you Folks think about this method of voting?

Win
The present system is intended to take account of people not voting on every entry, using a Bayesian algorithm to rank the entries. Basically there is nothing wrong with the voting system. If in a challenge with 100 entries, 100 people made the effort to vote honestly and fairly on a significant proportion of the entries, it would work fine. But when only 10 people can be bothered to vote, and 3 of them are dishonest, then you might just as well choose the winners and losers at random.

Joe
 
Probability of images getting the same number of favorites increases, up to having all top images getting the same number.

Worth considering though.
 
The present system is intended to take account of people not voting on every entry, using a Bayesian algorithm to rank the entries. Basically there is nothing wrong with the voting system. If in a challenge with 100 entries, 100 people made the effort to vote honestly and fairly on a significant proportion of the entries, it would work fine. But when only 10 people can be bothered to vote, and 3 of them are dishonest, then you might just as well choose the winners and losers at random.

Joe
The present Challenge system is flawed. There is no sense in having inexperienced & uneducated beginners hosting challenges, nor entrants voting in them...esp should not be able to vote in Challenges in which they have entered. Sheer comic lunacy.
 
Hi Folks,

Having read about so much negative voting patterns of late lets just cut the c*ap and do it the Michelin Stars way with taking out the 0 and just using 1 to 3 stars only.

If you wish to use the 0 = No Vote.

What do you Folks think about this method of voting?

Win
The present system is intended to take account of people not voting on every entry, using a Bayesian algorithm to rank the entries. Basically there is nothing wrong with the voting system. If in a challenge with 100 entries, 100 people made the effort to vote honestly and fairly on a significant proportion of the entries, it would work fine. But when only 10 people can be bothered to vote, and 3 of them are dishonest, then you might just as well choose the winners and losers at random.

Joe
Sorry to say and this has absolutely nothing to do with you...

... but this voting system is fundamentally wrong and unfair. It has come so far because some people thought they could build a statistical model of the psychology of photographer decision behavior, one way to describe this voting system.

Would the same people trust me, a total non-technician, to repair their precious camera and pay a few hundred dollars for it?

Don't think so. Because only a professional can do it.

--

TheBlackGrouse
Active outdoor photographer, trying to become better, studying user experience.
 
I really don't understand what the complaint is about. Each voter gets one vote per picture. Mark how you felt it looks and how it fits the challenge. How is this a bad thing?
 
I really don't understand what the complaint is about. Each voter gets one vote per picture. Mark how you felt it looks and how it fits the challenge. How is this a bad thing?
This is why: 'mark how you felt it looks and how it fits the challenge'.
 
You failed to say why it's a bad thing.

How is a person supposed to vote if it's not about how good the picture is and how well it fits the challenge?
 
Very good point. When voting, I start with how much an image is worth in pears. Then apples. It passes the pear test, only then do I assign it apples' worth.
 
OK, you make an interesting point. Even so, I still think it's a valid system.

For example, if I think a particular picture clearly doesn't meet the challenge, I'll give it .5 stars. I only do this if it's obvious. For example a picture of a cat in a challenge about dogs. Even if it's a great picture, it should never have been submitted so, it gets the lowest possible rating from me.

Next, I'll judge a picture on how I like it. Admittedly, I'm not an expert. Still, I like good photographs and can post an opinion on that. If one stands out, I'll vote higher. If it's just average, I'll put it in the middle. If it is a terrible picture, but meets the challenge, I'll vote it low.

That's just how I do it.

So, you're saying that the three start system is better? I don't see how.
 
You are actually quite right, the system is valid. The problem is that it is not reliable, i.e. it will not produce the same results each time. The reason for this is that the people who use the system are human, and, as we know, humans are very far from perfect.

The issue is that there are many here who are so desperate to win or gain a high placement that use unfair tactics to do so. If you have not read it then I would suggest you read the thread that I started recently on the topic:
It is known that some members will cast a large number of ½, 1 or 1½ votes and will not cast any votes higher. Such votes have the effect of lowering dramatically the overall average using the algorithm for scoring. One can only theorise as to why members act in this way, as in general the low marks for pictures are not generally deserved as there are very few pictures in any challenge which deserve a mark of below 2 if judged by normal standards. One theory clearly is that they look for images that might be better than their own and give them low marks to stop their chances or getting well placed. This in itself is bad enough, but unfortunately any other reason might indicate an even more serious psychological disorder.

It is also known that some members that some members have multiple accounts. It has been theorised elsewhere that these accounts can be used to boost personal scores in challenges while at the same time lowering the average marks of potential rivals. Other members have been found to be canvassing votes for their pictures in other places, so that the overall score is boosted and their chance of winning or being highly placed is enhanced.

There are threads covering all these issues somewhere within the “Challenges Discussion” forum, so it is an issue that is both well documented and well known. The truth is that DPReview seem to care little about the challenges as long as the “clicks” mount and they get revenue from advertising. This is a shame as many members have been deterred from participating in challenges because of their treatment by others who show little or no respect for anyone. Some others have left the site because of this problem.

