| Author | Thread |
|
|
02/20/2008 04:58:07 PM · #1 |
It has been interesting around here lately. About year ago I did a spread sheet on the average vote per-challenge that went back a year. At that time the average challenge entry vote was just over 5.5. I just averaged out the last 15 challenges, the average is down to 5.38.
So what's up?
Nothing
Trolls
Challenge Topic
Lower Quality shots
Stronger photographers leaving
More people new to photography (dpc style)
Or is this topic a dead horse that will never ride again....Lol
|
|
|
|
02/20/2008 05:18:49 PM · #2 |
maybe 15 isn't enough? did you do the last year since last time? It may also be interesting to note that if you go to the challenge history and rank the photos by lowest average score, you have to get to the 43rd challenge before you have a challenge that wasn't before 2006. It seems that even though the earlier challenges produced lots of high scoring ribbons. The rest of the field was treated more harshly. Also "The eyes have it challenge" is number 50 something on that list with a 5.0026 average vote with 221 votes. This is really low and has more votes than normal, having it as part of the 15 challenges would skew the results I think.
I would say a combination of "Nothing" and "More people new to photography(DPC style)" Perhaps people who got those cameras for Christmas are slowly starting to try and use them and going through the learning curve. My two cents |
|
|
|
02/20/2008 05:53:33 PM · #3 |
A) The challenge topics have been much less general and that makes them harder. Many people try so hard to meet the challenge they forget to make the picture look good.
B) Smaller #s of entries and smaller numbers of challenges in your samples make deviation from the mean more likely.
C) Although I believe trolls exist, I doubt they can change the average vote cast over 15 challenges.
|
|
|
|
02/23/2008 06:20:19 AM · #4 |
This thread sparked my interest... so I decided to do some investigation into the matter and created this graph from the data in the Challenges History.
(Please note this image is hosted on my own domain... if any members would like to add it to their workshop etc and let me know where it's located on DPC, I'll be happy to edit the URL, which should improve the upload speed etc of the image.)
From the graphs show, there has been no unusual drop is score... if anything, the average score and the lowest score have been on the up... and while the Moving Average lines suggest a drop in recent scores, such drops aren't unusual.
So... in short... I'd say "There's nothing to see here... move along" ;)
|
|
|
|
02/23/2008 06:49:25 AM · #5 |
5.38/5.5=97,8%
A decrease of 2.2% would fit into the 'nothing' category imo.
Besides: I think your missing a category:
"voters getting used to high quality, increasing their standards."
I think there is some evidence of that when you look at the higher scores further back. But a 2.2% decrease in the last year realy means nothing |
|
|
|
02/23/2008 08:08:11 AM · #6 |
Wow  Vlado, sweet graph! This proves to me that Non-Members are TROLLS!!!
|
|
|
|
02/23/2008 08:14:02 AM · #7 |
| ah vlado, the nerd in you arises... hehehehe |
|
|
|
02/23/2008 09:40:07 AM · #8 |
| It's interesting, from just a simple visual perspective; it looks like the gap between member and open challenges is steadily decreasing. Avg scores, and hi/lows, are really merging together gradually. |
|
|
|
02/23/2008 10:47:00 AM · #9 |
I wonder if there's a causal relationship between removing the low-vote nag screen (since the new year) and the latest down trend. That timing might just be coincidence, but as a relative newbe, I can tell you that the nag screen had an effect on my voting when I first got here. I didn't venture into low scores unless I had something to say about it, but I was happy to give a 9 or 10 because those were easy to comment on.
I think I've learned to comment more intelligently since I got here and now I'm more comfortable describing what I don't like in the lower scored shots, but that has been a goal of mine. I wonder if losing the nag screen has eased the way for lower scores.
IMHO, I would love to see "commenters awards" ranked based on the average commenter score. Those don't always line up with the ribbons and I find that the shots with the highest commenters score sometimes have more subtle artistic value that click-and-run voters might not take the time to apreciate. Of course, somtimes a mediocre photo might only get one enthusiastic comment that skews things, but having a system that valued commenters scores with a special place could encourage people to leave comments. Just my 2 cents (and no, I didn't look to see that this has probably already been suggested and discussed ad nauseum-sorry). |
|
|
|
02/23/2008 10:59:46 AM · #10 |
| vlado is my new hero. His graph shows that the overall mean score has been extremely stable over the life of DPC, with a strong trend for the extreme scores to slowly regress to the mean. IOW, the voting mean has been pretty constant with a slowly diminishing standard deviation. Pretty cool stuff. |
|
|
|
02/23/2008 11:44:09 AM · #11 |
| I just think it looks really pretty! (But there is the odd drop-off recently, isn't there?) |
|
|
|
02/23/2008 11:46:11 AM · #12 |
Originally posted by Melethia:
I just think it looks really pretty! (But there is the odd drop-off recently, isn't there?) |
What do you mean by recently? The last couple of millimeters? You could pick many, many similar drops throughout the mean line. Normal statistical variation. It goes up, it comes down. Wash, rinse, repeat. |
|
|
|
02/23/2008 12:14:31 PM · #13 |
Originally posted by Strikeslip:
Wow Vlado, sweet graph! This proves to me that Non-Members are TROLLS!!! |
No, I don't think so. The averages are for the member and open Challenges, not the member and open Voters. I realize that only members can vote on the member challenges, but I think the appropriate conclusion here is that the members are better photographers. It is very clear that the scores in the member challenges are higher than in the open challenges.
