| Author | Thread |
|
|
12/02/2005 03:27:16 PM · #1 |
I've noticed that several great shot on the site get 1-2 from some voters. Now the only way to explain this is Trolls, those who vote other photos low to boost thier own performance.
I propose a use of a weighted score for each voter. Where the final vote is determined by the mean or mode values of a users voteing pattern. Therefore, if a troll runs through voting every photo a 2, he'd basically be voting every photo a ten.
I'm sure if enough thought were put into a basic algorithm, a nice weighted vote scheme could easily be implemented.
|
|
|
|
12/02/2005 03:30:07 PM · #2 |
if someone votes all 2's or 3's, there votes dont count, this is noticed my either SC or the fathers, as for people who vote low to try and boost their score, it doesnt work, ther score is going to stay the same.
|
|
|
|
12/02/2005 03:30:15 PM · #3 |
There's already something like this in place. If someone votes all one's for instance, their votes are thrown out.
|
|
|
|
12/02/2005 03:46:25 PM · #4 |
But how about if they vote all 5-6's ...
What I'm kinda talking about is, if the highest vote they vote is a 7 for instance, shouldn't that count as a 10?
|
|
|
|
12/02/2005 03:50:50 PM · #5 |
Originally posted by fotomann_forever:
But how about if they vote all 5-6's ...
What I'm kinda talking about is, if the highest vote they vote is a 7 for instance, shouldn't that count as a 10? |
No. A ten means something different to everyone. To this person the best image in the challenge only deserved a 7.
Message edited by author 2005-12-02 15:51:34.
|
|
|
|
12/02/2005 03:51:50 PM · #6 |
Although they would be unduly hard to find, there are numeruous threads on essentially this same "issue." It all comes down to this:
- The voting patterns adhere to a statistically gaussian, or normal, distribution, to a very high degree
- The expected variation in voting will result in some high and some low votes on almost every image.
- There is an automated system in place to weed out votes from those who are voting disruptively.
- The final voting totals show no sign of any significant deviation from expected patterns, indicating that the true "troll vote" is a rare thing indeed, and need not be worried about.
|
|
|
|
12/02/2005 03:55:16 PM · #7 |
OK, both good explanations, I'm satisfied :-)
|
|
|
|
12/02/2005 03:58:10 PM · #8 |
| Remember that one six among hundreds of votes doesn't harm your score practically at all. Actually not even a one does. Let's say you've got two hundred votes with avg 5.2. One person gives you a one, you're still at ~5.18. Make it ten ones, and you're down to 5. As the automatic system already "weeds out disruptive votes" (thank you kirbic) you ain't going to get that more than one or two "troll votes" anyway. |
|
|
|
12/02/2005 04:01:18 PM · #9 |
That explains why I always notice a jump in my score from the last update to the after challenge average?
|
|
|
|
12/02/2005 04:02:03 PM · #10 |
Originally posted by cpanaioti:
Originally posted by fotomann_forever:
But how about if they vote all 5-6's ...
What I'm kinda talking about is, if the highest vote they vote is a 7 for instance, shouldn't that count as a 10? |
No. A ten means something different to everyone. To this person the best image in the challenge only deserved a 7. |
Also, I use 5,6,7 as my first-pass sorting. That way if I don't get back to fine-tune my votes all are still relative to a wider scale. If I do get back, 5's almost always become 1-5's, 7's become 7-10's and, oddly, 6's tend to stay as 6's. I don't see anything wrong with using a rough scale and refining it if possible.
|
|
|
|
12/02/2005 05:01:46 PM · #11 |
Originally posted by KaDi:
.........5's almost always become 1-5's, ...... I don't see anything wrong with using a rough scale and refining it if possible. |
The only time this sort of thing bothers me is when I see my score go diving in the last couple of hours of a challenge.
I always yell at my monitor: "Why the hell couldn't you do this 3 days ago..... it's nearly over, LEAVE ME ALONE, YOU MORON!!!"
Message edited by author 2005-12-02 17:02:17. |
|
|
|
12/02/2005 09:58:11 PM · #12 |
Originally posted by Beetle:
Originally posted by KaDi:
.........5's almost always become 1-5's, ...... I don't see anything wrong with using a rough scale and refining it if possible. |
The only time this sort of thing bothers me is when I see my score go diving in the last couple of hours of a challenge.
I always yell at my monitor: "Why the hell couldn't you do this 3 days ago..... it's nearly over, LEAVE ME ALONE, YOU MORON!!!" |
I *know* you weren't calling me a moron...but I would like to point out that I absolutely refrain from changing scores while I wait for the roll-over! I'm way too tired and would be far too rushed to tweak anything at that point. If I have any energy left I try to comment on photos--usually I just make wise-cracks and meaningless posts in the forums, though. Also, moving a vote from 5 to 4 (or even a 3) at that late time in the voting with so many votes being averaged in would be barely apparent, registering as a dip in the .00x range--the dive in votes are probably due to new votes, not adjusted ones.
|
|
|
|
12/02/2005 10:07:46 PM · #13 |
Originally posted by KaDi:
I *know* you weren't calling me a moron...
.....the dive in votes are probably due to new votes, not adjusted ones. |
I know you're correct with the first part, and you're probably right with the second one, too....... clever girl! |
|
Home -
Challenges -
Community -
League -
Photos -
Cameras -
Lenses -
Learn -
Help -
Terms of Use -
Privacy -
Top ^
DPChallenge, and website content and design, Copyright © 2001-2025 Challenging Technologies, LLC.
All digital photo copyrights belong to the photographers and may not be used without permission.
Current Server Time: 08/31/2025 04:23:26 AM EDT.
