GuruShots Survival Guide (Part III)

This is what some peo­ple were wait­ing for, the “get­ting rich quick” section.

Hold your hors­es! There is no quick and sure­fire way to get to “Guru” sta­tus if that is why you are read­ing this. Sorry.

But I can list a cou­ple of things I found use­ful to do to increase votes and to avoid time- and mon­ey wast­ing. Here we go.

This sec­tion is only about try­ing to win a chal­lenge. It is about game strat­e­gy (remem­ber, Gurushots is a game). It is not about how to get bet­ter at pho­tog­ra­phy. If you don’t have enough decent qual­i­ty images to put into chal­lenges, then you need more luck to get to the top rank. Hav­ing lots of great qual­i­ty images def­i­nite­ly helps.

Dis­claimer: I am like a blind talk­ing about col­or. I haven’t won a chal­lenge yet. But I was close (#4 place once), and I have 71 Guru Picks under my belt. So what I present to you are pieces in a puz­zle that con­stant­ly changes the image you want to assem­ble as Gurushots changes the soft­ware and algorithms.

Two ways to win

Every nor­mal chal­lenge has two dif­fer­ent modes of winning: 

• get the most pub­lic votes for one image, or for a set of images (2 or 4 image sets), or both
• get the top “Guru Top Pick”

Your strat­e­gy for win­ning one or the oth­er would be quite dif­fer­ent, as you tar­get dif­fer­ent decision-makers.

Let’s have a more detailed look at the two sec­tions and the win­ning conditions.

Most Public Votes

There are two sub-types here, the sin­gle-image speed chal­lenges, and the mul­ti-image set chal­lenges (2 or 4 images). The sin­gle image chal­lenges only have two win­ners, where­as the oth­er chal­lenges have three win­ners because there is also a win­ner for the most votes across all images you put into the set of 2 or 4 images.

They all have in com­mon that you need oth­er play­ers to vote on your images, not the guru who cre­at­ed the chal­lenge. That brings a whole new dynam­ic when you try to opti­mize your win­ning strat­e­gy. Basi­cal­ly, to max­i­mize the votes per view you want

• high pro­file play­er (mas­ter, guru) with high vot­ing pow­er (8 or 9) to vote
• a filled expo­sure meter at the right time
• a guru pick (50 votes) helps
• more views (boost­ed)
• images that appeal to the mass­es and are on-topic

These are the main fac­tors you need to opti­mize to get a good shot at win­ning the pub­lic vote in a chal­lenge. Keep in mind that the sys­tem tries to give each image in the chal­lenge the same amount of views, regard­less of when it was entered.

Let’s start with the first bul­let point. How do you attract high-pro­file play­ers, the “Gurus” and “Mas­ters”, to vote for your images? Let’s take a step back and look at how a vot­ing ses­sion in a web brows­er works. When you open the vot­ing screen, the sys­tem has deter­mined a set of 200 images to dis­play to you. As they nor­mal­ly do not fit all on your screen at once, they are only loaded as you would be able to see them on your screen. So a small­er screen might fit six images, then the first six images in that 200 image batch are loaded and dis­played. Is it ran­dom in which order the 200 images are pre­sent­ed? I believe not. That is where the expo­sure bonus meter comes into play. I believe the order of the images in the vot­ing ses­sion is most­ly deter­mined by their hid­den expo­sure meter val­ue. So the first cou­ple of images you see in a vot­ing ses­sion have a high expo­sure meter val­ue, where­as the last images in the 200^er batch have prob­a­bly a low expo­sure meter bonus value. 

Stay with me. Going back to how the vot­ing ses­sion works. Once you see the six images on your screen you then vote on the images you find wor­thy of your vote. You scroll down, the next images are loaded and you repeat the vot­ing until you have seen all 200 images, or you have decid­ed to end the vot­ing ses­sion some­where in between. You then click on ‘sub­mit vote’, and at that point, the list of seen images is sub­mit­ted back to the serv­er, togeth­er with the list of images you have vot­ed for. 

How do you vote? Do you take your time for each image, eval­u­ate it enlarged to full-screen view, use a col­or-cal­i­brat­ed mon­i­tor, think about if an appro­pri­ate com­po­si­tion has been applied, is it sharp where it should be, is the hori­zon straight, are sen­sor spots vis­i­ble, is the aspect ratio fit­ting for the sub­ject, where are lead­ing lines and does it have a good sto­ry?
Or do you take a split sec­ond to make a deci­sion on what to vote for in a rapid-fire bam-bam-bam way? To get it done as quick­ly as pos­si­ble, as you still have 20 oth­er open chal­lenges to vote on?

Now think about what most high-pro­file play­ers would do. Ide­al­ly, they would do the qual­i­ty vote style, but real­is­ti­cal­ly they do the quan­ti­ty vote style. That means they tend to get fin­ished as quick­ly as pos­si­ble, so they vote for a lot of images at the begin­ning of the vot­ing ses­sion (maybe the first 50–100 images), that is the quickest. 

Does that ring a bell? You want to be in the first half of the vot­ing screen to get vot­ed on by gurus.

What else do the expe­ri­enced play­ers do? They vote right at the end of the chal­lenge, min­utes before it ends. I can’t real­ly explain why that is, I can only guess. But I have seen it over and over. If you are in the lead 10 min­utes before the end of the chal­lenge, chances are you will not win the chal­lenge, and that is espe­cial­ly soul-crush­ing. I have been there. So try again next time to be at the top a lit­tle bit lat­er.
Hav­ing a full expo­sure boost meter at the end of the chal­lenge is key in my opin­ion. It makes sure your image is exposed to the full poten­tial. What do I mean by that?