--DanW
|
|
|
|
02/23/2008 02:42:55 PM · #14 |
Originally posted by marttila:
I wonder if there's a causal relationship between removing the low-vote nag screen (since the new year) and the latest down trend. That timing might just be coincidence, but as a relative newbe, I can tell you that the nag screen had an effect on my voting when I first got here. I didn't venture into low scores unless I had something to say about it... |
The decline has been shown to be stastically irrelevant, but nevertheless I think this is a good point.
|
|
|
|
02/23/2008 02:56:03 PM · #15 |
Originally posted by wheeledd:
Originally posted by Strikeslip:
Wow Vlado, sweet graph! This proves to me that Non-Members are TROLLS!!! |
No, I don't think so. The averages are for the member and open Challenges, not the member and open Voters. I realize that only members can vote on the member challenges, but I think the appropriate conclusion here is that the members are better photographers. It is very clear that the scores in the member challenges are higher than in the open challenges.
--DanW |
Member challenges are also typically advanced editing, and open challenges are typically basic editing. A quickie analysis I did a couple of weeks ago showed that advanced editing challenges score a couple of tenths higher, regardless of whether they're open or member challenges. |
|
|
|
02/23/2008 03:04:57 PM · #16 |
I don't believe that averages will show up trolls because I think that trolling practice is actually a very small percentage of the total active DPC users. They will just get lost in the averages...
To catch this minority of users (and if they actually exist) requires some monitoring of consistently low voting users and then someone to look at the photographs and determine if it's reasonable or not. Repeat offenders will keep voting above average images as 1, 2 or 3 - i.e. when your opinion is so significantly different from everyone else you're wrong and possibly a troll (in my most humble of opinions that is :))
N
|
|
|
|
02/23/2008 03:13:03 PM · #17 |
Or you could all just accept the fact that voting trolls are the same as other trolls, imaginary creatures. Myths. Products of overactive imaginations.
To whit: no proof that they exist. One has never been spotted and verified in the wild. No hypotheses attributed to troll behavior that can't be explained by simpler, perfectly reasonable causes. Occam's Razor applies. |
|
|
|
02/23/2008 03:20:22 PM · #18 |
Originally posted by strangeghost:
To whit: no proof that they exist. One has never been spotted and verified in the wild. No hypotheses attributed to troll behavior that can't be explained by simpler, perfectly reasonable causes. |
There is no proof because the evidence is not accessible. However, Langdon will have access to it - so just because there is no public proof doesn't mean trolling behaviour does not exist. In fact, from what I've read in these forums, it seems that there are anti-troll (and anti-friend voting etc) mechanisms in the scrubber...so it does exist, it's just not known as to what level.
Here's an alternative and controversial solution: transparency in voting, post end of challenge - i.e. show a list of people who voted on the image with the vote they gave. The thinking goes that with greater transparency users will feel more obliged to be more considered in their voting. Will PM wars break out? Maybe...hence why I doubt it will ever happen...
N
|
|
|
|
02/23/2008 03:27:59 PM · #19 |
Originally posted by Quasimojo:
There is no proof because the evidence is not accessible. However, Langdon will have access to it - so just because there is no public proof doesn't mean trolling behaviour does not exist. In fact, from what I've read in these forums, it seems that there are anti-troll (and anti-friend voting etc) mechanisms in the scrubber...so it does exist, it's just not known as to what level.
Here's an alternative and controversial solution: transparency in voting, post end of challenge - i.e. show a list of people who voted on the image with the vote they gave. The thinking goes that with greater transparency users will feel more obliged to be more considered in their voting. Will PM wars break out? Maybe...hence why I doubt it will ever happen...
|
So what then is a troll? Somebody who votes every image low and their own image high? Can't be done, because we can't vote on our own image. Any voter whose voting pattern is suspicious will be detected and their votes eliminated by the "vote scrubber," the talked about forum code that none of us really knows much about. But still, what is a troll? The most often discussed "evidence" is a great image - ribbon winner, popular, whatever - that in a challenge, receives a number of votes of 3 or less. People cry out in the forums, "who are those trolls who gave that GREAT image such a low vote?" That's really all the evidence I've ever seen offered, and there are MUCH better explanations for such voting behavior than creating a mythical creature. |
|
|
|
02/23/2008 03:34:30 PM · #20 |
Originally posted by strangeghost:
So what then is a troll? Somebody who votes every image low and their own image high? Can't be done, because we can't vote on our own image. Any voter whose voting pattern is suspicious will be detected and their votes eliminated by the "vote scrubber," the talked about forum code that none of us really knows much about. But still, what is a troll? The most often discussed "evidence" is a great image - ribbon winner, popular, whatever - that in a challenge, receives a number of votes of 3 or less. People cry out in the forums, "who are those trolls who gave that GREAT image such a low vote?" That's really all the evidence I've ever seen offered, and there are MUCH better explanations for such voting behavior than creating a mythical creature. |
This has been done to death in a million other threads, but in a nutshell my view is that if there are people voting on those excellent photographs as being 1, 2 or 3 REPEATEDLY then they are a troll...plain and simple. Look, we all have different opinions and views and whilst it's a bit odd to vote a 7+ image as a 1, that's okay. If it happens more than a couple of times then eyebrows have to be raised. This is what I mean by a troll.
