Grattis till lottovinsten!

Jag gillar att fota.
Jag gillar själva bildskapandet. Tekniken, siffrorna och prylarna lockar mig inte det minsta, men att göra en bild, att andra tycker om den, det är roligt.
Jättekul.

Få saker är så utvecklande för ens eget fotograferande som feedback från andra bildintresserade människor och att titta på andra duktiga människors bilder.

Därför gillar jag Flickr.

Flickr har en funktion, “Explore” där man med hjälp av någon sorts formel väljer ut 500 bilder varje dag, och att hitta sin bild bland dessa fem hundra är självklart roligt.
Det är SHITLOADS med bilder som laddas upp på flickr varje dag. Flera miljoner.

Så självklart blir man glad när det händer.

Men samtidigt konstaterar jag att det finns ett konstigt mönster i mina explore-bilder.

Får jag en bild som explore, får jag ofta en till någon eller några dagar senare.
Har jag varit inaktiv, och sedan laddar upp en bunt bilder är chansen till explore högre än om jag nöter på och laddar upp bilder regelbundet.

8 augusti 2012

16 augusti 2012

31 mars 2013

2 april 2013

16 Augusti 2013

17 augusti 2013

23 februari 2014

3 mars 2013

Den noggranne konstaterar att det ändå är några dagar mellan de sista två bilderna. Ja. Men jag har heller inte lagt upp några bilder, så då är det svårt att få en explore.

Jag undrar verkligen hur den där algoritmen ser ut…

Det ser ju onekligen ut som om den styrs mer av hur min aktivitet på flickr ser ut, än av “kvalitén” på de bilder jag laddar upp..?

  • Katinka Bille Lindahl

    …och samtidigt är dina bilder väldigt, väldigt bra! Jag har också märkt att en Explore gärna följs av en till. Har haft nio eller tretton tror jag. Men jag har aldrig tänkt så långt som du och analyserat saken. Intressant! Visst blir man fruktansvärt glad och smickrad. Ingen algoritm i världen kan ta ifrån dig att dina bilder är väldigt bra, och du har förtjkänat lottovinsten. :-)

  • OOF

    Flickr Explore explained: Skoj, men ytterst slumpmässigt. och inte alltid “rätt” bilder som kommer dit, blir ofta själv förvånad över vilka av mina bilder som hamnar på Explore, oftast inte de förväntade…. intressant läsning!

    Most people quickly realise that Flickr Explore is not a system designed to reward photographic excellence. Nevertheless, people come to desire the intense exposure an appearance in Explore often offers. It is this desire and people’s naturally competitive nature that leads to frustration with Explore. The usual line of thinking is: if exposure in Explore is desirable it should go to winners not just any participant. The fact of the matter it does go to winners, just not the kind of winners people think should win. It is clear that Flickr is not interested in providing a venue for amazing photographers to showcase their wonderful works, but rather it offers a system that exposes people to a broad spectrum of the work that finds its way into Flickr each day. As part of a larger a social network, Explore rewards good citizens not great photographers.

    Things we know to a high degree of certainty about Explore:
    1. Explore does not see the quality of your photos.
    2. Explore does not select images based on input from any Flickr representative; however, this does not negate the probability that a human reviews/edits the Explore list before going live.
    3. Images are selected for Explore by a simple, but ever changing, math equation that factors in Favourites, Comments and Views as well as the rates each accumulate.
    4. Explore’s current math equation is heavily weighted to Favourites and a high ratio of Favourites to Comments.
    5. Explore ignores any images not marked as “safe”.
    And the big one…
    6. You will be banned from Explore if your images are showing up in Explore too often.

