DPChallenge: A Digital Photography Contest You are not logged in. (log in or register
 

DPChallenge Forums >> General Discussion >> Effect of a single low vote
Pages:  
Showing posts 1 - 25 of 73, (reverse)
AuthorThread
05/06/2006 09:47:14 PM · #1
The common zeitgeist at DPC is that one or two low votes have no significant effect on your score. The following table is a fictitious vote sequence that illustrates the true effect:

Votes avg Cum. Vote
150 6.5000 975 start point
151 6.4834 979 4 new low vote (troll?)
152 6.4868 986 7 high vote
153 6.4901 993 7 high vote
154 6.4935 1000 7 high vote
155 6.4968 1007 7 high vote
156 6.5000 1014 7 high vote

Wow! It took five votes of 7 to recover from a single vote of 4!!
Trolls of the world unite! Destroy good scores with a single vote!

DPC script screws up the table alignment.

Message edited by author 2006-05-06 22:17:23.
05/06/2006 09:56:40 PM · #2
One of my entries has done this several times this week and it usually takes about 12-24 hours to get back to where it was before the troll vote and then bam another very low vote and then the slow creep back up.
05/06/2006 10:06:56 PM · #3
Some troll must have read my post, I just received a vote of 2!
05/06/2006 10:12:55 PM · #4
I suppose the only thing left to do is submit photos that receive 10s.
05/06/2006 10:14:54 PM · #5
Originally posted by ElGordo:

Some troll must have read my post, I just received a vote of 2!


Yep...both of mine just got a 2....thanks, El! LOL!
05/06/2006 10:15:56 PM · #6
Makes you appreciate the 7.5+ submissions even more doesn't it.

bazz.
05/06/2006 10:55:54 PM · #7
If you are after the elusive virtual ribbon, then you will be an unhappy artist.

If you just shoot for yourself, and know the work is good, then the trolls of world will disolve into what I like to think is like "alka-selter" in a shot of Jack Daniels. It will distract the headache, and still make you feel good.


05/06/2006 11:05:01 PM · #8
Originally posted by ElGordo:

The common zeitgeist at DPC is that one or two low votes have no significant effect on your score.


One or two low votes individually have no true significance. In large groups they mean your shot deserves a low score.

Originally posted by ElGordo:

The following table is a fictitious vote sequence that illustrates the true effect:

Votes avg Cum. Vote
150 6.5000 975 start point
151 6.4834 979 4 new low vote (troll?)
152 6.4868 986 7 high vote
153 6.4901 993 7 high vote
154 6.4935 1000 7 high vote
155 6.4968 1007 7 high vote
156 6.5000 1014 7 high vote

Wow! It took five votes of 7 to recover from a single vote of 4!!


I do not understand what you mean by recovering from a vote of 4... The only way this makes sense is if you feel that your shot does not deserve a 4. However you recieved a four which means that it does deserve one. After 5 more votes however, the vote of 4 had no statistical significance...

Originally posted by ElGordo:

Trolls of the world unite! Destroy good scores with a single vote!


Looks to me that your example still shows a very good score. Or are you trying to say that someone is a troll because they believe your shot deserves a four?

Originally posted by ElGordo:

DPC script screws up the table alignment.


You don't even like the sites coding...

Message edited by author 2006-05-06 23:18:57.
05/06/2006 11:17:27 PM · #9
When you get a 1, well its much worse.

Just one of those things.
05/07/2006 12:10:27 AM · #10
I wish there was a way to gaurantee some fairness though, here's what happened to me

Vote...133...5.8200
Vote...134...5.7915
Vote...135...5.7634

If you do the math, votes 134 AND 135 both had to be a 2! That means not only troll voting but two (or more) people voting together to control outcomes. This happened at midhnight

Score now is
Vote...163...5.7730
so recovery is long and slow

Brett

05/07/2006 12:48:49 AM · #11
Originally posted by ElGordo:

The common zeitgeist at DPC is that one or two low votes have no significant effect on your score.

Try this out... If you get one vote of 4 followed by one vote of 8 they exactly cancel each other out to. That means, of course, in this unique statistical situation that one vote of 8 is worth as much as 5 votes of 7. So what? A single vote of 8 counteracts 5 votes of 6 to. Who would be calling that troll voting? The bottom line is this... It is still only worth a difference of 0.0160 whether it be a single 4, a single 8 or 5 votes of 7 or 5 votes of 6 on a 6.5 average after 150 votes have been cast.

It probably does no good to suggest there is no such thing as a troll.

Since you specifically vote basically like most everyone else this is not picking on you. But, Gordon, do you know that a score of 4, that you called a "troll" vote, is barely 1.1 points lower than the AVERAGE score you give ALL images, including the best of the best?

How do you explain that? How do you think that you, and most everyone else at DPC, are affecting other people's average scores about every 4th or 5th vote they cast?

