| Author | Thread |
|
|
02/09/2008 07:30:33 PM · #1 |
Maybe just a paranoid delusion but I thought that I would mention it anyway.
Next time you entry goes to vote pay close attention to the rates you receive during the first six to eight hours. Once your entry has established its intial norm look for a short time span of very unrealistic low scores that tank your average after which the scores return to somewhat of a more realistic pattern. Such a tactic has nothing to do with the quality your entry it has to do mathematically altering overall scoring on a widescale. Entries that have not been hit by the lowball have a false yet true mathematical advantage.
Such a hunch has all of the makings of a parnoid delusion on my part but in today's world nothing is beyond possibility.
I hope I am wrong.
louisp
|
|
|
|
02/09/2008 07:32:10 PM · #2 |
can you repeat that in English?
R.
|
|
|
|
02/09/2008 07:33:22 PM · #3 |
I think I'll pass on what he's having.
|
|
|
|
02/09/2008 07:38:20 PM · #4 |
|
|
|
02/09/2008 08:20:39 PM · #5 |
Firing up the universal translator :)
I think he is saying this:
Hours (1-6): Normal voting pattern- your score will be around your average (say 5.45)
Hours (6.01-8): Average plummets due to abnormal number of low votes.
Hours (8.01- End): Votes received would be like the first 1-6 hours, but with the low votes from hours 6-8 you won't recover your previous average.
Did I come close?
|
|
|
|
02/09/2008 08:25:19 PM · #6 |
Unfortunately, it doesn't work that way.
In some cases hours 1-6 yield the lowest votes during the challenge
Hours 6.01-8 yield somewhat higher scores
Hours 8.01 to the end yield the adjustment votes (some up, some down) and comments |
|
|
|
02/09/2008 08:39:06 PM · #7 |
| ...just a paranoid delusion |
|
|
|
02/09/2008 08:39:36 PM · #8 |
Originally posted by cpanaioti:
Unfortunately, it doesn't work that way.
In some cases hours 1-6 yield the lowest votes during the challenge
Hours 6.01-8 yield somewhat higher scores
Hours 8.01 to the end yield the adjustment votes (some up, some down) and comments |
Hmmm...I am in Australia...so I am possibly considered upside down compared to you lot...and my scores reflect the upside version of what you say. Now...if only we could find a medium and only have the high scorers...and we would all be happy...wouldn't we???? Hmmmm....!
|
|
|
|
02/09/2008 08:45:37 PM · #9 |
I've had the full mix, Images that started high and stayed there, images that started high and tanked, mostly images that start lower than I expect and stay there.
I did have one that was a very low 6 after 40 votes and ended up over 7.
One thing is for sure, I've never been able to predict how a score will go during voting. ;-) |
|
|
|
02/09/2008 09:54:38 PM · #10 |
A tactic as described in my OP can be best described as voting not for but only against entries. Such a scheme has been seen in non-dpc challenges. In that contest the simplistic method of voting only positive or negative made it very easy to employ and more importantly detect a scheme of negative voting but regardless, the prinicpal remains the same, that being to unfairly influence the mathematical result.
In the case of dpchallenge uncovering lowball schemes could be much more difficult to uncover due to the high number of voters, entries, 1 to 10 range and normal voting range irregularity. Remember the purpose of the scheme is to decrease the average just enough to effect the outcome. To grastically alter the average would be too obivious.
Here are a few additional thoughts.
a.) For a lowball scheme at dpc to be effective multiple persons would have to be involved.
b.) Logic would suggest that lowball votes placed throughout the voting periods would make detection difficult unless each voter and their votes can be identifed over multiple challenges.
c.) Logic would also suggest that lowball votes placed in a limited period of time as suggested in my OP would make detection much easier. Now with that said, identifing undisputable lowball patterns will remain difficult unless each voter and their votes can be identifed over multiple challenges.
In conclusion, will what I have supposed uncover anything, most very probably not. In fact it would be best for us all if I am incorrect. Does it give us something to keep in mind as we enter challenges throughout the world wide web, yes it does.
louisp
Message edited by author 2008-02-09 22:06:34. |
|
|
|
02/09/2008 10:15:09 PM · #11 |
Okay, here's my take on it.
Initial votes are by people with high level of interest in the challenge. They tend to vote very high on good photos, and very low on everything else. A little later comes the voters who are not so emotionally attached to the challenge. Votes tend to moderate. Towards the end of the contest, people adjust their votes, and voters at this point just want to see good photos, both making the votes climb a little at this point.
I don't think there are any schemes, just different groups of voters. |
|
|
|
02/09/2008 10:29:16 PM · #12 |
Originally posted by louisp:
Maybe just a paranoid delusion but I thought that I would mention it anyway. |
Have you ever noticed that the histogram of scores is usually very normal? It usually will have a single central peak with a fairly equal distribution tailing to either side. This is a good sign that lowball schemes are either non-existent, so small that they have no effect or theyĆ¢re balanced out by voters that only use the top two thirds of the scale. If they had any significance they would likely show-up in the histogram as a skewed or bi-modal distribution.
Of course the exception to this is average scores that are on the very high or very low end of the scale. They're skewed by the fact that the distribution can't go beyond the hard limits of 1 and 10. So often a high score is skewed to the right and low score is skewed to the left.