Hosts, it is understood, do not know the names of the voters who perpetrate such scams and can therefore do nothing about it. It is believed that each voter is given a number or some other means of identification and the range of their votes, but not their identity is shown. From what is understood, DPReview does know the information but chooses not to share or act on it, so, to be truthful they are supporting the abuse of their members by their inaction in this situation. Does this make them equally responsible?

It is not understood what any person can gain from winning or being well place in a competition knowing that they have done so by using unfair means. The world sees how this is treated these days in sport, where those who seek to gain unfair advantage are banned for long periods, maybe even for ever. Using unfair means to gain an advantage is seen as wrong in all walks of life,

So, you are correct, the algorithm can be argued to be valid, though it is maybe overly complex for the purpose of judging a friendly competition. The problem is that the voting system is broken, because it is open to abuse, and those who govern it take no action to stop this. Maybe, at some point in the future, members will take their own action and walk away, fed up with this abuse by other members and those who refuse to take action about it – then what will be left of DPReview?
 
Well said anisah. I understand the problem of users having multiple accounts and how it can be used to sway the final score. Considering the "prize" that is awarded for winning (nothing) it staggers the mind that some would go to such lengths to win a challenge.

That said, what is a better system? Going to three stars won't change the issue you bring up. So, how do we fix it?

I like the challenges because they cause me to work outside my comfort zone. This in turn makes me a better photographer. Isn't that the goal here, to promote the art?

Even so, I see the problem where false voting can drive photographers away. If we were to remove the ability of an entrant to vote, it still doesn't solve the issue of multiple accounts. So, what's the solution?
 
Machiavelli would agree with me that the first step is to not allow participants in a challenge to vote on the other submissions. From all the suggestions I read on this forum only a few are willing to give up that power. Can someone that disagrees with me please educate me why is the current system better than what I propose?
 
Machiavelli would agree with me that the first step is to not allow participants in a challenge to vote on the other submissions. From all the suggestions I read on this forum only a few are willing to give up that power. Can someone that disagrees with me please educate me why is the current system better than what I propose?
Might banning participants from voting encourage those who do participate to have multiple accounts so that they can still influence the voting? Some people will be devious enough to want to try and use any system that is introduced to their own advantage. If they manage to win then their victories are hollow as they are not good enough to win without some form of cheating. Very sad.
 
anisah wrote:
If they manage to win then their victories are hollow as they are not good >enough to win without some form of cheating. Very sad.
This is really the crux of the situation. There is nothing to "win" here. No prestige, no trophy, no money, only the knowledge that some people liked the picture you made.

Also, wouldn't disallowing participants voting rights seriously reduce the number of those who vote? Would we even have anyone voting at all? The number of those who vote is already small.
--
Remember, you can post here because they died over there.
 
Machiavelli would agree with me that the first step is to not allow participants in a challenge to vote on the other submissions. From all the suggestions I read on this forum only a few are willing to give up that power. Can someone that disagrees with me please educate me why is the current system better than what I propose?
Might banning participants from voting encourage those who do participate to have multiple accounts so that they can still influence the voting? Some people will be devious enough to want to try and use any system that is introduced to their own advantage. If they manage to win then their victories are hollow as they are not good enough to win without some form of cheating. Very sad.

--
"The good man does not use others as a tool." (Kung-fu-tse)
"The only thing certain about life is uncertainty." (Rabbi Berel Wein)
"We must learn to love the man who differs from us in opinion." (Swami Vivekananda)
"Holding on to anger is like grasping a hot coal with the intention of throwing it at someone else - you are the one who gets burned." (Buddha)
Though I am not a tech wiz, you would be making it more difficult. No more easy .5 votes intended to hurt... There is a solution with minimal effort from DPREVIEW, post just the usernames who voted on the finished challenge. Do it without notice. A pattern will emerge...

Sometimes I can tell someone's shot because they have a voice, shoot at the same location, or are dedicated to a particular type of photography. I suspect I am not alone noticing this. It can play out in a positive or a negative way.

I assure you eliminating voting on the same challenge you participate in is essential. If someone with a winning shot received three .5 votes he or she will know it did not come from the participants. This way any antagonism between challenge participants is eliminated.
 
I assure you eliminating voting on the same challenge you participate in is essential. If someone with a winning shot received three .5 votes he or she will know it did not come from the participants. This way any antagonism between challenge participants is eliminated.
I'll speak for myself: I care not to know who voted 0.5 on my winning shot, where the 'who' here is not really a 'who'. If participants can vote, the 'who' is simply a category (participant, or non-participant.) If participants cannot vote, the 'who's remain unknown, but we'd know for a fact they're all non-participants. That's of little consequence.

Most importantly: if participants' ability to vote is taken a way, I'd have far, far less incentive to participate in challenges. I'd feel a sense of stake has been taken a way.

In my years here, I have not felt antagonism toward other participants, for the simple reasons explained above. I did feel baffled by few challenges' results. But then the results are really non-deterministic. I have gone as far as creating a challenge specifically to prove that point.
 

Keyboard shortcuts

Back
Top