Author | Thread |
|
02/08/2005 12:27:10 PM · #1 |
This is mine.
Sometime we enter a photo that is do mundane that the votes all seem to say "not bad but not great", in the above photo we can see this in the very narrow spead of votes that it go, almost exactly half were 5s.
For me I wuold rather get scores spread out all over the place then have no one really like it and no one really hate it.
This can be looked at mathematically, simply take the standard deviation of your scores, the smaller the standard deviation in general the less interesting the photo. The photo I posted here has a standard deviation of 1.0453.
What is your most mundane photo? Here is a link to the spread sheet I put together to calculate the standard deviation, just enter the number of votes for each score.
//www.sewcon.com/tools/std.xls |
|
|
02/08/2005 01:10:52 PM · #2 |
Mine was oddly enough in the mundane challenge:
Edit: And I wish they would add SD to the summary stat. I posted this as a suggestion a while back.
Message edited by author 2005-02-08 13:13:57. |
|
|
02/08/2005 01:44:35 PM · #3 |
Using the calculator provided, it is apparently this one with a standard deviation of 1.1334:
I didn't run all of my photos, yet. But I found that those that scored near a 5 average were likely to come closest to 1.x
Question: Are you calculating the standard deviation in relationship to 5 or to 5.5? |
|
|
02/08/2005 02:19:40 PM · #4 |
The standard deviation should be applied to a score of 5.5; there's no zero in challenge scoring, so the mean between 10 and 1 is 5.5...
Robt.
|
|
|
02/08/2005 02:29:50 PM · #5 |
Originally posted by bear_music: The standard deviation should be applied to a score of 5.5; there's no zero in challenge scoring, so the mean between 10 and 1 is 5.5...
Robt. |
The stanard deviation is not dependent of the starting point of the numbers, only the spead. So for example a vote of 3 3s 3 4s and 1 5 would have the same stanard deviation as 3 4s 3 5s and 1 6. |
|
|
02/08/2005 02:30:07 PM · #6 |
Originally posted by bear_music: The standard deviation should be applied to a score of 5.5; there's no zero in challenge scoring, so the mean between 10 and 1 is 5.5...
Robt. |
That makes sense. I looked at the formulas in the Excel sheet, but I've never learned the mathmatics of std. dev. I've only understood that it expresses deviation from the norm. In this case, I assume, the norm would be the average of 5.5 - but I wondered if this is actually expressing the average vote, in other words the number of 5's voted (vs. 6's or 4's)? |
|
|
02/08/2005 02:30:17 PM · #7 |
Sure we' re not talking mean, median, and mode here? |
|
|
02/08/2005 02:31:56 PM · #8 |
Originally posted by KaDi: Using the calculator provided, it is apparently this one with a standard deviation of 1.1334:
I didn't run all of my photos, yet. But I found that those that scored near a 5 average were likely to come closest to 1.x
Question: Are you calculating the standard deviation in relationship to 5 or to 5.5? |
As I said in a reply to bear_music the standard deviation is independent of the voting range, it only depends on the votes. |
|
|
02/08/2005 02:34:28 PM · #9 |
In the script provided by scottwilson, the standard deviation is (correctly) computed based on the actual scores so it measures not the deviation from 5, or 5.5, but from whatever average score you've got. This however poses a problem with interpretation of this number: I agree that if your score is 5 and std is 1, the shot is probably mundane. But if your average is, say, 9 and std is 1, it is probably a helluva shot. |
|
|
02/08/2005 02:34:30 PM · #10 |
OK, scott. Math isn't my strong point by a long shot. I'll bow to your expertise. For some reason I thought we were calculating deviation from the mean of the scoring range, which is how I'd define mundane. Using the correct definition of standard deviation (as you have expressed it) then an image that scored all 2's would be a standard deviation of 0.00 right? And this would not be a mundane image, it would be an awful image..
Or am I misunderstanding again? I am presuming the "most mundane image" would be the one that had the most 5's and the fewest deviations above and below that score, basically...
Robt.
|
|
|
02/08/2005 02:36:19 PM · #11 |
Originally posted by drydoc: Sure we' re not talking mean, median, and mode here? |
I do put out the mean as well but the main number I was after was the standard deviation, this is a measure about how much disagreement between all the voters as to the score for your photo.
If every one gave you a score of 7, for instance then, the standard deviation would be 0, there is no disagreement between voters, if on the other hand you get a lot of 1s and a lot of 10s you will have a high standard deviation.
|
|
|
02/08/2005 02:37:23 PM · #12 |
Without doing the calc, I'm pretty sure it'd be this one:
It has an almost perfect bell curve around a median score of 5. It's also one of my favourite challenge entries.
|
|
|
02/08/2005 02:40:58 PM · #13 |
So, bear_music has a point: the best way to measure mundaneness (?) of the shot is to calculate the deviation from 5.5
For those who understand LaTex: \sqrt{\sum(vote_i-5.5)^2/N} where vote_i are individual votes and "N" is the total number of votes
Message edited by author 2005-02-08 14:47:06. |
|
|
02/08/2005 02:42:24 PM · #14 |
1.1741
Tired to check more (already done 20 or something)
My average is pretty mundane :-(
|
|
|
02/08/2005 03:38:51 PM · #15 |
Based on the voting profile, easily the most "mundane" of my images: only 35 of 178 votes are NOT 4, 5, or 6.
1... 0
2... 5
3... 11
4... 46
5... 52
6... 45
7... 15
8... 1
9... 1
10.. 2
Robt.
Message edited by author 2005-02-08 15:40:33.
|
|
|
02/08/2005 03:45:44 PM · #16 |
Would not it be a lot more interesting to see the LEAST mundane photos? |
|
|
02/08/2005 04:10:47 PM · #17 |
Originally posted by ltsimring: Would not it be a lot more interesting to see the LEAST mundane photos? |
I wonder. This is my, statistically, most mundane:
followed by this one:
I wonder who the most mundane voter is.
|
|
|
02/08/2005 04:14:21 PM · #18 |
Standard Deviation of 1.18
Interesting thread, thank you, Scott. |
|
|
02/08/2005 05:03:46 PM · #19 |
I think my shot
deserves a "gray" ribbon: 1.1323 !
|
|
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: 08/09/2025 10:24:01 AM EDT.