Question Re Auto-Archive Feature
loading
DA Photo Contest Forum
 Home          Members     Calendar     Who's On

Welcome Guest ( Login | Register )
        



Question Re Auto-Archive Feature Expand / Collapse
Message
Posted 10/10/2007 2:57:42 PM Post #17337
 

Junior MemberJunior MemberJunior MemberJunior MemberJunior MemberJunior MemberJunior MemberJunior Member
Hi.  I am new to DA and was wondering if someone could help me understand what the term "two standard deviations" specifically means in the info regarding the Auto-Archive Feature "Additionally, the system will monitor your images and if they fall more than two standard deviations below the average for any particular award they are eligible for, they will be automatically archived."
Posted 10/10/2007 3:15:58 PM Post #17338
 

Supreme BeingSupreme BeingSupreme BeingSupreme BeingSupreme BeingSupreme BeingSupreme BeingSupreme Being
Rather than setting a hard line at a particular numerical value, the system will monitor all images in the system that have won awards at the particular gate in question and calculate the threshold for two standard deviations below the mean for that gate.

In common terms, this simply means that if your image is scoring significantly below winning images for that gate (I'll explain gates below), then it will be archived in an attempt to create more votes for everyone.

Gates: There are gates tied to each set of awards. You can see this pdf for a better understanding of the awards that make up each gate. For each 'check' within the gate, the standard deviation computation is performed.


Eric
DailyAwards.com
Posted 10/11/2007 5:03:32 AM Post #17347
 

Junior MemberJunior MemberJunior MemberJunior MemberJunior MemberJunior MemberJunior MemberJunior Member
Thanks Eric.  I think I understand this now :-)
Posted 10/11/2007 2:37:30 PM Post #17357
 

Junior MemberJunior MemberJunior MemberJunior MemberJunior MemberJunior MemberJunior MemberJunior Member
Hi Eric.  Sorry for being a pest, but just want to make sure I really understand...  If, lets say, the Average Winning Score [in any given category on any given day] was 7.5xxx, and an image in that same category, which had the required number of raw votes, but only scored 7.3xxx, would the image in question then be auto-archived?
Posted 10/11/2007 4:28:18 PM Post #17358
 

Supreme BeingSupreme BeingSupreme BeingSupreme BeingSupreme BeingSupreme BeingSupreme BeingSupreme Being
Most likely noStandard deviation is a statistical measure of dispersion.  The calculated StDev value is dependent on the values in the sample.   Two samples could have the same average (mean) yet have wide standard deviations.

Here are four Ten Vote Samples:

Group 1:  7,7,7,7,7,7,7,7,7,7 = 70  Mean=7.0   StDev = 0.00

Group 2:   8,8,8,7,7,7,7,6,6,6, = 70   Mean= 7.0  StDev = 0.816

Group 3:  10,10,9,9,7,7,6,5,4,3 = 70  Mean = 7.0  StDev = 2.49

Group 4:  10,10,10,10,9,9,3,3,3,3 = 70   Mean = 7.0 StDev = 3.46

In all four examples the average rating  is 7.0 but the Standard Deviation of the four samples ranges from 0.00 to 3.46.  WE

_________________________
Profile - Top 24 - All Images



Nikon D300 -  Nikon D70 - Nikkor 20mm  - Nikkor 50mm - Nikkor 24-120mm - Nikkor 28-70mm - Nikkor 85mm - Nikkor 80-200mm - Nikkor 70-300mm - Sigma 400mm

Posted 10/12/2007 6:02:05 AM Post #17362
 

Junior MemberJunior MemberJunior MemberJunior MemberJunior MemberJunior MemberJunior MemberJunior Member
Thanks!!
Posted 10/12/2007 7:52:43 AM Post #17363
 

Supreme BeingSupreme BeingSupreme BeingSupreme BeingSupreme BeingSupreme BeingSupreme BeingSupreme Being
It does depend though!

If the image has collected a lot of votes, then it is no longer being checked at the category level. So, if you are a subscriber, you can check the advanced image stats page to see if it has a 'PASS' on the category level. At the same time, it may have a 'FAIL' at the monthly level and be archived. The image is still eligible to compete for all awards at monthly and below, it simply cannot continue to collect votes (thus it helps to keep the voting pool lean and provide more votes for everyone).


Eric
DailyAwards.com
Posted 10/12/2007 10:08:45 AM Post #17366
 

Junior MemberJunior MemberJunior MemberJunior MemberJunior MemberJunior MemberJunior MemberJunior Member
Thanks to you both for helping me understand.
« Prev Topic | Next Topic »


Reading This Topic Expand / Collapse
Active Users: 0 (0 guests, 0 members, 0 anonymous members)
No members currently viewing this topic.
Forum Moderators: Admin, Zeke Smith, wemor

Permissions Expand / Collapse