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

DPChallenge Forums >> Challenge Results >> New stats - did you notice?
Pages:  
Showing posts 26 - 50 of 54, (reverse)
AuthorThread
 04/23/2008 10:59:52 AM · #26
Originally posted by MelonMusketeer:
... It's pretty evident that people with no camera or experience in photography see the images differently, and are more impressed with most of the images posted in the challenges. ...

From what did you draw that conclusion?

Earlier postings, when the no camera stat was available, complained that the "no camera" voters shouldn't be so critical (meaning low votes) if they don't even have a camera (which was a large assumption as well...).

Everyone that votes the in the "other" open challenge is a "non-participant"...

With all due respect, I'm not saying that you're statement isn't valid - but IMO I don't believe it applies to all, or even many.
 04/23/2008 11:17:26 AM · #27
Yeah! I suggested this change awhile ago.
(I know, I'm probably not the only one who did so.)
04/23/2008 11:19:50 AM · #28
Without going through all my entries, the biggest gap is likely also my highest rated image:

Rhythm III
Place: 1 out of 157
Avg (all users): 7.6794
Avg (commenters): 8.4000
Avg (participants): 7.0357
Avg (non-participants): 8.1600
 04/23/2008 11:34:05 AM · #29
I like the new stat. I looked at my past 20 entries and had to go to the 19th before I found one where non-participants rated higher (edit - LOWER) than participants, and that was only by .03. You can't tell the difference in spread, but you can tell the number of participants vs non-participants from the overall average, participant average, non-participant average and total voters. I never vote on challenges I don't participate in but based on the recent poll most do, and I may be changing my philosophy.
(totally messed up on first post mixing up who voted higher)

Message edited by author 2008-04-24 10:31:55.
 04/23/2008 11:47:00 AM · #30
This is a very interesting new feature.
Those two new numbers tell a story but I doubt that they tell whole story. It is obvious that particapants vote in general lower than non-participants and one might conclude that the delibaretly trying to rank their images higher by doing so. But one could also conclude that low-voters, people that generally vote low, participate more often than high-voters. It would be interesting to see user statistics of relation between average vote and number of challenges entered.
04/23/2008 11:48:46 AM · #31
Since I love numbers and I really like it when they reveal something spectacular, I have to say that there really isn't much sitting in this new statistic other than what many people were thinking.

I looked at my last 58 entries:
Min score = 4.87
Avg score = 5.91
Max score = 7.22

In all but two cases, or in 97% of these, the non-participant vote was higher than the participant vote - as many expected.
The average difference in these 58 cases is 0.34 (and the median is 0.32 so its not skewed in one direction).
There is absolutely NO relationship, overall, between my final score and the difference noted.
There may be small relationships lurking in there with regard to the type of shot required for the challenge, but I am not up for digging that deeply.

So, if everyone else finds a similar result by looking at a statistically significant number of and range of entries, the only diffence in using one value over the other would be that final score would change, but the placement would not.
 04/23/2008 11:53:57 AM · #32
What would make it a truly useful statistic, rather than the moderately "interesting" number it currently is, would be to include the raw number of voters for each category.

As seen now we must assume the vote distribution between the participant/non-participant is equal.

Any debate about voting patterns is moot until you can correctly correlate the number of voters between the groups.

In the good doctor's case for Rhythm III:

Place: 1 out of 157
Avg (all users): 7.6794
Avg (commenters): 8.4000
Avg (participants): 7.0357
Avg (non-participants): 8.1600
Views since voting: 11057
Views during voting: 337
Votes: 131
Comments: 76
Favorites: 49 (view)

Knowing there were 157 participants and only 131 votes, the non-participant number can be anywhere from 0 to 131. Perhaps it's 10. Or 2. I doubt it, but until known the debate is moot.
 04/23/2008 11:58:02 AM · #33
Originally posted by alfresco:
What would make it a truly useful statistic, rather than the moderately "interesting" number it currently is, would be to include the raw number of voters for each category.



i agree that would be useful to include it, but it is actually very easy to determine from the total average score, the participant and non-participant averages and the total number of votes. Algebra is fun!

