Author Topic: How easy is it to rig the MSG earnings poll? (Read 6557 times)

Sadstock · « **on:** November 11, 2011, 13:48 »

Wonder how easy is it to rig the earnings poll? I would guess that poll results influence where people upload with companies that move up in rankings getting a burst of uploads.

Leaf, is there any data you track or can share that would help the community identify trends that might be a bit suspicious? Graphs over time showing the number of votes per company, so a big spike in votes one month to the next would be easy to spot? Information about the number of voters that only vote for one company?

Know some people don't like/care about the poll, but as somebody else pointed out, what else do we have?

RacePhoto · « **Reply #1 on:** November 11, 2011, 15:17 »

Quote from: Sadstock on November 11, 2011, 13:48

Wonder how easy is it to rig the earnings poll? I would guess that poll results influence where people upload with companies that move up in rankings getting a burst of uploads.

Leaf, is there any data you track or can share that would help the community identify trends that might be a bit suspicious? Graphs over time showing the number of votes per company, so a big spike in votes one month to the next would be easy to spot? Information about the number of voters that only vote for one company?

Know some people don't like/care about the poll, but as somebody else pointed out, what else do we have?

Actually votes per poll or per company would be interesting because it would show how many people sell on those sites. Also maybe not perfect, but it would be representative. Lets say 100 people took the poll in Oct. and 99 of them sell on SS, 90 sell on IS, 80 sell on DT and FT, and 70 sell on DP and BigStock. All made up numbers. When it gets down to some of the smaller agencies and there are 30 people with images there, any one vote can make a difference, because it's more of a factor in the average. With the bigger ones, one vote has a smaller influence.

Then another interesting part of your question, say an unusual number of people vote for one agency and it's not that popular or big, maybe it would show that someone is trying to stack the numbers for that company. One way or another.

But for now, is number of votes showing someplace and I missed it? That would be interesting in itself.

cthoman · « **Reply #2 on:** November 11, 2011, 16:03 »

I'd think you'd need quite a few people to rig it. I've never seen my vote even make a dent in the numbers. That and some of the illustration only sites languish at the bottom despite regularly outperforming sites in the low, middle and top tier.

If you can't even get illustrators to have any influence on it, I don't think a few people with an agenda have much of a shot.

Sadstock · « **Reply #3 on:** November 11, 2011, 17:02 »

Quote from: RacePhoto on November 11, 2011, 15:17

But for now, is number of votes showing someplace and I missed it? That would be interesting in itself.

-----------------------------
Yes, at the bottom of the poll results side bar there is a link "More Poll Results" which takes you to this http://www.microstockgroup.com/index.php?action=pollsresults and it does show the number of votes per micro. But there is no historical data as far as I can see, so I can't tell if the number of votes for a given micro is up 50% from last month.

CD123 · « **Reply #4 on:** November 11, 2011, 17:29 »

Something I did suggest in the past (which is actually easily obtainable data) is a percentage indication per income category. You can then see out of the total number of people supplying an income vote in respect of specific agency, what percentage has indicated 1, 2, 3, etc.

Apart from making very interesting reading, one should be able to pick up if there is a sudden percentage rise in the high income category and or lack of middle income indicators. Just a thought...

Microstock Posts · « **Reply #5 on:** November 12, 2011, 12:23 »

I guess it seems that a lot of people see the polls as really representative of the situation. It is the only type of thing we have to get an idea, but I'm guessing the accuracy at times will be far from accurate. A poll with 1000 people will in general be more reliable than a poll with 100 people, for example.

I've noticed that people have suddenly started to upload to Photo Dune as they see it in middle tier now. However, of the other agencies that appear in the middle tier at present, the amount of people that voted (for earnings) range from 177 to 294 people. However, Photo Dune has just 58 votes, which is a long way off. Bigstock which is just out of the middle tier has 260 votes, maybe if it had 58 votes its position would be much higher or even much lower. Who knows.