N
|
|
|
|
02/23/2008 03:40:16 PM · #21 |
And just to be clear Quasi, this is not meant as an attack on you, just on the definitions in play:
Your definition then, of a troll, is somebody who
1) Votes differently than you do, or
2) Who votes differently than the majority of voters.
If we grant the benefit of the doubt and say that it's #2, a person who has consistently voted low on the site's most popular pictures, is it not possible that the issue is taste and artistic vision? Must it be assigned to a malicious motivation, e.g., the mythical troll?
I agree, it's a matter of definitions. Absolutely. But let's not define into existence a being who we assume must be malevolent to provide an explanation for a behavior that may have other explanations. |
|
|
|
02/23/2008 03:48:37 PM · #22 |
Originally posted by strangeghost:
If we grant the benefit of the doubt and say that it's #2, a person who has consistently voted low on the site's most popular pictures, is it not possible that the issue is taste and artistic vision? Must it be assigned to a malicious motivation, e.g., the mythical troll?
I agree, it's a matter of definitions. Absolutely. But let's not define into existence a being who we assume must be malevolent to provide an explanation for a behavior that may have other explanations. |
Yes, I think it's #2 as you put it. But I'd add:
2) Who repeatedly votes significantly differently that the majority of voters
And for clarity, when I say significantly differently I'm talking at the far end of the scale...1,2 and 3 in particular for images that the vast majority of users have decided is at least a 6 (i.e. a good average photo at worst). It's not about people who don't think the same as me, or anyone else for that matter - it's about people who consistently harshly vote excellent photographs.
It isn't about interpreting the results and assuming malevolence - I think it's quite clear that if someone feels so strongly that otherwise well regarded images are voted very poorly then this very small percentage of users are of a mindset so radically different to every other user here that one has to question the value of their opinion or vote. You can call it cluelessness, lack of taste, different taste, delusion, eccentric/exotic taste, individual, alternative....whatever you like....but ask, what is the value to you or to any DPC voter of someone contributing that kind of opinion or vote when it is consistently so contrary to all other belief.
N
|
|
|
|
02/23/2008 04:05:42 PM · #23 |
I promise this will me my last word on the subject:
You said:
think it's quite clear that if someone feels so strongly that otherwise well regarded images are voted very poorly then this very small percentage of users are of a mindset so radically different to every other user here that one has to question the value of their opinion or vote. You can call it cluelessness, lack of taste, different taste, delusion, eccentric/exotic taste, individual, alternative....whatever you like....but ask, what is the value to you or to any DPC voter of someone contributing that kind of opinion or vote when it is consistently so contrary to all other belief.
Your assumption here is that there is a group of people who are doing this consistently. A group of people who are all the ones who are giving those great images the 1s and 2s, in challenge after challenge. This is an assumption. There is not one shred of evidence that this is true. And those assumed offenders are labeled trolls in your definition. In my view, it is just as likely, (and I believe more likely) that those voters are different in every challenge. It makes sense to me, seems logical and rational to me, that this voting behavior is a reflection of tastes, and tastes are known to vary within a population.
I could add to this a discussion of the unlikelihood based upon the presumed effect: why would a voter repeatedly do this? How would he benefit? In a typical challenge, one individual vote may make a difference of a few hundredths of a point in the final score. To assume that this would make a difference, we'd have to posit a conspiracy, that there are groups of trolls acting in a united fashion to bring down the best images. Now we have a true conspiracy, and again, no evidence, just assumptions built on assumptions.
I just don't see it. I'm a troll agnostic, or rather a skeptic (if you will). Maybe there's some evidence out there that I'm not seeing. Or maybe I have blinders on and am unable to see something that's right in front of me.
I'm still waiting for even a modestly convincing argument. To me, trolls are in the same category as the fairies, angels, goblins and gods. |
|
|
|
02/23/2008 04:17:49 PM · #24 |
| this is great graph, great work Vlado. |
|
|
|
02/23/2008 04:21:34 PM · #25 |
Originally posted by strangeghost:
Your assumption here is that there is a group of people who are doing this consistently. |
Assumption, no....suspicion, yes. And as I said earlier, the evidence to support this theory is with DPC...so if it were me that's where I'd look, not in the averages as posted by someone else earlier.
As to why people do this, who knows or really cares? This isn't about some esoteric belief - this is about testing a very valid idea with empirical evidence.
N
|
|
