HOME | DD
Published: 2012-03-11 15:45:12 +0000 UTC; Views: 33540; Favourites: 235; Downloads: 0
Redirect to original
Description
body div#devskin0 hr { }
deviantART is a categories game.
According to deviantART's hq, "deviantART's algorithm for Popular browse results (ie. front page) is known as Fair Exposure. The goal is to maximize the display of the most popular deviations in a variety of topicsβ¦with Fair Exposure, the community is exposed to a varied cross-section of deviantART. "
What this boils down to is that certain art categories have inherent ranking advantages and others inherent ranking disadvantages on front page as a means of ensuring that a variety of works from across deviantART are visible on front page and not just that of one category. Sounds pretty good right?
Well let me say before I continue that I do not know deviantART's actual mathematical algorithm for front page ranking. However, I think that anyone who notices the composition of front pages (whether 8 hour or 24, main or by category) and the statistics of the pieces that populate those fronts pages can give you a pretty good guess.
You may have noticed deviations from weird categories such as Customization popping up on front page despite having relatively "few" favorites. I say relatively "few" as being less than 100 in a 24 hour window, which is quite a lot but nothing compared to its neighboring deviations on front page from categories such as Digital Art. This suggests that Customization is a "non-competitive" category. As you can probably guess, not many people submit to the Customization section. Most things there are screenshots or webcam pictures. But because of deviantART's Fair Exposure algorithm, there is a "spot", in a way, on front page for Customization. No, this spot isn't fixed as the number 5 slot or anything. But if a piece gets 100 favorites in Customization, it may be ranked 10th on front page to ensure that front page does have at least one Customization deviation as a part of Fair Exposure. Whereas if a piece gets 100 favorites in Digital Art, it'd be nowhere to be seen on front page and probably not even on the front page for the Digital Art category. Given this, Digital Art can be seen as a highly competitive category. So sorry digital artists, I guess you'll just have to work that much harder to get your work exposure. Doesn't seem as fair now does it?
As you can probably guess, there are also ranking advantages and disadvantages within categories. Take for example the Photography category, which I'm most familiar with. A deviation with 100 favorites in Photography/Macros/Nature will not rank as high as a deviation with 50 favorites in Photography/Macros/Objects and otherwise comparable statistics. Perhaps this is because there is another photography section for nature already, presumably Photography/Animals, Plants and Nature. I'm not really sure the "reasoning" is here. But what this seems to suggest is that if you have a nature-related photograph, perhaps it's better to avoid the Photography/Macros/Nature section and instead go for Photography/Animals, Plants and Nature. Or if you want to be a real player, submit it to something like Photography/Horror & Macabre (a surprisingly non-competitive subcategory for Photography). Or maybe even Customization! Who cares? deviantART sure doesn't seem to. So is this still fair?
Is deviantART's algorithm really this simple? Yes and no. The algorithm seems to also takes into consideration a number of things like the "age" of the submission, the "rate" of favorites (presumably both the number of favorites per view rate and number of favorites per time unit rate), the number of views, the number of comments, and so forth. But categories are a big part of the algorithm. And therefore a big part of deviantART's categories game.
But don't take my word for it. Read about deviantART's Fair Exposure algorithm which "explicitly attempts to select popular deviations that showcase as many popular topics as possible" here: hq.deviantart.com/journal/Popuβ¦
(So this is how popular deviants became popular?!!? Haha not really. deviantART only came out with this algorithm earlier this year. So I guess I'm actually a bit of a slowpoke. But maybe now the changes in front page trends since the beginning of this year make a bit more sense?)
If you say that you don't care about front page or favorites, good for you. Be above the machine (serious here, not sarcastic). If you say that you don't mind a bit of competition, even better.
But a lot of us are trying to get our artworks out there for the greater community to view and enjoy. And it's discouraging to have your artwork be put at a "disadvantage" simply because of the category its in (by this weird kind of art discrimination?). If you're a nature photographer, you'll simply have a harder time getting exposure than a horror photographer, not necessarily because people like nature less and horror more but because the system favors horror. And it's even more enraging when you see others take advantage of this system by playing the categories game while others are left out in the dark.