Leaf, so that people know how many people voted (for earnings), maybe you could put this figure beside the figure, which shows the percentage which an agency has risen or fallen. It is important for people to see just how many people actually voted.

RacePhoto · « **Reply #6 on:** November 12, 2011, 17:31 »

Quote from: cthoman on November 11, 2011, 16:03

I'd think you'd need quite a few people to rig it. I've never seen my vote even make a dent in the numbers. That and some of the illustration only sites languish at the bottom despite regularly outperforming sites in the low, middle and top tier.

If you can't even get illustrators to have any influence on it, I don't think a few people with an agenda have much of a shot.

If the illustration sites have less voters that doesn't change the amount of earnings. If a site has 50 people and someone votes $10, that's the same as six people on the top sites voting $10.

In other words, one high vote on an illustration site, counts for six times more than one high vote on a top tier site.

294 votes SS
275 votes IS
300 votes DT
275 votes 123RF
294 votes FT
26 votes ClipartOf (each vote counts 11 times what a SS, DT or FT vote counts)
177 votes DP
58 votes PhotoDune
250 votes CanSTock
49 votes Pond5
49 votes Veer
260 votes BigStock

So I might guess that as PhotoDune gets more members and ClipArtOf gets more members, it might level off a little instead of having a small number of very happy members?

DT has more than SS and FT by a small number, but still, it's more.

RacePhoto · « **Reply #7 on:** November 13, 2011, 15:03 »

Part Two Votes Cast November 2011

300 votes DT
294 votes SS
294 votes FT
275 votes IS
275 votes 123RF
260 votes BigStock
250 votes CS
177 votes DP
173 votes Veer
58 votes PD
49 votes P5
26 votes COf

Says something about how popular some sites are? Not much about sales or profits.

Someone screwed up (that's me again) 173 votes Veer not the 49 that I had copied in error.

Mantis · « **Reply #8 on:** November 13, 2011, 17:33 »

Quote from: RacePhoto on November 13, 2011, 15:03

Part Two Votes Cast November 2011

300 votes DT
294 votes SS
294 votes FT
275 votes IS
275 votes 123RF
260 votes BigStock
250 votes CS
177 votes DP
58 votes PD
49 votes P5
49 votes Veer
26 votes COf

Says something about how popular some sites are? Not much about sales or profits.

This is true. My metrics don't necessarily line up with what MSG shows, so there are inherent differences probably based on portfolio type, quantity, quality, and N where N=data points being measured.

Sadstock · « **Reply #9 on:** November 14, 2011, 07:11 »

Quote from: RacePhoto on November 12, 2011, 17:31

Quote from: cthoman on November 11, 2011, 16:03
I'd think you'd need quite a few people to rig it. I've never seen my vote even make a dent in the numbers. That and some of the illustration only sites languish at the bottom despite regularly outperforming sites in the low, middle and top tier.

If you can't even get illustrators to have any influence on it, I don't think a few people with an agenda have much of a shot.

If the illustration sites have less voters that doesn't change the amount of earnings. If a site has 50 people and someone votes $10, that's the same as six people on the top sites voting $10.

In other words, one high vote on an illustration site, counts for six times more than one high vote on a top tier site.

294 votes SS
275 votes IS
300 votes DT
275 votes 123RF
294 votes FT
26 votes ClipartOf (each vote counts 11 times what a SS, DT or FT vote counts)
177 votes DP
58 votes PhotoDune
250 votes CanSTock
49 votes Pond5
49 votes Veer
260 votes BigStock

So I might guess that as PhotoDune gets more members and ClipArtOf gets more members, it might level off a little instead of having a small number of very happy members?

DT has more than SS and FT by a small number, but still, it's more.

------------------------------------------
This makes a lot of sense to me. It would be helpful I think to have the number of votes on the first page. I think some type of chart showing the number of votes over time on the second page would be really useful as well.

leaf · « **Reply #10 on:** November 14, 2011, 08:51 »

thanks for all the thoughts.