Based on this I vote for paranoid delusion.
|
|
|
|
02/09/2008 11:12:03 PM · #13 |
Maybe the time of day in the locations where the voting is taking place has a bearing on the perceived phenomenon, since all the challenges begin at midnight, US eastern time.
It's mid day on the other side of the world at that time, so the first hours are the up late US voters, and the up early Pacific Rim voters and the mid day European/Mid East groups. The cultural differences in the voters on line at any one time may be showing in the votes.
After the first 12 to 20 hours or so, then things tend to average out and the score levels off.
Message edited by author 2008-02-09 23:13:42.
|
|
|
|
02/09/2008 11:39:17 PM · #14 |
| I'm far too obivious to grastically alter my average. |
|
|
|
02/09/2008 11:42:54 PM · #15 |
| I have noticed that at about twelve minutes past rollover There seems to be six votes that seem to knock my entries right down. then it starts to climb again slowly! But what i would like to know is how these six voters manage to vote so quickly on a huge challenge such as free style I know they only have to vote 20% to count but that is a lot of images! |
|
|
|
02/09/2008 11:57:16 PM · #16 |
Originally posted by sulamk:
But what i would like to know is how these six voters manage to vote so quickly on a huge challenge such as free style I know they only have to vote 20% to count but that is a lot of images! |
Aren't all votes counted as soon as they are made then scrubbed if the 20% is not hit?
Yes, I meant not. Thanks CP
Message edited by author 2008-02-10 08:27:17. |
|
|
|
02/10/2008 12:25:18 AM · #17 |
Originally posted by DJWoodward:
Originally posted by sulamk:
But what i would like to know is how these six voters manage to vote so quickly on a huge challenge such as free style I know they only have to vote 20% to count but that is a lot of images! |
Aren't all votes counted as soon as they are made then scrubbed if the 20% is hit? |
Yes. Votes count immediately and are only scrubbed at rollover if the 20% is not met (I'm sure you meant 'not'). |
|
|
|
02/10/2008 12:32:54 AM · #18 |
| Ok people, just move along now, there is nothing to see here ....... |
|
|
|
02/10/2008 01:11:29 AM · #19 |
Originally posted by jbsmithana:
Ok people, just move along now, there is nothing to see here ....... |
These aren't the droids we're looking for. |
|
|
|
02/10/2008 03:27:23 AM · #20 |
| I tracked my scores fairly closely over more than 20 challenges and, looking back now, I don't see any pattern as suggested by the OP. |
|
|
|
02/10/2008 07:51:12 AM · #21 |
Friends,
Another member has provided us with an alternate description of the issue I presented in my OP while two others have more importantly provided what could be logical explanations to both suppositions. Thank you all for commenting.
In closing, a "Paranoid Delusion" looks like a distinct possiblity yet I haven't been to a rock concert in years! Uh, even when I did go to concerts I never inhaled!
louisp
ALTERNATIVE DESCRIPTION:
Originally posted by sulamk:
I have noticed that at about twelve minutes past rollover There seems to be six votes that seem to knock my entries right down. then it starts to climb again slowly! But what i would like to know is how these six voters manage to vote so quickly on a huge challenge such as free style I know they only have to vote 20% to count but that is a lot of images! |
LOGICAL EXPLANATIONS:
EXPLANATION ONE:
NOTE: Explanation One directly addresses the supposition outlined in the "Alternative Description"
Originally posted by cloudsme:
Okay, here's my take on it.
Initial votes are by people with high level of interest in the challenge. They tend to vote very high on good photos, and very low on everything else. A little later comes the voters who are not so emotionally attached to the challenge. Votes tend to moderate. Towards the end of the contest, people adjust their votes, and voters at this point just want to see good photos, both making the votes climb a little at this point.
I don't think there are any schemes, just different groups of voters. |
EXPLANATION TWO:
Originally posted by DJWoodward:
Have you ever noticed that the histogram of scores is usually very normal? It usually will have a single central peak with a fairly equal distribution tailing to either side. This is a good sign that low-ball schemes are either non-existent, so small that they have no effect or theyĆ¢re balanced out by voters that only use the top two thirds of the scale. If they had any significance they would likely show-up in the histogram as a skewed or bi-modal distribution.
Of course the exception to this is average scores that are on the very high or very low end of the scale. They're skewed by the fact that the distribution can't go beyond the hard limits of 1 and 10. So often a high score is skewed to the right and low score is skewed to the left.
Based on this I vote for paranoid delusion. |
Message edited by author 2008-02-10 07:55:43. |
|
|
|
02/10/2008 08:47:18 AM · #22 |
this is why voting takes place over a week's time. There are groups of voters that could be classed by peak usership hours for their time zone. In the end however, it all works itself out pretty well.
There would be no real point to anyone running any sort of vote altering system all at once, especially with ISP monitoring and other things that are in place by the moderators of this site. Hence this data is all likely badly skewed.
With a very low number of votes, the score reflected will always be much more susceptible to skewing both up and down. Therefore it is far more likely that you would notice a strong shift in the earlier stages of voting as the vote "normalized" to whatever it will eventually land at. If there was no strong shift, you wouldn't notice anything.
No conspiracy here. This can further be borne out by watching the scores threads during voting. Some are experiencing upswings, others are 'tanking'. |
|
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: 08/25/2025 10:41:43 AM EDT.