So now if you want to play this game, go ahead. Now you know. If we're all informed then we're all on even playing grounds then that's what's fair...right?
Should you play the categories game? I'd recommend it. It's a good real world non-science application of the scientific method. Plus, if you think everything I've said is complete bull then try it yourself and see. So submit something. See where it ranks. (Maybe on the 5th page of a subcategory. The deviation doesn't have to be on a front page but you do need to be able to locate it and see where it ranks.) Change the category. Then wait half an hour or so (there's a delay in switching categories and all) and see where your deviation ranks under the new category. Observe the consequences, draw a correlation, and deduce the underlying pattern. Then write a rant-y journal like this one about what you found.
Do I play? At least enough times to make these boisterous claims anyway. Do others already play? I've seen many many people play "discretely" by switching their deviation's category between similar categories. For example, if your deviation is ranked first in the Photography/People and Portraits subcategory but then a more popular piece "overtakes" your spot as first, perhaps you will change your deviation's category to Photography/Miscellaneous so that it can be ranked first in the Photography/Miscellaneous category and therefore rank higher in the front page of the Photography category and subsequently the main front page. I've also infrequently seen others play more outrageously, blatantly submitting to the wrong category to gain an advantage. Traditional Art submitted into Photography/Miscellaneous. Photography submitted into Literature (a very non-competitive category mind you). You name it.
So why doesn't deviantART try to prevent this categories game? If you've ever tried to report a deviation for Misplacement (or anything), you probably know how slow and non-responsive deviantART can be. And who can blame them? I'd say 90% or more deviations on deviantART are misplaced. Most just don't ever make it to where they're visible. Other times, it's really hard to say who's playing the game. Who's to say this photo of a flower isn't really "Darkroom" (another fairly non-competitive subcategory for Photography)? Or even if you see someone switch categories every few hours, who's to say they didn't just realize their piece "belonged" in a different section. Maybe they just made a mistake? Of course, it happens. But maybe deviantART doesn't care because people don't seem to care. Maybe there are no reports to begin with. Or maybe people don't even notice the category something's in once it's on front page. Or more likely a bit of both.
So what should deviantART do? Perhaps a change of algorithms is due. deviantART's old algorithm based on raw popularity didn't seem to work too well since those who were popular simply gained more exposure and became more popular while those who were new couldn't even get their foot in the door. deviantART's new algorithm actually seems to favor what's new and is gaining momentum quickly (so a high rate of favorites) though this has a tendency of favoring fan-art, porn, stolen works, or whatever else has a large immediate wow factor. So what algorithm would be better? I honestly have no clue.
Then why am I complaining if I can't fix the problem? Eh, I'd rather think of this as disseminating information than complaining.
Take what you've learned here and do what you want with it. Judge me for it or do as I've recommended you do: play the categories game.
May the best artists players win?
Related content
Comments: 316
RosesInTheFlames [2023-02-06 15:30:57 +0000 UTC]
π: 0 β©: 0
Frenchs-Asylum [2022-09-22 02:36:21 +0000 UTC]
π: 0 β©: 0
dekelpoliva [2021-02-18 16:35:11 +0000 UTC]
π: 1 β©: 0
Kaldrinn [2020-01-18 16:49:14 +0000 UTC]
I wonder if there's even, a "good" algorithm on any web platform, if this is even possible to please all sections and artists and give them equal chances
It's just a natural process that popularity means more popularity, friends tell their friends etc (Switch that with retweets and journals sharings, same stuff)
Honestly have no clue
What about an algorithm who judges the quality of a piece and then decide to put it front page ?
I've seen very mediocre things on the font page of DA somedays, porn or multiple 3d renders of scenes the person has alreayd 56 screenshots of in their gallery, but I guess it's hard for a bot to see this type of thing, even maybe impossible.
π: 0 β©: 0
CatzK3 [2018-02-27 14:00:21 +0000 UTC]
THis entire site is such a load of horse shit...Β
π: 1 β©: 0
rei-kaa [2018-01-30 21:15:45 +0000 UTC]
This categroy thing is annoying me now, as itβs unfair for people who actually works hard and follow the rules who put their works in the correct category. Reporting does very little and there isnβt really punishment to people who obviously does that. There should be some punishment to users that constantly misplacing categories.