Yeah - the number of voters does affect the rating by nature. PhotoDune's sales (according to my stats) have increased every month but their rating has dropped quite a bit this month. Simply because more people (with lower earnings) are voting for the site. That said, where they currently appear in the list on the right is pretty accurate to my own experience.

The reason a site needs 51 votes to appear in the top area with the up / down arrows is exactly this reason - with very few votes a site can easily be 'spammed' to the top of the results. With 51 votes it is pretty hard and if things start to look fishy I'd go in an look at the votes - so far that hasn't ever been a problem. I don't put the number of votes on the main page to keep things simplified. If people want more detailed info, it is easy to follow the link on the bottom of the resutls or the link in the top menu.

Now - after all that is said. I also feel it is time for a reworking of the voting system so people can vote more accurately and a nice history is also shown.

Sadstock · « **Reply #11 on:** November 14, 2011, 12:03 »

Quote from: leaf on November 14, 2011, 08:51

thanks for all the thoughts.

Yeah - the number of voters does affect the rating by nature. PhotoDune's sales (according to my stats) have increased every month but their rating has dropped quite a bit this month. Simply because more people (with lower earnings) are voting for the site. That said, where they currently appear in the list on the right is pretty accurate to my own experience.

The reason a site needs 51 votes to appear in the top area with the up / down arrows is exactly this reason - with very few votes a site can easily be 'spammed' to the top of the results. With 51 votes it is pretty hard and if things start to look fishy I'd go in an look at the votes - so far that hasn't ever been a problem. I don't put the number of votes on the main page to keep things simplified. If people want more detailed info, it is easy to follow the link on the bottom of the resutls or the link in the top menu.

Now - after all that is said. I also feel it is time for a reworking of the voting system so people can vote more accurately and a nice history is also shown.

Leaf, Thanks for the info. I had not realized that you needed 51 votes to get into the up down arrow section. I had assumed that the lack of arrows meant the earnings were so low they were not measurable.

On the second page I'd love to see historical data in a graph or chart so overall trends per micro could be easily seen.

RacePhoto · « **Reply #12 on:** November 14, 2011, 15:43 »

Quote from: leaf on November 14, 2011, 08:51

thanks for all the thoughts.

Yeah - the number of voters does affect the rating by nature. PhotoDune's sales (according to my stats) have increased every month but their rating has dropped quite a bit this month. Simply because more people (with lower earnings) are voting for the site. That said, where they currently appear in the list on the right is pretty accurate to my own experience.

The reason a site needs 51 votes to appear in the top area with the up / down arrows is exactly this reason - with very few votes a site can easily be 'spammed' to the top of the results. With 51 votes it is pretty hard and if things start to look fishy I'd go in an look at the votes - so far that hasn't ever been a problem. I don't put the number of votes on the main page to keep things simplified. If people want more detailed info, it is easy to follow the link on the bottom of the resutls or the link in the top menu.

Now - after all that is said. I also feel it is time for a reworking of the voting system so people can vote more accurately and a nice history is also shown.

And just in case, I have no complaints, it's nice to look at. I never paid attention to the number of voters before. Kind of surprised that DT has the most.

Speaking of voting, how does Paxxion get 16 votes, when the site is invisible and appears to be shut down?

Microstock Posts · « **Reply #13 on:** February 15, 2012, 06:14 »

Quote from: leaf on November 14, 2011, 08:51

Now - after all that is said. I also feel it is time for a reworking of the voting system so people can vote more accurately and a nice history is also shown.

Any developments on this Leaf? MSG is the most popular microstock forum, but for me the the poll on the right is the thorn on MSG's side. People probably do rig the poll, I don't see how earnings on iStock can show as high as they do after the countless reports of huge drops in sales with both non-exclusive and exclusives, for at least a year. Yet it always has a poll level of 6 something, it never drops to 5.