By this logic the average DPC voter is a troll, not some small band of mythological, evil voters we can conveniently blame for everything.

Why don't more people complain about the undeserved 10s they get and how they skew their scores unrealistically upward?

Message edited by author 2006-05-07 00:51:59.
05/07/2006 01:25:32 AM · #12
My feeling is a vote is a vote. It is just an opinion, and I think we all know that little saying about opinions.. =)


05/07/2006 01:47:03 AM · #13
Originally posted by KiwiPix:

I wish there was a way to gaurantee some fairness though, here's what happened to me

Vote...133...5.8200
Vote...134...5.7915
Vote...135...5.7634

If you do the math, votes 134 AND 135 both had to be a 2! That means not only troll voting but two (or more) people voting together to control outcomes. This happened at midhnight

Score now is
Vote...163...5.7730
so recovery is long and slow

Brett


We have all been there. I find it extremely hard to believe that two or more people would be conferring on what shot to "attack". They have no idea if a shot is doing good or not, if ruining the chances for a shot. They have no idea who is behind each shot in a challenge if it is a personal thing.
My chrome shot did really well, my best result to date. During voting it took a few dives, but I can't imagine people sending mails or pm's to each other saying: Hey, there is a shot in chrome called sleek, I have a feeling it is doing good, let's give the owner that sinking feeling....
I do agree that it sometimes looks strange, with rather big dives after 150 votes or more and then that very slow climb up again. I might be very naîve ... are there trolls?
If there are ......' . substr('//www.dontfeedthetroll.de/images/dftt.gif', strrpos('//www.dontfeedthetroll.de/images/dftt.gif', '/') + 1) . '
05/07/2006 04:08:36 AM · #14
I think what El Gordo is trying to say is that every vote is important. Frequently people make posts in these forums that diminish the significance of an individual vote, or a small group of votes, in essence saying that they don't have much effect on the final score. If I am reading him correctly, El Gordo thinks that is not the case. I think he is correct, each vote is important and we should recognize how much each can effect scores, and the order of finish.
05/07/2006 04:34:37 AM · #15
If the voting is applied equally to all pictures, then it does not matter. ie if one person tends to give lower scores to all photos, it will not really hurt any of them. What is important is that the voter casts his vote relative to the other photos. A four should be not as good as a five. Even if the voter does not give higher than a five. I want to use the scores to try to measure if I am getting better and get frustrated sometimes if a picture does not score well, but I am learning also which comments and scores are meaningful and which are not. I need to be honest with myself as to whether or not it is as good as it could be and how I could do better for next time. This is not always easy, but can be done. We all want to be told all our shots are great. I now accept anything over five as good.
05/07/2006 04:51:34 AM · #16
The issue of strongly straying votes could be easily dealt with by using median instead of average. That would ignore the extremes and give the most frequently given vote as a result. As a downside, you would also lose the lift from the occasional tenner, of course ;-). In summary, as it has been pointed out many times before, I believe, the marks that the average voters give you carry far more weight than an odd stray one.
05/07/2006 06:08:00 AM · #17
Do people honestly think their photos won't be rated as a 4 by some voters? Heck I expect it, and plenty of them : ) Now where's the table that illustrates the effect of a single 10 on your score?

Way way back when I was a lad and had a score in the 8's, every 6 felt like a troll vote and I can't repeat what I thought about the voter who gave me a mere 5 ... Average? Moi? How very dare you.
05/07/2006 10:36:53 AM · #18
To all the naysayers: Of course you will get 4 votes. You will even get 1 votes occasionally. Do they affect your net score? You bet they do, so don't try to rationalize that a few low votes will not affect your final placement! Look at the point spread in the top 20 images, sometimes less than 1/10 of a point.
Once in a while some jackass gives everyone a low vote and that is irrelevent, those votes get thrown out. But the genuine troll doesn't do that. He doesn't want his vote to be thrown out, so he selects only the images that he feels are in the 'potentially winning' category for his spoilsport. The votes of 4 are not out of reason, but it is a very rare entry that rates a 1 or 2. The table cited is for reference to those that believe ocassional low votes do not effect them, even a perfectly reasonable 4 will affect your score.
Does the system work? Of course it does, mostly. But you can't convince me that once in a while a deserving image won't get unfairly displaced because of troll voting.

The example cited at the beginning of this thread is FICTITIOUS (as stated there)! My entries don't score anywhere near that well!

And to TooCool, if an entry has an average of 6.5 then a vote of 4 is clearly on the low side, though still within reason. This table is an example of the effect of that low vote, not a guide for jackasses.
05/07/2006 10:54:53 AM · #19
Is it just me or does all this talk of math give you a headache?