π: 0 β©: 0
itzikgur [2017-07-27 19:49:14 +0000 UTC]
many thanks. now I understand thisΒ algorithm. in my caseΒ Photography / Urban & Rural / Industrial with 100 fav apears in front page a long time, while other photography categories with more favs not.
π: 0 β©: 0
longestdistance [2017-03-06 00:35:50 +0000 UTC]
Apparently because of the Wix partnership Deviantart may be getting more staff members. We can only hope this means something will be done about the miscategorization problem on Deviantart. A lot of the most popular DA artists do it and it disgusts me. The whole point of the new algorithm was to give niche and lesser-known artists a fair chance at exposure. The popular artists were going to get on the front page anyway so why do they have to miscat? I think it says a lot about their integrity; they're willing to screw over other artists for their own selfish means. And personally that is where I draw the line and decide not to support them as an artist, no matter how talented and skilled they are otherwise.
π: 0 β©: 0
TheMadMoustache [2016-09-30 23:35:49 +0000 UTC]
I'm new to Deviantart, so I don't really have much experience ere, but this does make a lot of sense. Good writing, too.
I'm just keeping my page as a hobby, and (probably) eventually going to get 'competitive'.
And indeed, may the best players win. Β Β
π: 0 β©: 0
Angustus [2016-03-07 01:06:42 +0000 UTC]
Very disapointed in DA. they should advertise talent, but they rather show off mediocrity... way to go.
π: 0 β©: 0
TrappedGirl [2015-04-14 11:04:14 +0000 UTC]
If I made one of those visual art made up of text it should qualify for the literature category I bet hehehe
π: 0 β©: 0
Arthur-Ramsey [2014-04-30 20:02:21 +0000 UTC]
It is mostly about how many watchers you and what category and how many views. You can have no favs at all but if you have the rest your work will be seen.
π: 0 β©: 0
Xiaooyu [2013-05-11 17:02:39 +0000 UTC]
Ahh.
I'm always wondering about how some types of artworks are always up in the front page.
I guess maybe now people will cheat and put their digital art as "customization" or something? Who knows. I think the best way to get on the page is to have a high number of views, favourites and comments. At least its fair// '._.
π: 0 β©: 0
The-Golden-Knight [2013-02-03 19:02:45 +0000 UTC]
Revisiting this; 100 favorites is a lot. I'm lucky to even get 5!
π: 0 β©: 0
bib993 [2012-11-13 00:08:16 +0000 UTC]
Interesting. You forgot to say that the algorithm also seems to take into account the number or frequency of your submissions. I submit in the fractal category only. Recently I became very popular (don't know why) and all my deviations now almost instantly make it to the 24 hours front page (I guess thanks to Fair exposure stuff that want to show at least 1 fractal image on the front page).
But now I became "so" popular (more than 20,000 views every time in less than 48 hours) that my deviations do not appear anymore on the 3 days, 1 week and 1 month popular pages, like if I've been blacklisted, too many successful images, sorry man, you out of here, leave room for the youngster! However, my older deviations with much less views, favs (and lower fav/views ratios) are still there. Very weird!
π: 1 β©: 0
invictuzz688 [2012-04-07 04:21:28 +0000 UTC]
lol! i know this to be true, it makes sense!
but i see no point of playing the "game".
deviantart is a bit more of a repository for my works. an online vault for my portfolio. thats it.
thanks jean for putting this up
π: 0 β©: 0
ElenaHelfrecht [2012-04-01 13:51:56 +0000 UTC]
Well, I think you're quite right with this and captured this system in a very clever way. But I think the good thing with this is, that you get a greater band width of different categories. We all know that Nature photographs are more socially acceptable than Horror ones. And this adds a rich diversity of art to the frontpage... but that's just what I as a Horror-artist think.
π: 0 β©: 0
BrielleCoppola [2012-03-23 10:23:32 +0000 UTC]
You're a very clever person, and an excellent writer.