It has to do with the num­ber of play­ers par­tic­i­pat­ing in a chal­lenge grow­ing over time. The most poten­tial views your image can get are right at the last sec­ond because then the max­i­mum num­ber of play­ers in that chal­lenge is reached. But obvi­ous­ly, the vot­ing can’t all hap­pen in the last sec­ond, so you take the sec­ond-best option by expos­ing your images as late as pos­si­ble, and have the boost acti­vat­ed, but still with enough time before the chal­lenge ends, to give the play­ers time to vote for your image and get all the views (and hope­ful­ly votes) it is enti­tled to.

So to get the most bang for your buck you want your images displayed

• first on a vot­ing screen
• short­ly before the chal­lenge ends

To get both done at the same time, you want to fill your expo­sure meter a cou­ple of min­utes before the chal­lenge ends. How long before is an art form I haven’t mas­tered yet. It depends on the type of chal­lenge, how many votes it has, and when the chal­lenge ends. If you do it too late, you miss poten­tial votes. If you do it too ear­ly, you might get more not-so-high-pro­file-play­ers in the vot­er mix and you lose a few poten­tial views com­ing in at the late game.

Now let us cir­cle back and re-iter­ate: The sys­tem tries to give each image in the chal­lenge the same amount of views, regard­less of when it was entered.

Imag­ine you enter an image at the begin­ning of a chal­lenge. Your images get dis­played to vot­ers over time until they reach the same amount of views as all the oth­er images in the chal­lenge (that is a mov­ing tar­get as new images are entered by play­ers into the chal­lenge over time). But if you enter ear­ly, you get a lot of views from play­ers that have only low vot­ing pow­er. You want to avoid being exposed to these low-pow­er vot­ers. Why? Easy cal­cu­la­tion: if your image would only be dis­played 500 times to a begin­ner play­er (1 vote pow­er) and all vote for your image, you get 500 votes. If your image would only be dis­played to gurus and all vote for your image, you get 500 times 9 vot­ing pow­er= 4500 votes. In both cas­es, you got 500 views. But the votes tal­ly is dras­ti­cal­ly dif­fer­ent (500 vs 4500).

So you def­i­nite­ly want as many high-pow­er play­ers as pos­si­ble. How do you do that? The best way is to enter late in the chal­lenge. That is prob­a­bly one rea­son why the votes of high-pro­file play­ers are accu­mu­lat­ed towards the end of the challenge.

Anoth­er way to avoid the view ‘leak­ing’ dur­ing a chal­lenge is to wait until the first flur­ry of votes is gone, then swap it with a place­hold­er image, and then swap it back in at the end of the chal­lenge (“dou­ble-swap”). That way you do not have to use a ‘key’ to enter late, but instead, you have to use two swaps. The swapped image keeps the views it had at the point in time when it was swapped off. But as the view tar­get ris­es dur­ing a chal­lenge, a ‘view-gap’ is cre­at­ed for the off-the-chal­lenge image. When you swap it back into the chal­lenge, the sys­tem tries to close that gap and puts it in many vot­ing ses­sions. It is sim­i­lar to if you would put in a fresh image not used before in that chal­lenge. If that is at the end of the chal­lenge, you get exposed to high-pro­file play­ers as men­tioned above.

You still can’t force gurus to vote for your image, or pre­vent new­bies to vote for your image. You can only try to shift the bal­ance to what you believe is in your favor.

One note about the cost of the strate­gies. If you are low on resources, typ­i­cal­ly low on unlock-keys (as you only get a free key every 7 chal­lenges you par­tic­i­pate in), you can instead enter ear­ly with a place­hold­er image. Near the end of the chal­lenge, you swap that image to the image you real­ly want in the chal­lenge. That way you use a swap instead of a more expen­sive key. Is that as effec­tive as the dou­ble-swap strat­e­gy men­tioned above? Don’t know. The sin­gle swap at the end takes prob­a­bly more time as the image has to catch up more views. It might be hard­er to time cor­rect­ly. But it only cost one swap instead of two. I think it is worth test­ing out.

We have talked about the expo­sure bonus meter a num­ber of times, and how I believe it works. But we also have to know when to use it and when not to use it.

That is one com­po­nent that needs exper­i­ment­ing with. For exam­ple, is it enough to only put your expo­sure meter to the max at the end of the chal­lenge (and boost, if that is avail­able)? In the­o­ry, that should give you the max­i­mum expo­sure with the min­i­mum vot­ing time per chal­lenge. I need to put more test­ing into that to ver­i­fy if that is a valid assump­tion. Also, a lot of test­ing is need­ed to find out the best remain­ing time for dif­fer­ent kinds of chal­lenges to start vot­ing (1, 2 or 4 image chal­lenges, amount of votes already cast, end time rel­a­tive to Euro­pean prime time, exhi­bi­tion chal­lenge or not?). And then there is the boost. Do you turn that on at the same time you get your expo­sure meter to the max? Or wait until your poten­tial view count is reached, and then increase it by boost­ing the image? Does it make a dif­fer­ence? I don’t know yet. I will find out, eventually.

To sum­ma­rize:

To get a high pub­lic vote count,

• enter as late as possible
• have the expo­sure bonus meter filled at the end
• wait unti­ly­ou know which image is your best one, then boost that, also as late as possible
• if you want to enter ear­ly to avoid using the late entry unlock key, put in a place­hold­er image and replace it short­ly before the chal­lenge ends. This way you exchange the expen­sive unlock key for a cheap­er swap. In this case you need to give it a bit more time, as the replace­ment image needs some time to catch up to reach the same lev­el of views as all the oth­er images in the challenge.