And I like to have trolls to blame my low scores on. But I see what you mean, and they're EVIL.
05/07/2006 10:58:31 AM · #20
I'm sorry, but this grumbling in this whole thread gives me a headache...
05/07/2006 12:02:51 PM · #21
Originally posted by ElGordo:

To all the naysayers: Of course you will get 4 votes. You will even get 1 votes occasionally. Do they affect your net score? You bet they do, so don't try to rationalize that a few low votes will not affect your final placement!


The few low votes that you may get place your photo where it belongs in the rankings... If everyone scored every shot 6 or 7 as in your example, we would all tie at 6.5 and there would be hundreds of ribbons every week...
05/07/2006 12:10:25 PM · #22
Originally posted by ElGordo:

... if an entry has an average of 6.5 then a vote of 4 is clearly on the low side, though still within reason. This table is an example of the effect of that low vote, not a guide for jackasses.


What about the 'errant' high votes? In your example with a 6.5 average, a 4 will be balanced out with a single 9 vote. If the image is a '6.5 image' (which I use ironically, because I don't think it is so objective) then a 9 is just as unreasonably high as a 4 is low. And a 10 will balance a 3. The table could have been written with one 9 vote and then the five 6 votes that it takes to bring the average back down to 6.5.

The voting distributions are remarkably good bell curves, usually. We don't see ditributions like the one in your early example. Where there are a few low votes and then lots and lots of 7s to counteract. For the typical 6.5 there are lots of 5, 6, 7 and 8 votes, with a few 3 and 4s and 9 and 10s in the tails. Then, there is that peppering of 1 and 2s that everybody gets so hot and bothered about. But usually all of the top 10 have some of them, so everybody seems to get dinged by them.

Just my $.02.

Liza

Message edited by author 2006-05-07 12:13:00.
05/07/2006 12:14:44 PM · #23
Would you guys cut it out, I think the voters are experimenting :-P
05/07/2006 12:35:05 PM · #24
My Low Key Shot finished 12th. It had nine 1 votes. Only 5 images in the twelve that finished above mine had any votes of 1 (the most was 2 votes). 116 of 221 votes were 7 or higher.

The nine "punitive" votes were because my image contained 1/2 of a left nipple.

IMHO... That Just Sucks!!!

But i'm told DPC is about learning....so I have learned my lesson and only take colorful pictures of rainbows, flowers (with the obligatory water drops), puppies and children.

: - P
05/07/2006 12:39:53 PM · #25
Originally posted by elru21:

Originally posted by ElGordo:

... if an entry has an average of 6.5 then a vote of 4 is clearly on the low side, though still within reason. This table is an example of the effect of that low vote, not a guide for jackasses.


What about the 'errant' high votes? In your example with a 6.5 average, a 4 will be balanced out with a single 9 vote. If the image is a '6.5 image' (which I use ironically, because I don't think it is so objective) then a 9 is just as unreasonably high as a 4 is low. And a 10 will balance a 3. The table could have been written with one 9 vote and then the five 6 votes that it takes to bring the average back down to 6.5.

The voting distributions are remarkably good bell curves, usually. We don't see ditributions like the one in your early example. Where there are a few low votes and then lots and lots of 7s to counteract. For the typical 6.5 there are lots of 5, 6, 7 and 8 votes, with a few 3 and 4s and 9 and 10s in the tails. Then, there is that peppering of 1 and 2s that everybody gets so hot and bothered about. But usually all of the top 10 have some of them, so everybody seems to get dinged by them.

Just my $.02.

Liza


Think about what you are saying! Are there as many 10s as there are 4s? For the exceptional image there may be as many 7s as there are fours. Frequently the distribution of scores is not a bell curve but is skewed to one side or the other (that's a not necessarily bad) and when a controversial subject appears the distribution may even be bimodal.

I think many of you completely missed the point of this thread; that awareness of how votes impact an entry is useful to all, the voters as well as the entrants. I believe that a few voters vote with little or no thought about the quality or difficulty of producing a good image, but instead vote on the basis of their reaction to the subject of the photo. That is a legitimate criteria if you are selecting an image for publication. But my impression has always been that the avowed purpose of DPC is to cultivate the art and science of photography. Some of the world's most memorable photos have very unpleasant subject content but are nontheless great photos. I suspect they would not fare well at DPC.

Message edited by author 2006-05-07 12:41:13.
Pages:  
Current Server Time: 07/16/2019 12:09:22 PM

Please log in or register to post to the forums.


Home - Challenges - Community - League - Photos - Cameras - Lenses - Learn - Prints! - Help - Terms of Use - Privacy - Top ^
DPChallenge, and website content and design, Copyright © 2001-2019 Challenging Technologies, LLC.
All digital photo copyrights belong to the photographers and may not be used without permission.
Proudly hosted by Sargasso Networks. Current Server Time: 07/16/2019 12:09:22 PM EDT.