Edit to include equation:

TV = total votes
Tave = total average score
Pave = participant average score
NPave = non-participant average score

PV = participant votes
NPV = non-participant votes

Equation for determining votes per group:

Participant Votes = PV = TV x (Tave-NPave) / (Pave-NPave)

Non participant Votes = NPV = TV - PV

Message edited by author 2008-04-23 12:13:12.
 04/23/2008 12:12:18 PM · #34
Originally posted by alfresco:
What would make it a truly useful statistic, rather than the moderately "interesting" number it currently is, would be to include the raw number of voters for each category.

As seen now we must assume the vote distribution between the participant/non-participant is equal.

Any debate about voting patterns is moot until you can correctly correlate the number of voters between the groups.

Knowing there were 157 participants and only 131 votes, the non-participant number can be anywhere from 0 to 131. Perhaps it's 10. Or 2. I doubt it, but until known the debate is moot.


One can deduce there were 75 non-participant votes and 56 participant votes.
 04/23/2008 12:16:49 PM · #35
To amplify on Doc Achoo's post, you know:
- Your participant score
- Your non-participant score
- Your total vote count
- Your overall score
Given the above four data points, it is possible to dig out the number of participant and non-participant votes.
 04/23/2008 12:19:16 PM · #36
Here are the equations from my post earlier

TV = total votes
Tave = total average score
Pave = participant average score
NPave = non-participant average score

PV = participant votes
NPV = non-participant votes

Equation for determining votes per group:

Participant Votes = PV = TV x (Tave-NPave) / (Pave-NPave)

Non participant Votes = NPV = TV - PV

Originally posted by kirbic:
To amplify on Doc Achoo's post, you know:
- Your participant score
- Your non-participant score
- Your total vote count
- Your overall score
Given the above four data points, it is possible to dig out the number of participant and non-participant votes.
04/23/2008 12:32:24 PM · #37
Originally posted by kirbic:
To amplify on Doc Achoo's post, you know:
- Your participant score
- Your non-participant score
- Your total vote count
- Your overall score
Given the above four data points, it is possible to dig out the number of participant and non-participant votes.


Correct, and Peter posted the equation for those who forgot their high school algebra. However, it sure would be nice if the site did it for us.
 04/23/2008 12:48:02 PM · #38
Ah, yep, we were both doing the math at the same time, LOL! Peter was faster. I worked up a little excel spreadsheet to parse out the votes & scores from the text data copy/pasted from the site, and verified that for a number of my entries, it really does work. There are a couple caveats:
- In the case where the participant and non-participant averages are identical, the calculation results in a division by zero.
- In cases where the participant and non-participant averages are very close, there is a wide error band on the result.

ETA:
In order to test whether the two distributions of votes are statistically separable, we need to know one additional piece of information for each, and that is the standard deviation. Ideally, we need a separate standard deviation for each group (which we cannot calculate from the given data). We *do* have sufficient information (in the histogram) to calculate the standard deviation of all the votes. We could therefore assume that the S.D. is the same for both, and use the overall as an estimate. This may or may not be accurate. Better would be to break the vote totals in the histogram into participant and non-participant components.

Message edited by author 2008-04-23 12:54:43.
 04/23/2008 12:53:14 PM · #39
In the case when the averages of the participants and non-participants are close or the same, I don't think anyone would really care!!! LOL

Funny stuff - all us math geeks wanting to be first to figure the silly little equation out!

Originally posted by kirbic:
Ah, yep, we were both doing the math at the same time, LOL! Peter was faster. I worked up a little excel spreadsheet to parse out the votes & scores from the text data copy/pasted from the site, and verified that for a number of my entries, it really does work. There are a couple caveats:
- In the case where the participant and non-participant averages are identical, the calculation results in a division by zero.
- In cases where the participant and non-participant averages are very close, there is a wide error band on the result.
04/23/2008 01:20:05 PM · #40
Interestingly I didn't use an equation at all. I looked at the fractions on the scores. A quick spreadsheet from 1/1 to 1/100 revealed what the decimals look like for each # of voters. Seeing the non-participants was 0.1600 I could see it was a multiple of 25. I then simply multiplied the remaining number of votes (131-25,50,75,etc) by the participant score to see which resulted in a whole number. 75 and 56 both resulted in whole numbers so should be the answer. :)

Maybe a little more screwy way to arrive at the answer, but it made more intuitive sense to me and I posted the answer first. :P

Message edited by author 2008-04-23 13:20:15.
 04/23/2008 01:25:03 PM · #41
Y'all are a bunch of geeks. :-) (And I say that with great fondness.)
 04/23/2008 01:25:51 PM · #42
Your a geek for opening the thread! ;-P
Originally posted by Melethia:
Y'all are a bunch of geeks. :-) (And I say that with great fondness.)
04/23/2008 03:25:07 PM · #43
Originally posted by Melethia:
Y'all are a bunch of geeks. :-) (And I say that with great fondness.)


