![]() ...
·
![]() |
---|
I logged in to my account this morning and just happened to notice that my Image Index went down from 93 to 39? Not that it really matters but I was just curious. I had noticed that since I had participated in a group (Observable Photon's Group) that took a group collaboration of the Helix nebula. That participation caused my Index to jump a large number due to a very large viewing of the image by a lot of people. That is what shot it up a lot and into the 90's. Now it shot back to just under 40 in one day. Again, the index is nice and I watch it a little just to see the response of others to my work. That being said I'm no publicity hound so I certainly don't loose any sleep over it but the geek engineer in me always wants to ask "Why" when I see something like that. CS |
![]() ...
·
![]() |
---|
Non-inflationary algorithm, maybe? If you post something after that huge jump back to your previous rate then your index can fall although not as sharply in my experience but then I never had very large numbers all of the sudden.
|
![]() ...
·
![]() |
---|
The Index can drop, the way the calculation is, but I don't think I've ever seen it drop that much. The index calculation is very stable tho, and I'm pretty sure at this point there are no edge cases or bugs (tho I know one should never say that!) The reason why the index drop is that it takes into consideration only the images with a higher number of likes than your average number of likes. So outliers can cause havoc with your index, as they raise your average number of likes, and cause images to drop off the calculation, therefore actually lowering your index. In other words, the index rewards consistency. I'll check the logs in a bit to see if something is wrong, tho. |
![]() ...
·
![]()
·
3
likes
|
---|
The algorithm in the Image Index seems to contain an inherent flaw in that it applies too much weight to "likes" versus number of total images. This explains your large upward jump with just one image. The algorithm also causes single new members to sometimes immediately jump to the top of the list with only a handful of popular images, mostly from large group situations and often repeated postings of the same data repeatedly winning the same awards. I would say the aberration is the jump up more than the fall back. |
![]() ...
·
![]()
·
1
like
|
---|
@Phil Hoppes, I know this is probably not what you'd like to hear, but what I think happened is that the image you are a collaborator of got more likes, and made it so your average went over a threshold that made a lot of your images not be calculated in the index anymore, which made it drop. Unfortunately, I don't have the bandwidth to pinpoint this accurately (I would need to write a bunch of code to run simulations at different times, corresponding to each recent like to all of your images) but I'm quite confident in this theory. It does sound like the the 93 number might have been over-inflated but a specific lucky combination of numbers. I wish I could provide a more specific explanation! |
![]() ...
·
![]() |
---|
Salvatore Iovene: I have seen this happening with collaborations, especially when the other collaborating party decides to pull their image or even their full content from Astrobin. All the points for "likes" from that pool are deleted from your Index... |
![]() ...
·
![]() |
---|
Dear @Phil Hoppes @Salvatore Iovene and @Georg G Albrecht are right. In your precise case, I believe that your index (just as mine) went down as Andrea deleted its version of the Helix Nebula, where we were appearing as a collaborators. The likes corresponding to this image were cancelled and our indexes were adjusted accordingly. That is fair enough. CS Patrice |