π: 0 β©: 0
MidnighBlood [2012-03-22 13:56:48 +0000 UTC]
This blew my mind... o.o I would've never realized this!
π: 0 β©: 0
avalanchepark [2012-03-16 03:08:52 +0000 UTC]
little bit of chicken and egg here - in any system based on favs you need to get them to get on the front page to be seen to get them .
If they wipe out all the preference for less popular categories then the 10 bazillion people competing in the most popular one would have twice as many slots (say) in the first 10 pages (or however many you think get searched in the popular gallery). You folks are still out of luck !
The fact that any fav carries equal weight also sort of insures that 'serious' artists here will always be dissatisfied with who makes the top if it is a pure fav contest.
Some long time ago I tried to make a suggestion that they use some sort of social model that would look at the images an individual had fav'd , then look at all the other people who have faved each image , finally creating a page of art this pool of like minded viewers had fav'd . Sort of the Amazon "if you like this book you might also like ..." approach. A personal popular page based on the opinions of people whose opinions I might actually care about. I often go through the fav galleries of the artists I follow as one approximation of this approach.
And in the end I spend a LOT of time cruising the newest gallery because it gets me past all the screening and gives a chance for new discovery that isn't 30 pages back in some imperfect ranking system. Near as I can tell my tastes must be unique because I go through a lot of the popular gallery and it doesn't speak to me that often........
π: 0 β©: 0
M0THart [2012-03-14 21:58:35 +0000 UTC]
i'm so happy you wrote this! i've noticed something, but never try to understand better the situation, and you really made it! huge respect for you- i'm in the same situation as yours (since i follow you for years by now, i only completely agree with you)
π: 0 β©: 1
Argolith [2012-03-13 22:13:32 +0000 UTC]
I believe there are good ideas behind this so-called Fair Exposure algorithm... Such as making sure the front page is not completely filled with Twilight fan art, thus giving the impression that the whole dA member base consists of people who never made it out of puberty. However, the implementation may be far from ideal, and I've also noticed a lot of -- possibly intentionally -- misplaced deviations. Anyway, I never paid much attention to the front page, so I don't really care about this algorithm.
π: 0 β©: 0
AnnieGitUrGuns [2012-03-13 18:37:10 +0000 UTC]
At least your photos are seen, Jean. You are very popular on here due to your talent and I am sure this has lead to you getting better and better (or at least helped a smidgen). Maybe I should play with game of categorization, because I don't get much feedback and it is hard to get better if you get no feedback.
But it is very unfair in my opinion. Maybe Deviant Art is too big and they have no other way of getting people exposure. I don't know. Maybe it is time for people to join more specific art sites. I might write a rant as well. lol
π: 0 β©: 1
JeanFan In reply to AnnieGitUrGuns [2012-03-13 20:45:24 +0000 UTC]
Haha that's brings up an interesting question: which came first, the "talent" or the popularity? ie. does "talent" lead to popularity or does popularity bring out "talent"?
When I sift through other people's comments, I rarely see any constructive criticism or helpful feedback. So I don't think you're alone in not getting much feedback. I don't get much feedback either because, really, it's hard to get or give helpful feedback. People are vague.
So if you want to improve, you need to be your own biggest critic. Not wait on feedback from others
You should try joining a group. They're like specific art sites-ish. I think that may help
π: 0 β©: 1
BabylonSabby In reply to JeanFan [2012-03-16 00:00:29 +0000 UTC]
That shouldn't be to say that feedback doesn't help. If ever a person -does- get it, it should be handled first of all: maturely. If it's negative feedback, you shouldn't respond like, "Oh, yeah? Well fuck you! This is what I draw and I'm not changing it." Instead, you should handle it with maturity and really take in what the critique is saying, because it can be very helpful if and when you receive it. I see way too many people receiving constructive criticism and throwing a fit over it. Ticks me off. Until then, though, you're right about really being the only critic you've got.