Thank you. (And I saw that with great fondness also.)
 04/29/2008 07:01:21 AM · #44
The users who have put images in a challenge have given more thought to the challenge and into shooting a photo that fits the challenge, at least from their POV about it. It would seem that that alone would tend to make them a little more critical about the images in a challenge, not because they want to hold down the scores, but just that they have a "hands on" connection to the challenge subject, as participants have done a bit of work to produce an image that fits and hopefully will do well in voting.
It's pretty evident that people with no camera or experience in photography see the images differently, and are more impressed with most of the images posted in the challenges.

It could be that way... or it could be that the people who participated in the challenge have more of a closed mind to the challenge interpretation than those who didn't. In any event they are likely to more picky of technical or artistic points which they believe that their own entry has worked hard to overcome (rightly or wrongly).

The average for my challenge entries is about 0.5 difference, ranging from 0.3 to a staggering 0.8 points - all in favour of the non-participants.

Personally I don't vote in challenges that I am competing in since I cannot guarantee objectivity.
 04/29/2008 07:40:45 AM · #45
I've only looked at my most recent entry ("macro" challenge), and my results are dramatically flipped from what y'all are reporting...



Place: 26 out of 258
Avg (all users): 6.2803
Avg (commenters): 7.4286
Avg (participants): 6.3529
Avg (non-participants): 6.1944
Views since voting: 239
Views during voting: 324
Votes: 157
Comments: 20
Favorites: 3 (view)

Participants voted .15 HIGHER than non-participants.... For that matter, commenters voted a LOT higher than the average, much more so than I usually see...

R.
 04/29/2008 08:15:13 AM · #46
I checked all of my scores and in every instance the Particpant vote was lower than the Non Participant - in many cases by a wide margin.

The only conclusion I can draw is the infamous Trolls are probably participants.
 04/29/2008 08:33:52 AM · #47
Originally posted by citymars:
Yeah! I suggested this change awhile ago.
(I know, I'm probably not the only one who did so.)


I suggested it also (success has many fathers). I suggested it in this thread //www.dpchallenge.com/forum.php?action=read&FORUM_THREAD_ID=759514, and in another similar thread.

We all expected participant score to be lower. I think possible reasons are
1. More thought, interest, and critical review of the photograph by the participant.
2. Response to dissapointment in ones score (comparing my photo to theirs).
3. Trolling and trying to place higher.

I think 1 & 2 are for the most part the explanation.

05/12/2008 11:57:18 AM · #48
Originally posted by photodude:
I checked all of my scores and in every instance the Particpant vote was lower than the Non Participant - in many cases by a wide margin.

The only conclusion I can draw is the infamous Trolls are probably participants.


What I find interesting is that the general consensus seems to be that whichever group (participants or non-participants) is voting images higher is therefore voting images more correctly. I think this stat has revealed more about photographer bias than it really has illuminated any supposed voter bias. ;)
 05/12/2008 11:58:55 AM · #49
I also like to add that some of these trolls have figured out that if they vote 3s and 4s their vote will not be eliminated from the rollover. I noticed a lack of 1s and 2s but more 3s and 4s than usual.
 05/12/2008 12:35:53 PM · #50
They came for the participants and banned them, because participants were assumed to be trolls, and I said nothing, because I'm not a troll.

Then they came for the non-participants and banned them, because non-participants were assumed to be ignorant of what it means to create a good shot for the challenge, and I said nothing, because I'm not ignorant.

Then the connection to the old poem breaks down, but we're still left with no one being allowed to vote.
Pages:  
Current Server Time: 06/22/2025 10:02:21 AM

Please log in or register to post to the forums.


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: 06/22/2025 10:02:21 AM EDT.