Finally, it is NO coincidence that 123rf got ideas about reducing commissions once they saw themselves in the Top Tier. I mean it was so soon after getting in to the Top Tier on MSG, that they announced that they will be reducing commissions for newbies, they have since announced that they will reduce the commissions for almost everyone, including me. 123rf like the other sites are here almost everyday and they see the poll too and they know how much the contributors think it is gospel. The poll on the right helps to give agencies like 123rf a window for exercising their arrogance and as a consequence we can lose money from this.

CD123 · « **Reply #14 on:** February 15, 2012, 07:40 »

Sorry Microstock Posts, can not say I agree with your viewpoint.

Firstly, you are trying to state guesswork as fact "Finally, it is NO coincidence that 123rf got ideas about reducing commissions once they saw themselves in the Top Tier". Do you have insider info at 123RF? It is as much a co-incidence as it was one that it was the end of the year, when most businesses re-asses their positions. The whole industry has being dropping their commissions for quite a while now. So it is not so OBVIOUS that you can link their change to anything else than following an industry trend.

Secondly, it does not make business sense to change your recipe for success as soon as you start tasting it. There would have been 10x more reason for 123RF not to change what they where doing, based on their success. So deciding to drop their commission could not have been one easily taken.

Thirdly, I think a business like 123RF will trust their own business intelligence, own income and expenditure statements much more than a self admitted "inaccurate" website poll.

In my opinion, the poll is fine an serve the purpose it has been designed for and that is to give a broad guidance of general success for the majority of contributors represented here. Any improvement in accuracy will help and I fail to see that this pole can move a major player in the industry to change its business direction.

Microstock Posts · « **Reply #15 on:** February 15, 2012, 08:24 »

@CD123 Sure, these are my observations, but I believe that msg is of importance to them or they wouldn't spend their time here. I think you give too much credit to agencies for intelligent policy making, when quite clearly a lot of their policies are on whims. We all give 123rf lots of praise for entering the top tier, soon after they announce lowering commissions for new contributors, a few weeks later they scrap that idea and then say well we might as well get more from more of you, cos you love us now. Chopping and changing plans on a whim is not the sign of a well planned agency. The poll on the right has too much influence in the way that contributors and also agencies think. To exactly what degree I don't know, but it is definitely an influence. Hence, I think higher accuracy is imperative, but that depends on so many factors. At the very least to show the number of votes cast for each agency should appear on the main section which everyone sees.

CD123 · « **Reply #16 on:** February 15, 2012, 08:32 »

Will also still love to see the number of votes for every scale category. If, out of 10 votes 7 is on 7 and 3 on 1, it should tell you something.

Jonathan Ross · « **Reply #17 on:** February 15, 2012, 16:43 »

Hi All,

I think what Leaf is trying to accomplish is very honorable and helpful as is everything he does here for all of us. The one issue with these kind of polls is they are generated by people not drawn on from an actual polling system where people are approached directly for their information or gathered directly from agency data, so these numbers I believe are skewed to the high side. The people that " generally " are going to add to this poll are most likely people that do above average in the industry. These are the people that take the time to answer such poles as they are more involved in the process at a professional level. Great info to have but I think you have to take it as a fun addition and track your own numbers more than anyone else's, or get straight information from stock agencies that have very accurate info on this topic. Just my two cents, and thanks again Leaf you rock..

Best,
Jonathan

CD123 · « **Reply #18 on:** February 15, 2012, 23:02 »

MicrostockGroup Sponsors

Author Topic: How easy is it to rig the MSG earnings poll? (Read 6557 times)

Sadstock

RacePhoto

cthoman

Sadstock

CD123

Microstock Posts

RacePhoto

RacePhoto

Mantis

Sadstock

leaf

Sadstock

RacePhoto

Microstock Posts

CD123

Microstock Posts

CD123

Jonathan Ross

CD123

Related Topics

Sponsors

Microstock Poll Results

Sponsors