Speaking of more specific art sites, do you know of any? I have the same issue. I've got lots of page-views, but compared to how many things I have in my gallery, it feels like I could have more. It's hard to get especially noticed in this community for my original stuff, especially. I get way more views, comments, and favorites on fanart when I'd actually like to be known more for my personal, original art. Think I should start trying to get my name out in another community somewhere?
π: 0 β©: 1
JeanFan In reply to BabylonSabby [2012-03-16 00:15:17 +0000 UTC]
I don't think it matters whether you're in dA or any other art community, focus on getting better at your art and the views will naturally come.
π: 0 β©: 0
Satsumo [2012-03-13 15:48:34 +0000 UTC]
It is an interesting journal piece. The thing I will take away from it is that some people take the pageviews/ranking/popularity thing way too seriously. I view it like the number of friends people have on Facebook. I use DA as a place where I can see and discuss pieces of art that interest me. The only reason I might use the front page is if I'm researching something and I'm looking for a particular subject of art. The content of the front page is popular in the same way as Justin Beiber and Twilight, neither of which I'm burning to know more about.
I applaud what DA have done to try to make the front page more diverse. Its rare to see any site that isn't rushing toward the lowest common denominator in search of clicks, Facebook shares or whatever. If some people are so needy for ego stroking gimmicks that they will distort that system then they have a problem, not the system. Besides, people will always find a way to abuse a system for personal gain, the only way to defeat their game is not playing.
π: 0 β©: 0
moontearfall [2012-03-13 14:21:29 +0000 UTC]
I'm not an artist. I do like looking at deviations, however, and I favorite the stuff I like the most. That said, I have almost never really liked anything that shows up in the "popular deviations", so my method of looking for new stuff is either the "discover" feature, or looking directly at other deviants' favorites (which is the same thing, really).
So I think the algorithm is irrelevant, at least to me. I understand that it could put some artists at a disadvantage, but there will always be people looking at all deviations, adding them to their favorites and then giving them exposure by word of mouth, or just by having them in their favorites and collections.
π: 0 β©: 0
RoxyCockerSpaniel [2012-03-13 14:11:52 +0000 UTC]
Holy shit.
True.
HAI GAIZ CHEK UT MAI GALLARY!!!!!!!111111!!!!!!!!11111!!!!!!!!!11111!!!!ONE!!!!!!!1
I did that "chek ut mai gallary" thing on purpose.
π: 0 β©: 0
eBulwiaseK [2012-03-13 14:04:33 +0000 UTC]
How to win?
It's simple. Join a lot of groups, fav all of artworks, watch peoples. Someone sure will watch you. Then add your deviations and... tadamn!
π: 0 β©: 0
Vaporeon249 [2012-03-13 10:50:10 +0000 UTC]
If you think the Front Page - Popular In the Last * hours is bad, try the Search by Popular field. It's so bunked.
π: 0 β©: 1
JeanFan In reply to Vaporeon249 [2012-03-13 15:45:32 +0000 UTC]
Search is more so by keyword relevance than by popularity even if you search by Popular field The algorithms are completely different from front page ranking.
π: 0 β©: 1
Vaporeon249 In reply to JeanFan [2012-03-13 16:04:11 +0000 UTC]
If they were the same, then every last anime related search probably would result in a lot of craft-related things.
π: 0 β©: 0
Limaria [2012-03-13 09:45:01 +0000 UTC]
I only browse photography so I don't know about the other categories.
I don't care too much about blatant miscats like anime screenshots or scribbles because I just ignore them. What kinda irritates me though is the amount of photos from the most abused categories like "glamour portraits" and "artistic nudes"*. Before the change, those shots were usually left deep down in the abyss but now... I sure wish that galleries prone to this sort of stuff would get some weeding and pruning.
Anyway, I really don't know why dA went back to this system since it's causing precisely the same problems as several years ago.
* Blurry party pics are not "glamour" and there's a good chance that shoving a cellphone up your private parts produces photos that are probably nude (if the lens gets foggy, maybe your camera is too embarrassed) but let's not talk about the "artistic" part... That also includes that I don't want to know that you do not have matching covers for your bed, that you are too lazy to vacuum or that your cat just had a bit of fun with the waste basket. On the other hand, it does distract from the amateur posing.