    The list of factors below contains things commonly thought to be involved with Explore. Everyday you will find examples that call into question each of these statements. It is possible the individual impact of each factor waxes and wanes in a deliberate way to keep people from gaming the system to any level of precision. It is just as likely these things are akin to urban legends and have no foundation in reality. With this in mind here are some factors that have been re-occurring considerations in the functioning of Explore:
    1. Excessive “social” activity on your part may be detrimental to your chances of getting into Explore. (i.e. spamming favourites to all your contact’s images)
    2. Low “social” activity on your part may be detrimental to your chances of getting into Explore. (e.g. not commenting on other people’s photos)
    3. Certain groups, especially mandatory award groups, may be detrimental to your chances of getting into Explore.
    4. Adding your image to too many groups may be detrimental to your chances of getting into Explore. (current wisdom says 2 is safe).
    5. Geo locating your images (i.e. placing on the map) may improve interestingness therefore the chances of getting into Explore.
    6. Placing watermarks on your images may be detrimental to your ability to get into Explore (very few images in Explore have any kind of copyright notice or watermark)
    7. It does not appear that tagging affects interestingness directly, but it may entice people to visit your image via search.
    8. Who is doing favourites and comments on your photos and where those individuals have arrived from may affect the image’s interestingness. (e.g. non-contact activity being weighted more than contact activity)
    9. Images without EXIF data may be at a disadvantage for selection into Explore.

    There is also some evidence of a random factor. You get hints of this randomness when you see a dozen pictures of buses in Explore one day and none the next. This could also be seen as evidence tags do matter from time to time. In either case one should conclude there is enough complexity in the interestingness algorithm that no image will be a guarantee for getting into Explore.

    The changing nature and perhaps randomness of Explore should not be viewed as unfair. Flickr has elected to use the social aspect of its site to make the selections for that exposure. It should be no surprise then that your social group (i.e. contacts) will impact whether you can get into Explore. Ideally your contacts are routinely active, tend to use favourites as a vote of support vice comments, and respond to your uploads spontaneously (i.e. they do not wait til you fave first). The rationale for this mythical contact is engendered in the factors highlighted above. The simple fact of the matter is you want to receive as many favourites as you can in as short a time as possible. The comments are just gravy, and like gravy too much will spoil the meal.

    The vast majority of people who complain Explore is unfair miss the simple fact they are in control, to some degree, whether their image makes Explore. It is through cultivating good contacts and being an active participant in Flickr that sets the foundation for Explore success. A beautiful image with 5 faves and 300 award comments is not going to get into Explore. An image with 5 favourites and one comment is actually more likely to make it into Explore than the preceding example, not because it is better looking but rather because it meets the current selection criteria. While this latter image won’t likely appear as Explore #1, it might get Explore #345. Whether it stays there will be largely predicated on the response it gets from other Explore users vice the author’s contacts. In other words: your contacts are generally those responsible for getting you into Explore; whether you stay there will depend on how interesting your image is to the wider, and typically much larger, Explore audience.

    When an image gets and remains in Explore, a window can be considered to have opened. During the next 14 days your uploads may continue to get into Explore. The window will close on the 14th day if you have too many images still in Explore. What “too many” is exactly is open for debate. Observations indicate it can be as low as two. Regardless, once the window closes you are effectively banned from Explore for the next 6-8 weeks. During the ban none of your uploads will be displayed in Explore even if it is universally loved and favourited by thousands of people. The bans are temporary in all but the rarest of cases. When the ban is lifted the slate is wiped clean and you are able to open that window again with your next upload.

    There are those who may seem to you to be exempt from being banned. This is actually quite a complex issue and not one I am prepared to discuss here.

    Some people will chafe at the use of the term ban. Others will be fume at what they consider an arbitrarily unfair exclusion. However, if you remember that Flickr is a social(ist) experience you may consider the ban a form of wealth re-distribution :-) For Star Trek folks, Flickr has decided the needs of the many outweigh the needs of the one.

    Care to see the actual Flickr Explore “interestingness” patent? See link:

    appft1.uspto.gov/netacgi/nph-Parser?Sect1=PTO1&Sect2=…

    .
    Originally posted at 12:53AM, 1 November 2013 CET (permalink)