π: 0 β©: 0
NibaAudreyZebra [2012-03-13 08:18:49 +0000 UTC]
my gpa feels like a pice of shit compared to this
π: 0 β©: 0
BlissfullySarcastic [2012-03-13 07:21:24 +0000 UTC]
I...want to start with this:
My opinion is going to be biased, considering I DO post to the Customization category (legitamately, not playing the game, as you'd say) as an active emoticonist on this site.
While I do agree with a lot of points you are making, I DO find it fair that some categories are favored over others. It gives people who are a part of a small category to begin with a chance to actually SHOW people that their category is at least worth glancing over. Hopefully, because there are things like Fair Exposure, the category will grow because more people will be aware it's there (same things DDs for separate categories do--bring exposure to new artists, new pieces, and the category as a whole). If the smaller categories become larger because people know they're there and active, then eventually it'll become less weighted towards one category or another because each category would grow. If people REALLY find placing into the Customization category is a way to make them become more popular on the site, they will place more work in the Customization category and devout less time towards larger categories, tipping the scale more towards the Customization category and making it slightly harder each time that scale does tip to get on the front page.
I do not think it's fair that people could change categories so often, but that's more to do with the way the categories work. If there was a limit to how often you could switch categories for a deviation or something similar in place, then maybe the flip-flopping would be reduced, but then they could just place in the wrong category from the getgo. That is not a problem I would know how to fix, but it's because it's not an easy problem TO fix.
But strictly from the stand-point of someone who is from a small community on this site:
We need whatever boosts we can get. We don't really HAVE that big a community to bounce work back and forth with, a lot of people to compare with, et cetera. We are a small community with limited amounts of people, which, if you saw only DigiArt was getting exposure, for instance, you would never really want to step foot in. We cannot grow if no one knows where we are or how to find us. The most popular artforms will be the ones getting the most views, most favs, most comments, et cetera, because, honestly, if you're a digital artist, you will be a lot more likely to post on a fellow digital artist and have friends who are also digital artists. Even if you AREN'T a digital artist, it's a huge category to explore with a lot of people and a lot of artwork, so you can easily find yourself commenting/browsing there, so it'll just continue to grow as a powerhouse category and the littler categories, such AS Customization will just rot in a corner with no one new to go into it because they simply won't KNOW about it. I didn't know the Customization category existed at all until I saw `CookiemagiK 's, `LeoLeonardo 's, and `Synfull 's emotes, which made me START getting into emoting.
So yes, I can see how you think it's unfair, as people can use it to their advantage. But for those who do not use it for their own advantage and for those who make art and place it legitmately without the leverage, it really does help the smaller categories out a TON. It's like you said: I see the problem you stated, but I really do not know how to fix it. For now, I'm happy with it until it's played around with some more by $taff.
π: 0 β©: 1
SadisticIceCream In reply to BlissfullySarcastic [2012-03-18 21:56:40 +0000 UTC]
As a Lit deviant, 100% agree.
π: 0 β©: 1
BlissfullySarcastic In reply to SadisticIceCream [2012-03-23 04:18:01 +0000 UTC]
Yay for smaller communities!
π: 0 β©: 1
xX-NIGHTBANEWOLF-Xx [2012-03-13 06:41:50 +0000 UTC]
how do you get artwork featured thoguh?
π: 0 β©: 2
MetellaStella In reply to xX-NIGHTBANEWOLF-Xx [2012-03-13 11:41:26 +0000 UTC]
If someone suggests you to an admin, and they choose your work.
π: 0 β©: 1
Vaporeon249 In reply to xX-NIGHTBANEWOLF-Xx [2012-03-13 10:53:02 +0000 UTC]
Actually being unique and decent.
π: 0 β©: 1
xX-NIGHTBANEWOLF-Xx In reply to Vaporeon249 [2012-03-13 12:09:35 +0000 UTC]
Okay, thanks
π: 0 β©: 0
willyamPax [2012-03-13 04:30:18 +0000 UTC]
i learned alot in this journal...well said.....
π: 0 β©: 0
| Next =>